Broadcom CEO doubles down on private cloud at VMware Explore

Copyright TechTarget - All rights reserved https://cyber.law.harvard.edu/rss/rss.html Techtarget Feed Generator en Sun, 14 Jun 2026 05:52:29 GMT https://www.techtarget.com/searchdatacenter editor@techtarget.com LAS VEGAS -- Hock Tan didn't mince words when he took the VMware Explore 2025 stage Tuesday. The Broadcom CEO's message was clear: VMware's future is tied to increasing private cloud demand. That message kicked off a slew of VMware Cloud Foundation <del datetime="2025-08-27T13:10" cite="mailto:Biscobing,%20Jacqui">(VCF) </del>launches and upgrades, including an expanded partnership with software firm Canonical, virtualization for the latest AMD GPUs, bundled VMware Private AI Services into VCF 9.0 and a host of Tanzu additions. "Last year<ins datetime="2025-08-27T13:10" cite="mailto:Biscobing,%20Jacqui">,</ins> when I was here<del datetime="2025-08-27T13:10" cite="mailto:Biscobing,%20Jacqui">,</del> we <a href="https://www.techtarget.com/searchdatacenter/news/366609324/Hock-Tan-locks-in-VMwares-private-cloud-future">talked about private cloud</a> being the future of the enterprise. Twelve months later, the future is here," Tan said. "You want to invest in private cloud. But here's the challenge: VMware innovated for years, but they never truly integrated the building blocks of a cloud." For Torsten Volk, an analyst at Enterprise Strategy Group, now part of Omdia, Tan was laying out a clear strategy. <a href="https://www.techtarget.com/searchcloudcomputing/definition/cloud-repatriation">Cloud repatriation</a>, moving workloads from public cloud to hybrid or private cloud, is gaining momentum as companies look to control costs and add security. "The main pitch was very blatantly that, 'We are better than public cloud. We can do everything public cloud can, and we do it more securely, we do it with more control, and we do it more effectively," Volk said. "He was basically saying that VMware is better in everything than public cloud, which is very interesting … because everybody else is talking about hybrid cloud." <section class="section main-article-chapter" data-menu-title="Bolstering VCF with AI, security, partnership"> <h2 class="section-title">Bolstering VCF with AI, security, partnership</h2> At the event, Broadcom unveiled the inclusion of VMware Private AI Services as a standard component of <a href="https://www.techtarget.com/searchvmware/news/366626077/Wait-is-over-for-Broadcoms-VMware-Cloud-Foundation-9">VCF 9.0</a>, making it an AI native platform. That gives VCF customers access to AI tools such as its support assistant Intelligent Assist for VCF, <a href="https://www.techtarget.com/searchitoperations/news/366621932/Model-Context-Protocol-fever-spreads-in-cloud-native-world">Model Context Protocol</a> support to connect with external AI tools and services, and the ability to run AI models on a range of hardware and GPUs from Nvidia and AMD, the company said. "They clearly had to do that," Volk said. "Because agentic AI is the No. 1 topic for everybody." Chris Wolf, Broadcom's global head of AI and advanced services for VCF, told attendees that Private AI Services has evolved since the company first launched it three years ago. "The world has caught on to the notion that you can bring your models to wherever the data resides, and you can run those models at a lower cost without having to sacrifice privacy or control of your data," he said. On the security side, the company touted its VCF Advanced Cyber Compliance with updates in VMware vDefend and VMware Avi Load Balancer for bolstered private cloud compliance and security in regulated industries. Broadcom also announced an expanded partnership with Canonical that will give customers access to the <a href="https://www.techtarget.com/searchdatacenter/definition/Ubuntu">Ubuntu</a> Pro, the leading operating system for Kubernetes. Volk said that will put VCF in striking distance of competitors like Red Hat for Kubernetes workloads. "I think that partnership could be a big deal because the key thing is to get people to believe that they can get from legacy to cloud native easily," Volk said. <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/tan_hock_vmware_2025-f.jpg "> <img data-src="https://www.techtarget.com/rms/onlineimages/tan_hock_vmware_2025-f_mobile.jpg " class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/tan_hock_vmware_2025-f_mobile.jpg 960w,https://www.techtarget.com/rms/onlineimages/tan_hock_vmware_2025-f.jpg 1280w" alt="Hock Tan on stage during VMware Explore 2025 keynote" data-credit="Shane Snider/Informa TechTarget" height="357" width="560"> <figcaption> Broadcom CEO Hock Tan continued his private cloud pitch to VMware customers at VMware Explore 2025. </figcaption> <div class="main-article-image-enlarge"> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="A VCF pitch to SMBs"> <h2 class="section-title">A VCF pitch to SMBs</h2> After <a href="https://www.computerweekly.com/feature/Broadcoms-VMware-acquisition-explained-The-impact-on-your-IT-strategy">Broadcom's $69 billion VMware</a> purchase in 2023, SMBs and organizations faced sticker shock when services were bundled and exclusively made available through subscriptions. <a href="https://www.techtarget.com/searchdatacenter/news/366610359/ATT-sues-Broadcom-over-VMware-support-licensing">Customers</a> suddenly faced with eight to 15 times price increases for renewals, according to multiple reports. During Explore 2025, VMware tried to show that even smaller companies could benefit from a broader portfolio of virtualization services. The company included Jeremy Wright, director of IT infrastructure at Iowa insurance company Grinnell Mutual, in its on-stage lineup, who talked about how his company came to take advantage of VMware's offerings. "I had heard what people were saying about pricing, so I came [to last year's VMware Explore event] looking for answers," Wright said. "I really started to think, 'This looks like it's a good fit for really big companies, but is it powerful enough and a fit for a team like mine?' And that doubt lingered." Wright said that he found that switching from <ins datetime="2025-08-27T13:20" cite="mailto:Biscobing,%20Jacqui">discrete </ins>storage clusters to VMware's vSAN software-defined storage offering would save his company $1 million. "VCF started to feel like it's not going to be so big for us<ins datetime="2025-08-27T13:20" cite="mailto:Biscobing,%20Jacqui">,</ins> and that it's just going to perfectly scale to my small team," he said. Wright said he started looking to add VMware by optimizing other software renewals and cutting spending elsewhere. He presented his plan to his company's C-suite, and it approved his plan. "VCF is transforming our infrastructure from the ground up … it's a unifier for us," he said. "Stop thinking about it as servers and infrastructure and start thinking and talking about it as a private cloud experience." Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> Hock Tan talked only for several minutes during the keynote at VMware Explore 2025, but in that time, he laid the groundwork for a host of VMware updates. https://cdn.ttgtmedia.com/rms/onlineimages/cloud_g1297025236.jpg https://www.techtarget.com/searchvmware/news/366629924/Broadcom-CEO-doubles-down-on-private-cloud-at-VMware-Explore Wed, 27 Aug 2025 09:03:00 GMT Broadcom CEO doubles down on private cloud at VMware Explore A new deal that gives the federal government a stake in Intel could help fend off a potential doomsday scenario for the tech industry, but it doesn't mean the struggling company -- or the semiconductor industry in the U.S. -- is out of the woods yet. Intel's <a href="https://www.techtarget.com/whatis/feature/Intels-rise-and-fall-A-timeline-of-what-went-wrong">downward spiral</a>, particularly over the last year, has been well-documented, attributed to its failure to remain competitive amid the rise of GPUs for generative AI applications, which has been dominated by Nvidia. Intel has also lost market share in CPUs to rivals such as AMD in recent years. Still, the company previously had a commanding market share in both servers and PCs for decades, and Intel CPUs remain a major component of enterprise IT infrastructure worldwide. "Everyone focuses on Nvidia and AI, which is all well and good and great, and making a ton of money, but there are a whole lot of servers out there running databases on Intel CPUs in enterprise IT shops," said Jack Gold, principal analyst at J.Gold Associates. "So if Intel fails, the whole industry is in trouble." Meanwhile, Intel has something of interest to the U.S. government that rivals such as AMD do not: U.S.-based chip factories under its Intel Foundry business. The company has struggled to find large-scale outside customers for its latest 18A and 14A chip manufacturing processes, but the U.S. government has national security and economic reasons to want computer chip manufacturing that's based in the U.S. rather than Taiwan or China. AMD partners with Taiwan Semiconductor Manufacturing Company (TSMC) for chip manufacturing. Propping up U.S.-based semiconductor manufacturing is a notion that predates the current Trump administration, but as of Friday, that administration changed the terms of funding for Intel and other chipmakers set forth by the <a href="https://www.techtarget.com/searchdatacenter/news/366615473/CHIPS-Act-expected-to-survive-change-under-Trump">U.S. CHIPS and Science Act</a> during the Biden administration. Now, Intel will issue additional shares in the company that represent a 9.9% non-voting stake in the business to the U.S. Department of Commerce in exchange for the release of $8.9 billion in funds the company had been granted under the Biden-era legislation. <section class="section main-article-chapter" data-menu-title="Industry reacts to news of Intel government stake"> <h2 class="section-title">Industry reacts to news of Intel government stake</h2> The government stake in Intel prompted <a target="_blank" href="https://www.wsj.com/politics/policy/trumps-intel-move-blurs-party-lines-on-economic-intervention-3c7665cc" rel="noopener">widespread discomfort</a>, given the capitalist economic culture in the U.S., where government ownership over the means of industrial production is met with trepidation. An Intel <a target="_blank" href="https://www.intc.com/filings-reports/all-sec-filings?form_type=&year=2025" rel="noopener">Form 8-K</a> released Monday indicated that the U.S. Department of Commerce will not have representation on the Board of Directors, quelling those fears somewhat, but many outstanding questions about the timing of the transaction and additional conditions required to secure the funding remain unanswered, according to the 8K. The form also outlined the potential risks of the government deal, especially its possible effect on Intel's business outside the U.S. "Sales outside the US accounted for 76% of the Company's revenue for the fiscal year ended December 28, 2024," the form states. "Having the US Government as a significant stockholder of the Company could subject the Company to additional regulations, obligations or restrictions, such as foreign subsidy laws or otherwise, in other countries." In the short term, the funding injection could restore confidence among enterprise IT buyers in the chipmaker amid a 50% drop-off in its stock price last year and <a target="_blank" href="https://www.bbc.com/news/articles/c70x6602pdyo" rel="noopener">previous calls</a> from President Trump for its CEO, Lip-Bu Tan, to step down, said Patrick Moorhead, CEO and chief analyst at Moor Insights and Strategy. "Intel technically wasn't teetering on the brink of financial calamity, but the difference between perceptions and reality can be pretty stark," Moorhead said. "I have heard, 'Oh my gosh, how long can Intel survive? How long can they do this?' They look at the President calling for the firing of the CEO, [which] would have meant even more churn inside of Intel. And then if you're [invested in] Intel as an enterprise, you might wonder, 'What does the future look like?'" <div class="imagecaption alignLeft"> <img src="https://cdn.ttgtmedia.com/rms/onlineimages/dickens_steven.jpg" alt="Steven Dickens, CEO and principal analyst, HyperFrame Research">Steven Dickens </div> Tech industry analysts said it's clear that Intel is "too big to fail," echoing a catchphrase from the global financial crisis in 2007 and 2008 that saw widespread U.S. government investments to bail out various industries. While enterprise IT buyers are two or three degrees removed from chip foundries, which typically supply the server hardware suppliers they buy from, the downstream effect of chip foundries on tech and the global economy is potentially massive, according to Steven Dickens, CEO and principal analyst at HyperFrame Research. "Intel is an edge case," Dickens said. "The means of production should stay with companies, but it's strategically important that the U.S. has a solid foundry provider -- it's so foundationally important to the modern structure of our world." Not just the U.S., but much of the Western world also has its semiconductor supply chain concentrated in Taiwan, which is potentially subject to conflict with China, Dickens said. </section> <section class="section main-article-chapter" data-menu-title="Systemic problems remain for Intel and the industry"> <h2 class="section-title">Systemic problems remain for Intel and the industry</h2> While the Intel government stake acts as a potential short-term boost, it's merely a step toward solving the company's biggest problem, which is finding an outside, high-scale customer for its next-generation <a href="https://www.computerweekly.com/news/366629333/Whats-going-on-inside-Intel">14A chip manufacturing process</a>. Without such a customer, Intel would only be funding a new foundry facility for its own internal needs, the economics of which aren't feasible. <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> Imagine getting all our wheat from one field in Nebraska, for example -- [sourcing the global semiconductor supply chain solely in Taiwan] is the same kind of scenario. But [Intel] can't fill those new factories just by itself -- it's like having a 100-acre farm just to grow your own food. </figure> <figcaption> Jack GoldPrincipal, J. Gold Associates </figcaption> </div> </blockquote> "Imagine getting all our wheat from one field in Nebraska, for example -- [sourcing the global semiconductor supply chain solely in Taiwan] is the same kind of scenario," Gold said. "But [Intel] can't fill those new factories just by itself -- it's like having a 100-acre farm just to grow your own food." In fact, prior to the government stake announcement, Intel's Tan said in a <a target="_blank" href="https://newsroom.intel.com/corporate/lip-bu-tan-steps-in-the-right-direction" rel="noopener">public statement</a> that Intel will only further invest in the 14A manufacturing process "based on confirmed customer commitments." The $8.9 billion afforded by the government stake also won't account for the <a target="_blank" href="https://siliconheartland.newalbanyohio.org/" rel="noopener">estimated $28 billion</a> investment it would take to finish building out Intel's newest factory near Columbus, Ohio. Without that new facility, Intel won't be able to regain its status as an innovator in semiconductor manufacturing, according to Gold. "Intel was two to three years ahead of the industry over the last decade or so, then they just totally blew that. They screwed up, and that's being polite. So how do they get that back?" he said. "They're investing in 18A to begin with, and 14A afterwards, and they have got to be successful in that space." On the chip design side of the Intel business, it could potentially capture a new opportunity to revitalize its CPUs as generative AI workloads become more efficient and move away from massive clusters of GPUs to more conventional x86-based infrastructure, Gold said. <div class="imagecaption alignLeft"> <img src="https://cdn.ttgtmedia.com/rms/onlineImages/crawford_tim.jpg" alt="Tim Crawford, CIO strategic advisor, AVOA">Tim Crawford </div> If not, this gap could potentially be bridged by competitors such as AMD and Qualcomm, but shifting to new chip suppliers could be disruptive to enterprises, said Tim Crawford, CIO strategic adviser at Avoa, a research and advisory services firm in Rolling Hills Estates, Calif. "The [government] stake in Intel provides Intel with more runway to maintain their business, but it does not provide a long-term solution to Intel's woes," Crawford said. Intel must decide whether to separate its foundry and design businesses, according to Moorhead. Either way, the foundry business will need beefed-up oversight from qualified advisors to right the ship, he said. Even if all goes perfectly for Intel as a result of the government stake, it won't necessarily propel the U.S. semiconductor industry forward, Crawford said. "Taking an equity stake in Intel does not encourage further research and development, nor production of semiconductor technology within the U.S.," he said. "The bigger opportunity would be to invest in research organizations that are leading the innovation in the space and encourage long-term investment [there]. The CHIPS Act was intended to do some of this. <a target="_blank" href="https://www.reuters.com/world/us/trump-weighs-using-2-billion-chips-act-funding-critical-minerals-sources-say-2025-08-21/" rel="noopener">Removing funding for research institutions</a> takes away from this." Beth Pariseau, a senior news writer for Informa TechTarget, is an award-winning veteran of IT journalism covering DevOps. Have a tip? <a href="mailto:beth.pariseau@informatechtarget.com?subject=News%20tip">Email her</a> or reach out <a target="_blank" href="https://x.com/PariseauTT" rel="noopener">@PariseauTT</a>. </section> A deal giving the U.S. federal government a 10% stake in Intel in exchange for funds won't necessarily ensure the company's ultimate survival, according to industry analysts. https://cdn.ttgtmedia.com/rms/onlineimages/legal_g941378956.jpg https://www.techtarget.com/searchdatacenter/news/366630032/Experts-Intel-government-stake-wont-fix-long-term-woes Mon, 25 Aug 2025 18:01:00 GMT Experts: Intel government stake won't fix long-term woes AWS this week launched its Amazon Elastic Compute Cloud (EC2) R8i and R8i-flex instances -- new virtual servers powered by custom Intel Xeon 6 processors. The eighth-generation instances promise 15% price performance gains and 2.5 times more memory bandwidth compared with previous Intel-outfitted instances. The customized <a target="_blank" href="https://www.intel.com/content/www/us/en/content-details/845771/intel-xeon-6-processor-family-product-brief.html" rel="noopener">Xeon chips</a> are available on only AWS and feature DDR5 7200MTps memory operating at up to 3.9 GHz, according to AWS. Intel said that firepower will offer flexibility that minimizes the need for specialized accelerators. The chips also provide built-in AI acceleration with integrated Advanced Matrix Extensions (AMX), delivering up to two times AI inference and machine learning performance gains. Intel introduced its <a href="https://www.techtarget.com/searchenterpriseai/news/366587618/Intel-launches-Xeon-6-for-AI-data-centers">Xeon 6 processors</a> for the data center in April. While Intel declined to provide specifics on Xeon 6 customization, a spokesperson said the company worked closely with AWS to optimize performance and software implementation. "Beyond the hardware itself, we worked across the system stack covering firmware, hypervisors, virtualization layers and software frameworks to ensure that these capabilities are fully optimized and accessible to cloud workloads," the spokesperson said via email. Matt Kimball, vice president and principal analyst at Moor Insights & Strategy, said the offerings will be attractive to enterprise customers. "If I were an IT professional, the decision to stay on R7i or migrate to R8i would be a no-brainer," he said in an email. "The Granite Rapids architecture has been greatly improved over previous generations to deliver great performance for latency-sensitive workloads." <section class="section main-article-chapter" data-menu-title="Competitive advantages for Intel, AWS"> <h2 class="section-title">Competitive advantages for Intel, AWS</h2> The continued relationship <a href="https://www.techtarget.com/searchdatacenter/news/366611032/Intel-gets-boost-from-AWS-government-contracts">between AWS and Intel</a> benefits both companies, analysts said. For Intel, the nod from AWS for its new generations of instances "is a powerful endorsement, especially as [Intel CEO] Lip-Bu Tan executes his strategic vision for Intel," said Ron Westfall, an analyst at HyperFrame Research, in an email interview. "It demonstrates that Intel's technology can be the choice for demanding, mission-critical workloads such as databases and analytics, reinforcing its market position." With the collaboration, Amazon can target growing customer demand for increased memory capacity and bandwidth, Westfall said. "Mission-critical enterprise applications, including large-scale business intelligence and data warehousing, require robust and high-performing infrastructure. The R8i instances, with their large instance sizes and bare-metal options, provide the necessary power and reliability," he said. For Intel, which has seen GPU king Nvidia grab significant market share while its own data center presence shrinks, the nod from AWS is a <a href="https://www.techtarget.com/searchdatacenter/news/366613916/Intel-AMD-form-x86-group-in-effort-to-combat-Arm">win for x86 architecture</a> -- where the company hangs on to a lead in the data center despite growing CPU competition from Advanced Micro Devices (AMD). "The deal further demonstrates the integral role x86 CPU server processors play in AI training and inference computing devices for 2025 workloads," Westfall said. Kimball said AWS is diversifying its ability to meet customer needs. "From an AWS perspective, the company is very focused on delivering customer choice," he said. "Be it AMD with Turin, Graviton or Intel with Xeon 6 -- the company continues to ensure that every customer's use case is being met." Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> The companies tout a collaboration that will offer better price performance and memory bandwidth for memory-intensive workloads. https://cdn.ttgtmedia.com/rms/onlineimages/maze_g676210320.jpg https://www.techtarget.com/searchdatacenter/news/366629885/AWS-launches-virtual-servers-with-custom-Intel-Xeon-6-chips Fri, 22 Aug 2025 12:15:00 GMT AWS launches virtual servers with custom Intel Xeon 6 chips Nvidia this week launched its latest Blackwell server GPU aimed at smaller enterprises using more traditional IT infrastructure, a development the company said will enable on-premises AI compute power at a lower cost. Partnering with Cisco, Dell Technologies, HPE, Lenovo and Supermicro, enterprises will also have access to 2U Nvidia RTX Pro servers available in multiple configurations for workloads such as agentic AI, content creation, data analytics, graphics, scientific simulation and both industrial and physical AI. For smaller, on-premises data centers, the <a href="https://www.techtarget.com/searchenterpriseai/news/366616254/Nvidia-Blackwell-GPU-production-ramps-up-amid-revenue-surge">CPU is still the workhorse</a> handling IT infrastructure needs. But with demand for AI reaching a fever pitch, RTX Pro 6000 Blackwell Server Edition GPUs could be an attractive upgrade for smaller outfits, analysts said. The RTX Pro 6000 GPUs could give smaller enterprise operations access to powerful accelerators that match their power and cooling capabilities. The lower-density GPUs combined with the more compact RTX Pro Server or other server configurations allow for air-cooling and fit into existing racks, providing plug-and-play upgrade possibilities. "I see the RTX PRO 6000 and RTX Pro Server as Nvidia bringing Blackwell to the masses," said Matt Kimball, vice president and principal analyst at Moor Insights & Strategy, in an email interview. He said RTX 6000 GPUs offer similar performance to the company's top-of-line <a href="https://www.techtarget.com/searchdatacenter/tip/Top-AI-hardware-companies">Blackwell Ultra</a>, or B300, at a lower price tag. "The RTX Pro Server brings all of these capabilities, right-sized for the non-large enterprise customer -- think about the 5,000-employee, 1,000-server organization that wants to deploy agentic AI across the enterprise," Kimball said. <section class="section main-article-chapter" data-menu-title="Nvidia looks to increase GPU footprint"> <h2 class="section-title">Nvidia looks to increase GPU footprint</h2> Justin Boitano, Nvidia's vice president of enterprise AI, said Nvidia wanted to broaden its GPU and server portfolio based on demand from enterprise customers. "A lot of enterprises have data centers that might have lower power densities or have air-cooled data centers," he said in an email interview. "These RTX Pro Servers end up fitting brilliantly in that kind of traditional IT footprint. As we roll this out, we're seeing a lot of interest across every vertical industry." Boitano said the GPU boost will enable enterprises to power virtual desktops and design applications, <a href="https://www.techtarget.com/searchcio/definition/synthetic-data">synthetic data</a> generation, digital twins and more. "This infrastructure offers an accelerated computing platform that has the broadest capability," he said. "As we really bring AI into this world of building agents and put them to work within our companies, we're bringing the compute to the data." CPU-only servers are becoming stretched to their limits as enterprises explore use cases with more power needs. At the same time, more enterprises are moving back to on-premises or hybrid architectures to save on costs and increase security. CIOs were faced with the option of reinvesting in CPU-only architectures that are limited in AI workload capability or moving workloads to the cloud, according to Boitano. "So it's really not a huge risk to their business at this point, even to pause and reinvest that capital into AI infrastructure to get their company on this path of building and deploying AI," he said. Kimball said aiming at smaller enterprise customers was a good move for Nvidia's overall market strategy. Main GPU rival AMD does not have a product aimed at smaller enterprises, he said. That left a big opening for Nvidia. "I think Nvidia is smart in what it is doing with these platforms," he said. "It is effectively seeding the commercial enterprise segment with an accelerated compute platform that can deliver value today -- and tomorrow, when that organization is ready to adopt agentic AI enterprise-wide." Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> The company's RTX Pro 6000 Blackwell Server Edition GPU and RTX Pro Server offer companies using smaller-scale enterprise infrastructure a way to do more AI on-premises. https://cdn.ttgtmedia.com/rms/onlineimages/storage_g1180684429.jpg https://www.techtarget.com/searchdatacenter/news/366629130/Nvidia-introduces-entry-level-RTX-Pro-GPU Fri, 15 Aug 2025 14:12:00 GMT Nvidia introduces entry-level RTX Pro GPU Nvidia and AMD have agreed to U.S. President Donald Trump's planned 15% fee on chip sales to China, according to several published reports. The companies are willing to pay fees for export licenses for <a href="https://www.techtarget.com/searchenterpriseai/news/366623018/Tariffs-not-likely-to-undermine-Nvidias-AI-dominance">Nvidia's H20</a> and AMD's MI308 -- GPUs that could boost China's AI capabilities, according to a report by <a target="_blank" href="https://www.ft.com/content/cd1a0729-a8ab-41e1-a4d2-8907f4c01cac" rel="noopener">Financial Times</a>. The administration had previously restricted sales of the chips to China, citing national security concerns. The companies stand to make billions of dollars as China's AI appetite fuels powerful GPU demand. "We follow rules the U.S. government sets for our participation in worldwide markets," an Nvidia spokesperson said in a statement. "While we haven't shipped H20 to China for months, we hope export control rules will let America compete in China and worldwide. … America's AI tech stack can be the world's standard if we race." <section class="section main-article-chapter" data-menu-title="Bargaining chips"> <h2 class="section-title">Bargaining chips</h2> Last month, after meetings with the administration, Nvidia and AMD both said they would resume chip sales, while not mentioning potential fees being part of the agreement. In May, Nvidia said it had incurred $4.5 billion in charges due to excess inventory for its H20 processors in one quarter alone. The company said that, without restrictions, it would have gained $2.5 billion in sales. Nvidia's 2025 revenue from China was $17.11 billion before April's H20 ban, accounting for 13.11% of its total global sales. And, at AMD, "AI business revenue declined year over year as U.S. export restrictions effectively eliminated MI308 sales to China," CEO Lisa Su told investors on an earnings call last week. Alvin Nguyen, senior analyst at research firm Forrester, said the absence of American chipmakers has strengthened Chinese firms, like Huawei, as the country races to <a href="https://www.techtarget.com/searchcio/news/365532997/Tech-competition-with-China-remains-top-of-mind-for-US">compete in the AI arms race</a> and fill the GPU void. "Progress by Chinese semiconductor companies in their AI chip efforts have been impressive and would likely decrease demand for non-Chinese GPUs long term, creating more competition and weakening the position of current Western chip manufacturers over time," Nguyen said in an email. <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> China is not going to sit back and do nothing. </figure> <figcaption> Steven DickensCEO and analyst, HyperFrame </figcaption> </div> </blockquote> Steven Dickens, CEO and analyst at HyperFrame Research, said the semiconductor firms benefit, despite the increase in fees. "These are fundamental weapons in the race for AI supremacy," Dickens said in an interview. "You can stop Nvidia and AMD from selling completely -- and that hurts those companies by turning off a revenue stream. Does it stop the problem? Strip the politics out of it. … It's still a good deal for AMD and Nvidia. China is a key market that has rampant demand for their companies." Huawei has made significant strides in its own GPU offerings, claiming its Ascend 910B is on par with Nvidia's A100 GPUs. Nvidia still has a price and platform advantage, with developers already familiar with its well-established CUDA software ecosystem. "China is not going to sit back and do nothing," Dickens said. "Nvidia is better and cheaper than Huawei. And China is a key market with rampant demand. Who is the loser here?" </section> <section class="section main-article-chapter" data-menu-title="Potential legal struggle"> <h2 class="section-title">Potential legal struggle</h2> Trump's latest move will likely face a legal battle as the U.S. Constitution forbids export taxes, experts said. The Trump administration has used emergency powers to push its trade agenda -- with <a href="https://www.techtarget.com/searchcio/news/366618727/US-tariffs-may-stymy-executives-product-decisions">sweeping tariff hikes</a> bypassing the normal congressional approval process. Article I, Section 9, Clause 5 of the U.S. Constitution states, "No Tax or Duty shall be laid on Articles exported from any State." But the U.S. does allow fees for permitting and costs associated with exports. Export controls are normally enacted for national security reasons without regard to producing government revenue. William Reinsch, senior adviser at the Center for Strategic and International Studies and a former Clinton administration undersecretary of commerce, worried about the national security implications. "From a policy perspective, it essentially says our national security is up for sale, which is not a good message to convey," he said in an email. AMD and the White House did not immediately respond to a request for comment. Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> The administration's unprecedented move may conflict with the U.S. Constitution's rules against export taxes. https://cdn.ttgtmedia.com/rms/onlineimages/legal_g90787303.jpg https://www.techtarget.com/searchdatacenter/news/366628975/Trump-fee-for-Nvidia-AMD-China-exports-could-face-legal-battle Mon, 11 Aug 2025 14:42:00 GMT Trump fee for Nvidia, AMD China exports could face legal battle Democratic senators are pushing for a new review of HPE's <a href="https://www.techtarget.com/searchnetworking/news/366626897/DOJ-clears-road-for-HPEs-14B-Juniper-Networks-acquisition">approved acquisition</a> of Juniper Networks after the U.S. Department of Justice fired two top antitrust deputies for "insubordination" over the $14 billion deal. DOJ antitrust deputies Roger Alford and Bill Rinner were fired earlier this week, a Justice Department spokesperson confirmed to Informa TechTarget. Alford worked in the DOJ under the first Trump administration, and Rinner was a top legal advisor to Makan Delrahim, former head of the antitrust division under Trump. An Axios report citing an anonymous national security official on Wednesday claimed the U.S. intelligence community stepped in just before the merger was set to go to trial this month, persuading the Justice Department to reach a settlement. U.S. officials worried that blocking the merger would give China's Huawei Technologies an advantage over American companies, according to the report. That intervention, along with reported internal lobbying efforts, created infighting at the DOJ that led to Alford and Rinner being fired, the report said. The alleged infighting would illustrate a divide between antitrust hawks and pro-business forces within the Trump administration. <section class="section main-article-chapter" data-menu-title="Firings lead to call for scrutiny"> <h2 class="section-title">Firings lead to call for scrutiny</h2> In a statement, DOJ spokesman Gates McGavick denied that backroom dealmaking influenced the settlement. "The department has consistently reiterated that the resolution of this merger was based only on the merits of the transaction," he said. In response to the controversy, U.S. Sens. Elizabeth Warren (D-Mass.), Amy Klobuchar (D-Minn.), Cory Booker (D-N.J.) and Richard Blumenthal (D-Conn.) wrote to U.S. District Court Judge P. Casey Pitts -- an appointee of former U.S. President Joe Biden -- urging judicial review of <a href="https://www.techtarget.com/searchnetworking/news/366565834/HPE-to-acquire-Juniper-Networks-for-14-billion">the merger</a>. The senators called for heightened review under the Tunney Act (formally known as the Antitrust Procedures and Penalty Act), a 1974 law that allows courts to hold hearings and take action that could affect the merger. The act was first used in 2019 to review the CVS-Aetna merger. While that merger stood, the proceedings led to new precedents for judicial antitrust scrutiny. "A settlement to resolve HPE's proposed acquisition of Juniper should not be made on the backs of the American people while enriching well-connected lobbyists," the senators wrote in a <a target="_blank" href="https://www.warren.senate.gov/imo/media/doc/letter_from_senator_warren_to_judge_pitts_on_hpe-juniper_merger_and_tunney_act.pdf" rel="noopener">letter</a> to Pitts. Trump allies Mike Davis and Arthur Schwartz were lobbying on behalf of HPE, according to filings and several published reports. Schwartz is known to be a friend and confidante of U.S. Vice President JD Vance. In a statement, HPE said its merger with Juniper would promote competition in the WLAN market. "The transaction was appropriately approved with certain remedies by the U.S. Department of Justice, and it was unconditionally approved by 13 other antitrust regulators around the world," an HPE spokesperson said. </section> <section class="section main-article-chapter" data-menu-title="HPE's big win in jeopardy?"> <h2 class="section-title">HPE's big win in jeopardy?</h2> Analysts widely saw the June settlement as a big win for HPE, despite caveats of the settlement, which included the company selling off its Instant On business and licensing of Juniper's AIOps for Mist source code. Instant On, which offers a line of wireless access points and switches, is aimed at small and medium-sized businesses. "It appears that the divestiture does not allay concerns arising from the combination of HPE and Juniper Networks' enterprise-grade products, which would merge two current competitors and create an apparent duopoly," the senators wrote in the letter to the court. Renewed scrutiny could slow or even jeopardize the merger. Ron Westfall, an analyst at HyperFrame Research, said the potential new review could prove to be a major setback. "It's not good, overall, as HPE and Juniper teams have already started embarking on integration plans. This has already been scrutinized heavily," he said. "It's not good for the overall wireless LAN industry, let alone the tech industry. I think there's some irony here that there's political pressure here to decide if there were politics involved with the decision." In the letter to the judge, the senators said the HPE settlement did not address concerns about competition and market consolidation. After the deal, HPE and its closest competitor, Cisco, would hold a 70% WLAN market share. Notably, a Tunney Act review has never overturned an antitrust settlement. Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> Reported involvement from intelligence agencies, lobbyists may have helped push the merger through before a planned antitrust trial, causing infighting at the Justice Department. https://cdn.ttgtmedia.com/rms/onlineimages/legal_g929185540.jpg https://www.techtarget.com/searchnetworking/news/366628372/DOJ-firings-spark-fresh-HPE-Juniper-deal-scrutiny Thu, 31 Jul 2025 14:41:00 GMT DOJ firings spark fresh HPE-Juniper deal scrutiny AWS has withdrawn plans for a 7.2 million-square-foot data center -- which would have been one of the world's largest -- in Louisa County, Va., after neighbors voiced concerns about its potential impact. Amazon scuttled the plan last week after citizens protested the 1,370-acre project -- an area the size of more than 1,000 football fields -- citing concerns about water availability and overall effect on the rural landscape. "We have heard the community and appreciate the desire for more robust input in any future projects that may be brought forward in the county," said Charles W. Payne, Jr., an attorney for AWS, in a statement. The Louisa County community's win to block the project highlights a potential pushback on U.S. tech companies' ambitious efforts in a global AI race. And the fight comes on the heels of U.S. President Donald Trump's <a href="https://www.techtarget.com/searchdatacenter/news/366628044/Trumps-fossil-fueled-AI-Action-Plan-pushes-deregulation">AI Action Plan</a>, which pushes deregulation efforts to support fast-tracked data center builds throughout the country. Alan Howard, principal analyst in Omdia's cloud and data center practice, said companies need to do more to communicate with stakeholders early on. Omdia is a division of Informa TechTarget. "Local activism can paint a bleak picture on how data centers or any industrial facility will be a blight in the neighborhood," he said. "Regulations are, in general, a good thing to protect residential neighborhoods … but an overzealous NIMBY ['not in my backyard'] influence, combined with legislators or regulators that don't understand data centers, simply chases away economic opportunity to other markets." AWS said its proposed data center would have employed up to 864 people and added up to $144 million in gross tax revenue. <a target="_blank" href="https://louweb.louisa.org/LCDocs/_CUP/2025-05/Application.pdf" rel="noopener"></a>For residents opposed to the project, those benefits would not outweigh the potential risks. <section class="section main-article-chapter" data-menu-title="The Trump card"> <h2 class="section-title">The Trump card</h2> The Trump administration has promised lighter federal permitting regulations for projects costing more than $500 million, but the Louisa County fight shows localized efforts could still provide substantial hurdles. Additionally, Trump's AI Action Plan calls for data center construction on federal lands. AWS already has two data centers under construction in Louisa County, a rural community south of Washington, D.C. Those projects, representing $11 billion of a $35 billion AWS data center statewide effort, did not require the same permitting scrutiny and community input. In 2018, the county developed a <a target="_blank" href="https://www.louisacounty.gov/3244/Technology-Overlay-District-TOD" rel="noopener">Technology Overlay District</a>, opening the area to data center construction "by right" in designated zones without the need for additional use permits. But the county's board of supervisors now requires new projects to file a conditional use permit and gather community input. </section> <section class="section main-article-chapter" data-menu-title="Lessons for both sides"> <h2 class="section-title">Lessons for both sides</h2> Louisa is a two-hour drive from Loudoun County, home to the so-called "Data Center Alley" -- the largest concentration of active data centers in the world. At the meeting held at a local Baptist church last week, community members passed out stickers reading "Don't Loudoun my Louisa." Amazon's proposed Louisa data center would have been one of the largest in the world and would have tied for the largest in North America. Switch's Nevada data center -- the <a target="_blank" href="https://brightlio.com/largest-data-centers-in-the-world/" rel="noopener">Citadel Campus</a> -- also has 7.2 million square feet planned, with an estimated cost of $1.3 billion. China's Inner Mongolia Information Park sits atop the list with a 10.7 million-square-foot data center. While AWS does not disclose data center footprint specifics, an ABI Research <a target="_blank" href="https://www.abiresearch.com/blog/data-centers-by-region-size-company?utm_source=chatgpt.com" rel="noopener">report</a> says the cloud hyperscaler operates the most worldwide, with 126 sites as of 2024, accounting for 25% of the market. Will Townsend, vice president and principal analyst at Moor Insights & Strategy, said AWS may try again after a community outreach effort. "The general population is unaware of what is actually involved in cooling data centers with the reuse of water and recirculating systems," he said. "My guess is that AWS recognizes the need to educate communities before going to the permitting process -- and they pulled this one as a lesson learned." "Sustainability is a valid concern given the power consumption of AI data centers, but silicon and infrastructure are rapidly advancing to address these issues," he added. Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> The proposed 7.2 million-square-foot operation -- one of the world's largest -- would have added to Amazon's $35 billion data center plan in Virginia. https://cdn.ttgtmedia.com/rms/onlineimages/storage_g1197646065.jpg https://www.techtarget.com/searchdatacenter/news/366628222/AWS-tables-Va-data-center-after-community-pushback Wed, 30 Jul 2025 10:51:00 GMT AWS tables Virginia data center after community pushback Shrinking semiconductor giant Intel Corp. will nix planned European factories and slow U.S. construction, the company said Wednesday after its latest earnings report showed flat revenue year over year for the quarter, and a loss on earnings per share. Intel's <a target="_blank" href="https://www.intc.com/news-events/press-releases/detail/1745/intel-reports-second-quarter-2025-financial-results" rel="noopener">second-quarter 2025 results</a> beat revenue expectations with $12.86 billion, but the company reported a net income loss of $2.9 billion. Intel's <a href="https://www.techtarget.com/searchenterpriseai/news/366571038/Intel-Foundry-launches-as-enterprise-AI-surges">foundry business</a>, which launched last year, suffered an operating loss of $3.17 billion on $4.4 billion in revenue. CEO Lip-Bu Tan said the company will cancel manufacturing operations in Germany and Poland while consolidating testing and assembly operations in Costa Rica into existing sites in Vietnam and Malaysia. The company will also slow construction of a $28 billion chip factory in Ohio as it searches for customers. "There are no more blank checks," Tan said about Intel's foundry business as part of a <a title="https://newsroom.intel.com/corporate/lip-bu-tan-steps-in-the-right-direction" target="_blank" href="https://newsroom.intel.com/corporate/lip-bu-tan-steps-in-the-right-direction" rel="noopener">memo</a> to employees Thursday. "Over the past several years, the company invested too much, too soon -- without adequate demand. In the process, our factory footprint became needlessly fragmented and underutilized. We must correct our course." <section class="section main-article-chapter" data-menu-title="Mass layoffs at Intel near end"> <h2 class="section-title">Mass layoffs at Intel near end</h2> Tan said Intel's deep layoff plan was nearly complete, with 21,400 total job cuts expected by the end of 2025. The company will slim its workforce to 75,000 from the most current headcount of 96,400. The chipmaker said it has reduced the number of "management layers" by 50%. "We are making hard but necessary decisions to streamline the organization," Tan said in his memo to employees. "These actions are critical to strengthening our competitive position going forward, but it means we are saying goodbye to valued colleagues." Jack Gold, president and analyst at J. Gold Associates, in an email said the completion of layoffs "should help with profitability both with fewer impairment charges, as well as simply reduced costs of payroll. It's sad for the workers that are let go, but Intel needed to prune the payroll." </section> <section class="section main-article-chapter" data-menu-title="Focus on data center CPUs"> <h2 class="section-title">Focus on data center CPUs</h2> Part of the course correction, the company said, will be to wrestle back market share from data center rival Advanced Micro Devices by focusing on its <a href="https://www.techtarget.com/searchdatacenter/news/366613916/Intel-AMD-form-x86-group-in-effort-to-combat-Arm">x86 chip business</a>. While Intel is still the leader in data center CPUs, AMD has captured significant market share in just a few years with its <a href="https://www.techtarget.com/searchenterprisedesktop/ehandbook/AMDs-Ryzen-Epyc-power-surge-in-desktops-laptops-data-centers">Epyc processor line</a>. In the third quarter of 2024, AMD for the first time outsold Intel in the data center with $3.5 billion in revenue, compared with Intel's haul of $3.3 billion over the same period. "Companies like Intel don't turn on a dime, so it's going to take some time for the direction to emerge and play out," said Stephen Sopko, an analyst at HyperFrame Research. "So much of what Lip-Bu is doing is sorting through a storied company looking for value … AMD in the data center has thrived with great products, but also by capitalizing on Intel missing the rapid expansion of AI." Tan plans to address the competitive shortfall by reintroducing simultaneous multithreading (SMT) -- a technique that improves CPU performance by allowing multiple threads to execute simultaneously on a single processor. Intel moved away from SMT for its Lunar Lake design, opting for increased power efficiency to target mobile and AI PC use cases. That decision was made during former CEO Pat Gelsinger's tenure. <a href="https://www.techtarget.com/searchcio/news/366616633/Intel-CEO-Pat-Gelsinger-out-board-searches-for-new-CEO">Gelsinger retired in December</a>. "Moving away from SMT puts us at a competitive disadvantage," Tan wrote. "Bringing it back will help us close performance gaps." Tan said future chip designs will require his final approval. "Can Intel take back market share? Yes, of course," said Matt Kimball, an analyst at Moor Insights & Strategy. "The biggest impact to market share numbers would be in that cloud/hyperscale segment that has very little loyalty or stickiness to a certain vendor and is far more focused on performance per watt and performance per dollar." </section> <section class="section main-article-chapter" data-menu-title="Renewed AI focus"> <h2 class="section-title">Renewed AI focus</h2> Intel has struggled with its AI workload processing business, with <a href="https://www.techtarget.com/searchenterpriseai/news/366571732/Intel-Nvidia-aim-latest-systems-on-a-chip-at-AI-workstations">GPU maven Nvidia</a> and AMD offerings outselling in large-scale AI model training. "The company is exposed on the AI acceleration front," Kimball said. "Nvidia is at a $4 trillion market cap by owning the market." Tan said the company will address its AI shortfalls with an emphasis on <a href="https://www.techtarget.com/searchitoperations/news/366616013/Intel-Nvidia-vie-for-dominance-with-agentic-AI-blueprints">inference and agentic AI</a>. "Our starting point will be emerging AI workloads -- then we will work backward to design software, systems and silicon that enable the best customer outcomes," he said in his memo. That seems like a good plan, Kimball said, noting that there's more space to compete in areas like inferencing. "That's where the long-term play is," he said. Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> CEO Lip-Bu Tan says the company's layoff plan is mostly complete as Intel tries to rebound after financial woes in recent years. https://cdn.ttgtmedia.com/rms/onlineimages/money_g1250581414.jpg https://www.techtarget.com/searchdatacenter/news/366628049/Intel-cuts-spending-eyes-data-center-recovery-after-mixed-Q2 Fri, 25 Jul 2025 12:59:00 GMT Intel cuts spending, eyes data center recovery after mixed Q2 The White House on Wednesday unleashed its AI Action Plan and accompanying executive orders, which include sweeping deregulation efforts the administration said will enable the U.S. to "dominate" in the artificial intelligence arms race. One of President Donald Trump's executive orders revokes former President Joe Biden's January EO establishing AI infrastructure guidance. <a href="https://www.techtarget.com/searchenterpriseai/news/366628092/White-House-AI-plan-places-scrutiny-on-state-AI-laws">Trump's plan</a> would let tech companies <a href="https://www.techtarget.com/searchnetworking/tip/How-to-handle-environmental-regulations-and-green-networking">skirt Clean Water Act and Clean Air Act regulations</a> to fast-track data center and chip manufacturing construction. The plan would also allow large-scale data centers and semiconductor manufacturing on federal lands. Current environmental laws restrict the use of federal lands for commercial construction. It also instructs the Environmental Protection Agency to assist in "expediting permitting" on federal and nonfederal lands by developing or modifying current regulations for data center projects that cost at least $500 million. "America's environmental permitting system and other regulations make it almost impossible to build this infrastructure in the United States with the speed that is required," according to the action plan, which expands "Categorical Exclusions" to National Environmental Policy Act regulations. NEPA requires data center developers to conduct and submit for review environmental impact assessments. Kristian Stout, director of innovation policy at the International Center for Law and Economics, said the action plan is aimed squarely at competing with <a href="https://www.techtarget.com/searchenterpriseai/news/366627679/China-startup-Moonshot-AI-rivals-US-with-cheap-open-model">China's AI efforts</a>. China relies heavily on coal to power its electricity demands, while the U.S. leans on natural gas. The U.S. grid system is fragmented, while China has a national system. "The goal is to be able to do what China seems to be able to do, which is stand up energy quickly and stand up infrastructure for data centers quickly, which helps their companies get access to large data sets to spin up new startups quickly," Stout said. Politics aside, a comprehensive national AI strategy was needed, according to Steven Dickens, CEO and principal analyst at HyperFrame Research. "This is the administration realizing that there's an existential threat and mobilizing the U.S. tech sector," he said. "We are completely architecting the tech stack -- that's everything from power generation through data centers, and right up to the software layer." <section class="section main-article-chapter" data-menu-title="Death knell for sustainability?"> <h2 class="section-title">Death knell for sustainability?</h2> Trump's AI agenda puts AI advancement ahead of corporate sustainability efforts. "[W]e will continue to reject radical climate dogma and bureaucratic red tape, as the Administration has done since Inauguration Day," according to Trump's action plan. Many big tech companies have developed aggressive <a href="https://www.techtarget.com/whatis/definition/environmental-social-and-governance-ESG">environmental, social and governance</a> (ESG) policies to reduce or eliminate their carbon footprints and address global climate change concerns. The administration's push to increase nonrenewable energy sources like coal and natural gas could make those plans harder to achieve, critics warned. "This U.S. AI Action Plan doesn't just open the door for Big Tech and Big Oil to team up, it unhinges and removes any and all doors -- it opens the floodgates," said KD Chavez, executive director of the Climate Justice Alliance, in a statement. "We need more corporate and environmental oversight, not less." Tech companies including Google, Microsoft and Apple have pledged net-zero carbon emissions by 2030. But the surge in <a href="https://www.techtarget.com/searchdatacenter/news/366627578/Meta-Google-unveil-massive-AI-data-center-investment-plans">power-hungry AI data centers</a> has companies reshaping ESG goals. A recent <a target="_blank" href="https://www.conference-board.org/press/sustainability-under-scrutiny-2025" rel="noopener">report</a> by The Conference Board, a nonprofit think tank, found that 80% of 125 large U.S. and multinational companies surveyed have changed ESG goals since Trump's inauguration. Given the race to build out data center projects and the rules to do so, clawing back regulatory oversight could prove tricky even with the next change in administrations, Stout said. "That's going to be difficult to wind that back if you actually get these data centers built and the energy permits granted ... then you're going to be in a court process -- you can't just revoke them on the spot," he said. Stout added that if the plan finds success in making AI gains against China, "it's going to be hard for a future administration to attack that." Dickens believes that business efficiency will still drive renewable energy innovation, pointing to the need for more efficient cooling options to drive down power consumption and costs. But in the short term, he said, natural gas will be the most attractive solution. "Renewables don't give us the constant power that GPUs need, but they will still be a key part of the grid," Dickens said. </section> <section class="section main-article-chapter" data-menu-title="AI infrastructure safety, security addressed"> <h2 class="section-title">AI infrastructure safety, security addressed</h2> The action plan also aims to <a href="https://www.techtarget.com/searchsecurity/tip/How-to-secure-AI-infrastructure-Best-practices">safeguard AI infrastructure</a> by ensuring that "the domestic AI computing stack is built on American products" and that energy and telecommunications sectors are free from foreign software and hardware. The Trump administration also worked with several nonpartisan AI safety advocates for input earlier this year. "These measures, combined with the plan's emphasis on domestic chip production and export control enforcement, create a robust framework for maintaining America's AI advantage," said Varun Krovi, executive director at the Center for AI Safety Action Fund, in a statement. He said the action plan "offers a coherent and comprehensive roadmap for ensuring U.S. leadership in the development of safe, transformational AI." The security effort extends to data centers for military and intelligence use. The plan calls for the creation of new AI data center security standards to be led by government agencies and advancing the adoption of "classified compute environments" that support AI workloads. </section> <section class="section main-article-chapter" data-menu-title="Changes to CHIPS Act"> <h2 class="section-title">Changes to CHIPS Act</h2> The AI Action Plan also looks to bolster the effort to encourage a domestic semiconductor supply chain, promising to remove "extraneous" policy requirements for the <a href="https://www.techtarget.com/searchdatacenter/news/252523061/CHIPS-Act-takes-step-forward-on-long-road-to-production">Biden administration's CHIPS Act</a>. "The Trump administration will lead that revitalization without making bad deals for the American taxpayer or saddling companies with sweeping ideological agendas," according to the plan. The CHIPS Act, which Biden signed into law in 2022, promised nearly $53 billion in grants and incentives for domestic semiconductor manufacturing. The act included rigorous labor, security and environmental standards. Semiconductor projects with a minimum $500 million investment would also qualify for expedited federal permitting and federal land use, according to Trump's executive order. Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> The Trump administration's AI Action Plan and executive orders include deregulation that would fast-track data center and chip buildouts while boosting the power grid. https://cdn.ttgtmedia.com/rms/onlineimages/legal_g90787303.jpg https://www.techtarget.com/searchdatacenter/news/366628044/Trumps-fossil-fueled-AI-Action-Plan-pushes-deregulation Thu, 24 Jul 2025 17:10:00 GMT Trump's fossil-fueled AI Action Plan pushes deregulation Seoul-based chip startup FuriosaAI is looking to threaten Nvidia's data center AI chip dominance and has scored its first major contract -- powering LG's Exaone large language model. LG AI Research will offer FuriosaAI's RNGD chip-powered servers to enterprise customers across electronics, finance, telecommunications and biotechnology for <a href="https://www.techtarget.com/whatis/definition/large-language-model-LLM">LLMs</a>. FuriosaAI's <a target="_blank" href="https://developer.furiosa.ai/latest/en/overview/rngd.html" rel="noopener">RNGD chip</a>, a neural processing unit, is designed to support LLMs and other deep learning models on the inference side, rather than training. <a href="https://www.techtarget.com/searchenterpriseai/tip/AI-inference-vs-training-Key-differences-and-tradeoffs">Inference uses trained models</a> to make predictions. RNGD is based on tensor contraction processor architecture and is manufactured using Taiwan Semiconductor Manufacturing Co.'s 5 nm process node. The company claims that RNGD achieves 2.25 times better per-watt performance for inference over GPUs, which have become the most popular processors for AI workloads. "I think FuriosaAI is smart to go after the inference market as its primary target ... It is far more open and lucrative than the training market," said Matt Kimball, vice president and principal analyst at Moor Insights & Strategy, in an interview. <section class="section main-article-chapter" data-menu-title="Nvidia's hold on data center"> <h2 class="section-title">Nvidia's hold on data center</h2> Toppling GPU juggernaut Nvidia is easier said than done. Nvidia enjoyed a 98% data center GPU <a target="_blank" href="https://siliconanalysts.com/nvidia-ai-accelerator-market-outlook-2023-2027" rel="noopener">market share in 2023</a>, according to Silicon Analysts. Competitors claiming performance wins over Nvidia's hardware still have a mountain to climb -- Nvidia's popular CUDA software helps the company maintain its comfortable grip on AI workloads. GPU challenges from AMD and Intel have done little to loosen that grip. "Training is dominated by Nvidia GPUs and CUDA, with AMD showing potential to eat some of that share on the enterprise side," Kimball said. "Does [FuriosaAI's RNGD] pose a threat to Nvidia? I don't think so. ... RNGD seems to have scored in its niche area." Nvidia leads in the inference market with its A100, H100, L4, L40 and Blackwell GPUs. But the inference side is more diverse, with competition from Google using custom-built TPU v4i chips, Amazon with in-house Inferentia chips, Intel CPUs, AMD, Qualcomm and several startup chipmakers. </section> <section class="section main-article-chapter" data-menu-title="Finding a niche in AI inference"> <h2 class="section-title">Finding a niche in AI inference</h2> Inference is gaining momentum and is expected to be a larger market than training in the coming years. The global AI inference market was estimated at $106 billion in 2025 and projected to grow to $255 billion by 2030, according to a MarketsandMarkets <a target="_blank" href="https://www.marketsandmarkets.com/Market-Reports/ai-inference-market-189921964.html" rel="noopener">report</a>. Kimball said FuriosaAI's technology is aimed at a very specific low-latency inference, and that Nvidia's GPUs will still be valuable on the inference side, especially for extremely large models. "RNGD has an architecture that lends itself to high-query-per-second and lower-precision kind of inference. ... Think recommendation engines, customer service bots, programmatic advertising, etc.," Kimball said. Other big tech players clearly see potential in the growing inference market. In March, FuriosaAI turned down Meta's $800 million bid to acquire the company. "I believe the inference market is wide open. Partly because of its size, and partly because of the diversity of use cases and deployment scenarios," Kimball said. Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> The South Korean semiconductor startup scored LG as its first major customer, as companies compete to take advantage of the AI boom. https://cdn.ttgtmedia.com/rms/onlineimages/maze_g676210320.jpg https://www.techtarget.com/searchdatacenter/news/366627769/FuriosaAI-to-fuel-LG-Exaone-LLM-Is-it-a-challenge-to-Nvidia Tue, 22 Jul 2025 16:37:00 GMT FuriosaAI to fuel LG Exaone LLM: Is it a challenge to Nvidia? Meta and Google are the latest to trade blows in the AI data center arms race, promising billions of dollars in new investments for massive infrastructure buildouts. Meta CEO Mark Zuckerberg on Monday said the first in a series of AI data "superclusters" will come online next year as part of a plan to invest "hundreds of billions of dollars" to secure its spot as a dominate player in artificial intelligence. Superclusters are <a href="https://www.techtarget.com/searchdatacenter/feature/How-the-rise-in-AI-impacts-data-centers-and-the-environment">massive AI data centers</a> powered by the most advanced AI accelerators. Zuckerberg said in a <a target="_blank" href="https://www.threads.com/@zuck/post/DMF6tMAxkX8?xmt=AQF0I6-Ctn0leigUTf6ExxSz9KzG2zhB17D-AQ5kfFzBZQ" rel="noopener">post on Threads</a>, Facebook's social media platform, that Prometheus would be one of several sites to provide multi-gigawatt computing power. Another planned supercluster, Hyperion, could be scaled to 5 gigawatts of power. <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> Meta isn't just building data centers. It's building bargaining power. </figure> <figcaption> Pradeep SanyalAI and data leader, Capgemini </figcaption> </div> </blockquote> "Meta isn't just building data centers," said Pradeep Sanyal, AI and data leader at Capgemini, in an email interview. "It's building bargaining power. … Compute supremacy is now the new battleground for frontier AI. With gigawatt-scale clusters, Meta is buying its way into a class of infrastructure that only a handful of players can match." Big tech's appetite for AI advancement has hyperscalers and AI companies racing to build such clusters. Meta's AI <a href="https://www.techtarget.com/searchenterpriseai/news/252512373/Meta-AI-supercomputer-looks-toward-the-metaverse">Research SuperCluster</a> has been operational since 2022. That cluster uses 16,000 Nvidia A100 GPUs for large language model (LLM) training. Amazon and Anthropic announced a joint supercluster this year -- <a target="_blank" href="https://www.aboutamazon.com/news/aws/aws-project-rainier-ai-trainium-chips-compute-cluster" rel="noopener">Project Ranier</a> -- that will use AWS' Trainium2 chips. Elon Musk's xAI effort, <a target="_blank" href="https://x.ai/colossus" rel="noopener">Colossus</a>, is powered by 200,000 Nvidia chips that require <a target="_blank" href="https://energy.media/climate_deals/elon-musks-data-center-company-xai-to-import-overseas-power-plants/" rel="noopener">300 megawatts of power</a>. <section class="section main-article-chapter" data-menu-title="Google's power play"> <h2 class="section-title">Google's power play</h2> The propulsion of AI computing requires a massive amount of power, and big tech companies are competing to shore up <a href="https://www.techtarget.com/whatis/feature/Three-tech-companies-eyeing-nuclear-power-for-AI-energy">energy pipelines</a>. On Tuesday, Google said it would invest $25 billion in data centers and AI infrastructure within the nation's <a target="_blank" href="https://www.ferc.gov/industries-data/electric/electric-power-markets/pjm" rel="noopener">largest U.S. electric grid</a> -- which covers 13 states -- over the next two years. Google also said it would spend $3 billion to update two hydropower plants in Pennsylvania to help power the energy-hungry data centers. The Pennsylvania investment is part of <a target="_blank" href="https://bam.brookfield.com/press-releases/brookfield-and-google-sign-hydro-framework-agreement-deliver-3000-mw-homegrown" rel="noopener">an agreement</a> with Brookfield Asset Management to buy 3,000 megawatts of carbon-free hydroelectric power in the U.S. Google's parent company, Alphabet, earlier this year said it would <a target="_blank" href="https://www.cfodive.com/news/alphabet-cfo-says-capex-reach-ai-push-grows/739355/" rel="noopener">spend $75 billion</a> in 2025 on data centers and infrastructure. In June, the <a target="_blank" href="https://blog.google/feed/ctc-global-partnership-us-electrical-grid-capacity/" rel="noopener">company said it</a> would partner with CTC Global Corp. to use new technology to free up grid capacity across the country. Sanyal said there's a stark contrast between the AI strategies from Meta and Google. "While Meta is betting on bespoke AI superclusters, Google is taking a more balanced route," he said. "They are pairing aggressive data center expansion with long-term clean energy procurement. Their $3 billion hydro deal secures up to 3 gigawatts renewable power, giving them not just capacity but predictability." While Meta is going in on scale-first buildouts, Google is threading compute, sustainability and grid alignment, he said, adding that "both are aiming for leadership." </section> <section class="section main-article-chapter" data-menu-title="Meta's AI gamble"> <h2 class="section-title">Meta's AI gamble</h2> Sanyal said Meta's strategy might be to horde AI infrastructure and talent. The company has already poached top talent from competitors, including OpenAI and Apple, offering lucrative signing bonuses. Meta hired former Safe Superintelligence CEO Daniel Gross after a failed attempt to buy the company, according to published reports. "Meta is trying to force its way back to the center of the AI conversation using compute, capital and headcount," Sanyal said. "This also comes with serious execution risk. Scaling that kind of infrastructure is hard. Talent integration at this level is even harder. And alignment, safety and public trust will only get trickier the more power you throw at the system." Whether Meta executes on its strategy is still to be determined, but the outcome could be critical either way. "If Meta executes, this could reset the pecking order in AI," he said. "And if they don't, it could be the most expensive HR experiment in history." Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> Meta and Google tout aggressive AI infrastructure investments focused on data center builds and power. https://cdn.ttgtmedia.com/rms/onlineimages/storage_g1180684429.jpg https://www.techtarget.com/searchdatacenter/news/366627578/Meta-Google-unveil-massive-AI-data-center-investment-plans Tue, 15 Jul 2025 16:41:00 GMT Meta, Google unveil massive AI data center investment plans Oracle and AWS have partnered to run Oracle's database services in Oregon and Northern Virginia, with plans to soon expand availability. Oracle Database@AWS enables customers to run Oracle Exadata Database Service and Oracle Autonomous Database on Oracle Cloud Infrastructure within AWS. The move allows easier data integration between Oracle Database tools and AWS Analytics. The companies said the partnership expands customers' ability to run Oracle databases in the cloud. This is Oracle's latest effort to expand <a href="https://www.computerweekly.com/news/366610215/At-CloudWorld-Oracle-completes-hyperscaler-trifecta-with-AWS">partnerships with hyperscalers</a> as part of its cloud transformation strategy. Oracle's <a target="_blank" href="https://investor.oracle.com/investor-news/news-details/2025/Oracle-Announces-Fiscal-2025-Fourth-Quarter-and-Fiscal-Full-Year-Financial-Results/default.aspx" rel="noopener">financial results</a> for the fourth quarter of 2025 showed its cloud infrastructure revenue ballooned 52%, while database multi-cloud database revenue from Amazon, Google and Azure surged 115% quarter over quarter. Early adopters include Fidelity Investments, Nationwide and SAS. "Oracle is the data management platform for so many enterprise customers -- customers that may have already bought into AWS … and using services from [cloud service providers] like AWS Redshift for analytics," said Matt Kimball, data center analyst at Moor Insights & Strategy. "Rather than force those customers to migrate their Oracle environments to OCI [<a title="https://www.techtarget.com/searchcloudcomputing/feature/oracle-cloud-infrastructure-targets-devs-and-installed-base" target="_blank" rel="noreferrer noopener" href="https://www.techtarget.com/searchcloudcomputing/feature/Oracle-Cloud-Infrastructure-targets-devs-and-installed-base" id="menurkt4" aria-label="Link Oracle Cloud Infrastructure">Oracle Cloud Infrastructure</a>], they can meet them where they are." Oracle will bring the 20 additional regions online over the next 12 to 18 months, according to Kambiz Aghili, Oracle's vice president of product for multi-cloud at OCI. "When they come live, they will come fully live with everything we offer," he said. Aghili said Oracle's strategy is to give customers more choice. The company previewed Database@AWS starting in December, which showed that customers could move workloads to the cloud that were not possible previously, he said. "The strategy of the multi-cloud for us is making our technology available and interoperable, regardless of whether customers want to run it on OCI, on our data centers or in third-party or partner data centers." Rolling out Database@AWS in Northern Virginia and Oregon makes sense from a strategic standpoint. Ashburn, Va., known as "Data Center Alley," has the largest concentration of data centers in the world, and all hyperscalers -- Meta, Google and Amazon -- have built large data centers in Oregon as well. But Steven Dickens, CEO and analyst at HyperFrame Research, noted this is just the beginning. "Ultimately, this is going to play out in all the regions for all the cloud providers globally," he said. "Europe would be another obvious step from a sovereignty point of view, and it will be interesting to see if it gets rolled out in the Middle East and Japan." <section class="section main-article-chapter" data-menu-title="Partnership boosts AI capabilities"> <h2 class="section-title">Partnership boosts AI capabilities</h2> Oracle Database@AWS will enable customers working with OCI in AWS to use <a href="https://www.techtarget.com/esg-global/blog/oracle-announces-the-oracle-database-23ai-for-ai-workloads/">Oracle Database 23ai</a> with embedded AI vector capabilities, according to the company. With zero extract, transform and load integration, data moves between Oracle Database services and AWS services, letting customers combine data with AWS analytics, machine learning and generative AI. "Data and its residency are key," Dickens said. "This [partnership] answers the fundamental question, 'Do we move data to the AI or AI to the data.' This just fixes that dilemma." OCI users will also gain access to AWS tools including Management Console, Command Line Interface, APIs and monitoring tools. Oracle Database@AWS includes support for Oracle applications, including E-Business Suite, PeopleSoft, JD Edwards EnterpriseOne, Oracle Enterprise Performance Management and Oracle Retail Applications. "In these partnerships, Oracle is natively deploying OCI, including Exadata, in its partner data centers so they can deliver secure, low-latency connectivity to services," Kimball said. "This is the very definition of true multi-cloud." Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> Oracle Database@AWS combines the companies' tools to eliminate costly and complex data pipeline buildouts while adding to its collection of agreements with hyperscalers. https://cdn.ttgtmedia.com/rms/onlineimages/cloud_g1135435124.jpg https://www.techtarget.com/searchdatacenter/news/366627494/Oracle-AWS-partner-for-cloud-database-boost Thu, 10 Jul 2025 15:47:00 GMT Oracle, AWS partner for cloud database boost LAS VEGAS -- Under the backdrop of a towering Sphere display, HPE CEO Antonio Neri on Tuesday laid out the company's vision for networking in the agentic AI age with new hardware and partnerships for its key infrastructure products. Pointing to product launches and upgrades for GreenLake, Aruba Networking, and Private Cloud AI, Neri said HPE is focused on providing enterprises a seamless experience in any stage of AI adoption. Early-stage adopters like Brandon O'Driscoll, an Azure systems engineer for clean energy company NextEra Energy, said HPE's product launches were welcome news. "We're still doing a lot of assessment, but to have a turnkey solution -- that would be really helpful," O'Driscoll said. <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/hpe_ceo_antonio_neri_at_hpe_discover_2025-f.jpg"> <img data-src="https://www.techtarget.com/rms/onlineimages/hpe_ceo_antonio_neri_at_hpe_discover_2025-f_mobile.jpg" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/hpe_ceo_antonio_neri_at_hpe_discover_2025-f_mobile.jpg 960w,https://www.techtarget.com/rms/onlineimages/hpe_ceo_antonio_neri_at_hpe_discover_2025-f.jpg 1280w" alt="HPE CEO Antonio Neri at HPE Discover" data-credit="Shane Snider/Informa TechTarget" height="373" width="560"> <figcaption> Antonio Neri, CEO at HPE, speaks Tuesday at the company's Discover conference in Las Vegas. </figcaption> <div class="main-article-image-enlarge"> </div> </figure> <section class="section main-article-chapter" data-menu-title="Agentic AI focus"> <h2 class="section-title">Agentic AI focus</h2> Similar to other major IT infrastructure vendors recently, HPE's key releases had a common thread: Agentic AI will be at the core of its networking and hybrid cloud efforts for the foreseeable future. "With Agentic AI, AI is no longer just a tool -- it is actively engaged, automating workflows, making real-time decisions and driving business efficiency," Neri said during a keynote presentation at HPE Discover. "Soon, it will be IT's responsibility to manage a digital workforce of thousands of specialized agents working across the enterprise." <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> Soon, it will be IT's responsibility to manage a digital workforce of thousands of specialized agents working across the enterprise. </figure> <figcaption> Antonio NeriCEO, HPE </figcaption> </div> </blockquote> Touting a growing list of partners, <a href="https://www.techtarget.com/searchenterpriseai/news/366626405/HPE-beefs-up-AI-factory-fueled-offerings-with-Nvidia-upgrades">including Nvidia, Accenture</a>, AMD and many others, Neri said the rapid innovation fueling HPE's ambitious AI goals addresses a growing need for the entire IT ecosystem. From new servers that will take up less rack space as demand pushes the data center to its physical limits to unleashing AI agents, HPE claims its portfolio aims to ease the pressure on IT leaders to adopt AI quickly. Agentic AI has been a common thread among enterprise technology providers this year, with agentic systems powering everything from apps and <a href="https://www.techtarget.com/searchsecurity/opinion/RSAC-2025-to-center-on-agentic-AI-GenAI-in-security">security software</a> to <a href="https://www.techtarget.com/searchnetworking/news/366625781/Cisco-unleashes-agentic-AI-global-partnerships-for-data-center">data center infrastructure</a>. Aristos Dimitriou, an architect at aerospace manufacturing company Raytheon, said agentic AI capabilities could lead to new efficiencies that speed up workflows between companies and their customers. "There are a lot of possibilities," Dimitriou said, noting that his company mostly works with the government. "Government can be really, really slow. So, what HPE is doing is a key driver to improving speed through those new offerings. </section> <section class="section main-article-chapter" data-menu-title="Unleashing AI agents for Aruba"> <h2 class="section-title">Unleashing AI agents for Aruba</h2> Neri said HPE's products address key pain points for customers who are juggling AI needs with rising costs and rapidly changing infrastructure. "Much of today's infrastructure is made up of layers of legacy IT that's been built over decades and is increasingly difficult to manage and modernize," he said. "Your cloud and virtualization costs are rising, and many of you are pushing back on commercial business models and solutions that lock you in." Neri said the company's new agentic AI mesh in Aruba Central, the cloud-based network management platform, would arm enterprises with secure networking automation. "Paired with a networking copilot, these AI agents leverage a collection of reasoning models context-tuned for security-first, AI-powered networking," Neri said. "Think of them as an extra admin team that never sleeps, always working for you. The result is a network that thinks for itself -- autonomously identifying problems, offering solutions and getting smarter over time." </section> <section class="section main-article-chapter" data-menu-title="Building a modern hybrid cloud"> <h2 class="section-title">Building a modern hybrid cloud</h2> Neri pointed to research from Enterprise Strategy Group, now part of Omdia, that showed <a href="https://www.techtarget.com/searchnetworking/feature/AI-adoption-in-networks-is-the-norm-despite-its-infancy">AI has forced the majority of enterprises</a> to rethink deployment plans. Neri said GreenLake allows control over multi-customer, multi-cloud environments, across multiple vendors and multiple generations of IT infrastructure. "You need the flexibility to run your workloads where it makes the most sense -- across clouds, [colocations] and the edge -- maintaining speed and agility, but with more control," he said. "That is the value of hybrid cloud. … For AI to be a truly transformative force in your organization, you need an integrated operational hybrid stack that flows as one with your infrastructure." Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> The infrastructure titan launched several new products and services aimed at bolstering customers' appetite and abilities for AI workload optimization. https://cdn.ttgtmedia.com/rms/onlineimages/ai_a252657224.jpg https://www.techtarget.com/searchdatacenter/news/366626446/Agentic-AI-at-the-core-of-HPE-networking-GreenLake-updates Tue, 24 Jun 2025 18:36:00 GMT Agentic AI at the core of HPE networking, GreenLake updates <div class="podcastdownload alignCenter"> <audio id="podcastPlayer" src="https://cdn.ttgtmedia.com/rms/podcasts/0618.TexasInstrumentsChips.mp3" type="audio/mp3" controls="controls"></audio> Listen to this article. This audio was generated by AI. </div> Texas Instruments on Wednesday launched a $60 billion plan to boost domestic semiconductor manufacturing across seven facilities in Utah and Texas -- the largest effort by a single U.S. company in history. The vendor said the investment will help create a <a href="https://www.techtarget.com/searcherp/news/252523892/CHIPS-Act-should-simplify-semiconductor-supply-chains">domestic supply chain</a> that will support semiconductor needs for everything from vehicles to smartphones and data centers. In addition to the manufacturing investment, partnerships with Nvidia, Apple, Ford and SpaceX will also help spur 60,000 new U.S. jobs, according to the company. "TI is building dependable, low-cost 300mm capacity at scale to deliver the analog and embedded processing chips that are vital for nearly every type of electronic system," Haviv Ilan, president and CEO of Texas Instruments, said in a statement. Low-cost 300mm refers to the size of the silicon wafer. The investment will bring two more fabrication plants to Sherman, Texas, upping the company's total to four at that site. The company will also use the funds to increase production at its sites in Richardson, Texas, and Lehi, Utah. TI said hundreds of millions of chips will be produced daily at the sites in Sherman, Texas, which will provide 1.3 million square feet of clean room space for manufacturing. The Sherman sites alone will cost up to $40 billion of the planned investment, according to the company. Stephen Sopko, an analyst at HyperFrame Research, said the implications for the domestic <a href="https://www.techtarget.com/whatis/feature/The-history-of-semiconductors-and-the-chip-making-industry">semiconductor supply</a> are substantial, but realizing gains from increased manufacturing will take time. "The fabs aren't built quickly," he said. "The [types of advanced semiconductors] announced are key for U.S. domestic electronics and industrial manufacturing to expand in the coming years." <section class="section main-article-chapter" data-menu-title="300mm wafer supply boost"> <h2 class="section-title">300mm wafer supply boost</h2> AI infrastructure gets a boost from 300mm silicon wafers in several ways, as they are used in manufacturing complex processors and chips required for AI accelerators used in advanced AI systems and data centers. AI GPU leader Nvidia uses the 300mm wafers as a foundation for its AI chips. "Nvidia and TI share the goal to revitalize U.S. manufacturing by building more of the infrastructure for AI factories here in the US," Jensen Huang, founder and CEO of Nvidia, said in a statement. In April, <a href="https://www.ciodive.com/news/nvidia-domestic-ai-infrastructure-chip-manufacturing-tariffs/745441/">Nvidia announced plans</a> to build AI supercomputers in Texas with Nvidia Blackwell chips produced at Taiwan Semiconductor Manufacturing Co.'s Arizona plant. Texas Instruments is already a major supplier for Apple, which uses TI chips in its devices for power management, audio processing and sensor interfaces. TI chips are found in Apple's specialized devices, such as the Apple Pencil Pro and Vision Pro VR headset. SpaceX also uses TI's 300mm SiGe technology manufactured in Sherman, Texas, for its global Starlink satellite internet service. Sopko said TI is strengthening its position in the semiconductor ecosystem. He said the fab expansions are "essential for their customer's power management, microcontrollers and sensors in just about all electronics manufacturing. Few companies build high-volume 300mm wafer capacity for these technologies in the U.S., so TI strengthens its dominance in the analog segment." TI credited President Donald Trump's administration efforts and support. The administration has used ongoing tariff negotiations to encourage domestic manufacturing. Former President Joe Biden signed the CHIPS and Science Act, which <a target="_blank" href="https://bidenwhitehouse.archives.gov/briefing-room/statements-releases/2024/08/09/fact-sheet-two-years-after-the-chips-and-science-act-biden-%E2%81%A0harris-administration-celebrates-historic-achievements-in-bringing-semiconductor-supply-chains-home-creating-jobs-supporting-inn/" rel="noopener">allocated</a> $53 billion to support domestic semiconductor manufacturing, into law in 2022. In the past, Trump has <a target="_blank" href="https://www.youtube.com/watch?v=1zIHZtD3Geg" rel="noopener">criticized</a> the law, telling Congress during a March address, "You should get rid of the CHIPS Act," and calling the Act "horrible." And his administration has been working to renegotiate grants to semiconductor firms. In December, TI said it would receive up to $1.6 billion in CHIPS Act funding to support the sites in Texas and Utah. A company spokesperson said those funds would not be used in this latest endeavor. Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> The company will partner with Nvidia, Apple, Ford, SpaceX and others to spur domestic semiconductor production. https://cdn.ttgtmedia.com/rms/onlineimages/maze_g676210320.jpg https://www.techtarget.com/searchdatacenter/news/366626281/Texas-Instruments-injects-60B-into-US-chip-manufacturing Wed, 18 Jun 2025 14:29:00 GMT Texas Instruments injects $60B into U.S. chip manufacturing Hewlett Packard Enterprise added two new server platforms to its Nonstop Compute line that promise to increase overall performance for business users needing uninterrupted processing power. The entry-level Nonstop Compute NS5 X5 and flagship NS9 X5 aim to boost performance for critical enterprise operations such as payment processing, <a href="https://www.techtarget.com/searchsecurity/definition/fraud-detection">fraud detection</a> and smart manufacturing execution systems. The platforms will use <a href="https://www.intel.com/content/www/us/en/products/details/processors/xeon/scalable/bronze.html">Intel Xeon Bronze</a> 3400 series and <a href="https://www.intel.com/content/www/us/en/products/details/processors/xeon/scalable/gold.html">Gold 6400</a> series processors to increase performance capacity by up to 15%, the company said. HPE's Nonstop Compute platforms are aimed at financial services institutions, payment processors, retail merchants, manufacturing companies and transportation businesses that require constant transactional processing. The NS5 X5 and NS9 X5 platforms offer up to 8 TB of memory, more than twice the capacity of predecessor units, and more than double networking bandwidth, the company said. The release of the updated platforms coincides with the company's OS update, which adds support for <a href="https://www.techtarget.com/searchsecurity/definition/multifactor-authentication-MFA">multifactor authentication (MFA)</a>. The new features and support will help enterprises comply with mandated security standards, including the EU's GDPR, Payment Card Industry Data Security Standard, HIPAA standards and Service Organizational Control 2 auditing. "MFA is now seen as table stakes," said Steven Dickens, CEO and principal analyst at HyperFrame Research, noting tougher security requirements in the U.S. and abroad. "MFA is becoming mandatory, not optional." Casey Taylor, HPE vice president and general manager of the company's Nonstop Compute business, said the releases offer a refresh while maintaining HPE's goal of providing products that offer uninterrupted operation. The new units can be clustered with the past two generations of Nonstop Compute platforms. "This is another converged, bare metal system with enhanced software and hardware that is shipped to our customers in a white-glove way," Taylor said. "We need to keep up with our customers' roadmaps -- and these platforms are <a name="_Hlk200982879"></a>backward-integrated with our private platforms and latest software releases. The hardware comes with new features and functions that our customers are looking for." Dickens said the Nonstop Compute segment of HPE's business is important, if unheralded, especially for government and critical business infrastructure. <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> It's absolutely vital for those basic customers that these platforms keep moving forward and evolving. </figure> <figcaption> Steven DickensCEO and principal analyst, HyperFrame Research </figcaption> </div> </blockquote> "There are scenarios where failure and outages are not an option," he said. "These mission-critical systems have a place for certain workloads, whether that's payment processing, credit card transactions -- use cases where downtime for a second of the day, planned or unplanned, can have a huge impact on the business." He added that, while nonstop computing is often overshadowed by supercomputing, the advances in HPE's portfolio display a commitment to businesses that rely on the technology. "It's absolutely vital for those basic customers that these platforms keep moving forward and evolving," he said. <section class="section main-article-chapter" data-menu-title="AI 'adjacent' compute power"> <h2 class="section-title">AI 'adjacent' compute power</h2> While the new platforms are not meant to offer AI workload processing power -- neither platform uses GPUs or data processing unit processors -- they offer efficiencies meant to work in tandem with AI-intense operations. Banks and other enterprises are already running AI to aid in fraud detection, and the upgraded Nonstop platforms complement those uses with reliable data processing. "Customers are not running AI on the platform itself, but they want AI-adjacent ability," Taylor said, noting one customer's use of AI in concert with Nonstop platforms. "We're plugging AI into it with those types of periphery services and software, but not running an AI workload." She added, "Every business and every company is looking at how they can harness the power of AI -- and our customers in the Nonstop world are no different. The difference is that for us, it's about the products and services that we can wrap around this product that can complement AI." Dickens said HPE is sticking to <a href="https://www.techtarget.com/searchdatacenter/news/366613916/Intel-AMD-form-x86-group-in-effort-to-combat-Arm">x86 technology</a> because it works well for the intended use -- even as competitors like Dell and IBM add GPUs to their nonstop offerings. "It's a trade-off. Where do you want AI to work in this stack? These are transactional systems at the end of the day," he said. "It's interesting to see vendors taking different approaches. Nonstop customers probably don't need to be on the absolute bleeding edge. They are making the trade-off of having better availability." Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> The company says its latest releases will fuel efficiency for enterprise customers, with the help of Intel Xeon processors. https://cdn.ttgtmedia.com/rms/onlineimages/storage_g1197646065.jpg https://www.techtarget.com/searchdatacenter/news/366626160/HPE-doubles-Nonstop-server-platform-memory-bandwidth Mon, 16 Jun 2025 16:29:00 GMT HPE doubles Nonstop server platform memory, bandwidth Amazon on Wednesday said it would invest $10 billion to expand data center infrastructure in North Carolina to boost artificial intelligence and cloud computing technology, including plans to build a 20-building campus. The company said its investment would support AI initiatives creating 500 new high-skilled jobs. Amazon has already invested $12 billion in North Carolina, supporting 24,000 full- and part-time jobs directly. During the company's fourth-quarter earnings call in February, Amazon CEO Andy Jassy said the company planned to boost capital expenditures to $100 billion this year, mostly in AI investments. <a href="https://www.techtarget.com/searchaws/definition/Amazon-Web-Services">AWS</a> is the current market leader as a cloud provider, with stiff competition from <a href="https://www.techtarget.com/searchitchannel/news/366612147/Hyperscalers-mass-AI-role-dominate-partner-ecosystems">Google Cloud Platform, Microsoft Azure and others</a> jockeying for position in the AI race. "We don't procure it unless we see significant signals of demand," Jassy said during <a target="_blank" href="https://www.fool.com/earnings/call-transcripts/2025/02/06/amazoncom-amzn-q4-2024-earnings-call-transcript/" rel="noopener">the call</a>. "And so, when AWS is expanding its Capex, particularly in what we think is one of these once-in-a-lifetime type of business opportunities like AI represents, I think it's actually quite a good sign, medium to long term, for the AWS business." Alan Howard, principal analyst for data centers and colocation services in the data research practice at Omdia, a division of Informa TechTarget, said Amazon is creating partnerships with states as part of a broader data center strategy. "Amazon has been at this game for a long time," he said. "When they move to a community with a data center footprint, it can be rather significant ... They've learned that the road is much less bumpy if they approach the market as a partner, investing not just in infrastructure and jobs, but supporting the community with programs, renewable energy projects and other forms of just being a good neighbor." Howard added, "Many data center companies need to take this kind of approach to market entrance appropriate to their scale." <section class="section main-article-chapter" data-menu-title="Adding to N.C. tech center"> <h2 class="section-title">Adding to N.C. tech center</h2> North Carolina is home to a thriving tech sector, with Raleigh and the Research Triangle Park attracting significant tech job growth in recent years. Amazon said its previous investments in North Carolina have brought $13.1 billion to the state's gross domestic product, according to a <a target="_blank" href="https://www.aboutamazon.com/news/aws/aws-investment-north-carolina-ai-cloud-infrastructure" rel="noopener">blog post</a>. Amazon does not disclose the number of data centers currently in the state, but the company has been actively investing in North Carolina since 2010. As part of the latest investment, the company said it plans to support workforce development training programs with data center technician programs and <a href="https://www.techtarget.com/whatis/definition/STEM-science-technology-engineering-and-mathematics">STEM</a> initiatives aimed at K-12 schools. Jobs supported by the investment will include data center engineers, network specialists, engineering operations managers, security specialists and other technical roles, Amazon said. "The additional high-tech investments and the need for advanced STEM skills is a perfect fit," said Marc Hoit, vice chancellor for information technology and CIO of North Carolina State University, in an email interview. "We look forward to supporting Amazon's growth and investment in N.C." Brooks Raiford, CEO of the North Carolina Technology Association (NC Tech), said Amazon's data center play will have a big impact in the state. "The Amazon announcement furthers North Carolina's reputation as a leading region for data centers and artificial intelligence," he said. "It's exciting to see their plans to engage with educational institutions and workforce development organizations." NC Tech's latest <a target="_blank" href="https://www.nctech.org/_files/_pdf/talent/NC%20Tech%20State%20of%20the%20Technology%20Industry%20Report%202025.pdf" rel="noopener">state-of-the-industry report</a> showed that as of 2023, the technology industry employed more than 323,000 of North Carolina's 4.8 million workers, or 6.7%. The state was ranked eighth nationwide in tech industry employment growth from 2018 to 2023, according to the report. Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> As Amazon races to compete with Big Tech rivals for AI dominance, its Tar Heel State investment is part of a $100 billion capital expenditure effort slated for 2025. https://cdn.ttgtmedia.com/rms/onlineimages/storage_g1197646065.jpg https://www.techtarget.com/searchdatacenter/news/366625221/Amazon-to-launch-10B-data-center-upgrade-in-North-Carolina Wed, 04 Jun 2025 17:31:00 GMT Amazon to launch $10B data center upgrade in North Carolina <div class="podcastdownload alignCenter"> <audio id="podcastPlayer" src="https://cdn.ttgtmedia.com/rms/podcasts/06052025_sDC_Meta_inks_20-year_nuclear_deal.mp3" type="audio/mp3" controls="controls"></audio> Listen to this article. This audio was generated by AI. </div> Meta Platforms has signed a 20-year nuclear energy agreement with Constellation Energy for Meta's Illinois-based data center, drawing from the Clinton Clean Energy Center. Hyperscalers, including Meta, Microsoft, Google and Amazon, are racing to meet unprecedented enterprise AI demand spurred by the success of <a href="https://www.techtarget.com/searchenterpriseai/definition/OpenAI">OpenAI's ChatGPT</a>. In 2023, <a href="https://sustainability.atmeta.com/wp-content/uploads/2023/07/Meta-2023-Environmental-Data-Index.pdf">Meta disclosed</a> that it had used 2.6 million kilowatt-hours of electricity and emitted 1,000 tons of carbon dioxide while developing its <a href="https://www.techtarget.com/whatis/feature/Meta-Llama-4-explained-Everything-you-need-to-know">Llama</a> large language models. According to Statista, Meta's global power consumption soaked up more than 15 terawatt-hours in 2023, a 33% increase over the previous year. The company has a goal of using nuclear energy for up to 4 gigawatts of energy consumption in the coming years as a generative AI arms race heats up. Meta's deal with Constellation will start in 2027 and explore other nuclear energy opportunities. It remained operational after the Illinois state government introduced a zero-emissions credit program, which is set to expire in 2027. In December, Meta sent out a <a target="_blank" href="https://sustainability.atmeta.com/blog/2024/12/03/accelerating-the-next-wave-of-nuclear-to-power-ai-innovation/" rel="noopener">nuclear energy request for proposal</a> to meet its growing power needs. On Tuesday, the company said in a <a target="_blank" href="https://about.fb.com/news/2025/06/meta-constellation-partner-clean-energy-project/" rel="noopener">blog post</a> that it received more than 50 submissions from nuclear energy operators across more than 20 states. "Through these projects, we aim to activate investment in new nuclear across multiple grids and technologies, seeding new reliable firm power to support future data centers," according to the blog post. "Our investments in nuclear energy ensure that we will have the robust energy infrastructure needed to power the AI innovations that are set to spark economic growth and prepare our communities for the future." The Clinton Clean Energy Center was slated to close in 2017, after years of financial losses. Constellation said Meta's deal will ensure long-term operations at the plant without ratepayer support. "[Meta] figured out that supporting the relicensing and expansion of existing plants is just as impactful as finding new sources of energy," Joe Dominguez, president and CEO of Constellation, said in a <a target="_blank" href="https://www.constellationenergy.com/newsroom/2025/constellation-meta-sign-20-year-deal-for-clean-reliable-nuclear-energy-in-illinois.html" rel="noopener">statement</a>. Meta said such deals are crucial to its future. It has already invested more than $1 billion into its DeKalb Data Center in Illinois, which broke ground in 2020. "Securing clean, reliable energy is necessary to continue advancing our AI ambitions," Urvi Parekh, Meta's head of global energy, said in a statement. Details of the Constellation deal were not disclosed. Google has also unveiled <a target="_blank" href="https://blog.google/feed/google-and-elementl-nuclear-energy-site-development/" rel="noopener">plans to fund</a> three new nuclear projects to meet compute needs. Last year, Microsoft <a target="_blank" href="https://www.constellationenergy.com/newsroom/2024/Constellation-to-Launch-Crane-Clean-Energy-Center-Restoring-Jobs-and-Carbon-Free-Power-to-The-Grid.html" rel="noopener">announced a plan</a> to reopen Pennsylvania's Three Mile Island nuclear power plant -- also owned by Constellation Energy -- to meet its growing AI needs, and <a>Amazon </a> <a target="_blank" href="https://www.aboutamazon.com/news/sustainability/amazon-nuclear-small-modular-reactor-net-carbon-zero" rel="noopener">is eyeing</a> small modular nuclear reactors. While Meta is making strides on its renewable energy commitment, it cannot seem to shake carbon-emitting energy entirely. The company's planned Louisiana data center -- a $10 billion project that will be Meta's largest data center -- will construct three natural gas turbines along with renewable energy sources. <section class="section main-article-chapter" data-menu-title="What the nuclear deal means for IT leaders"> <h2 class="section-title">What the nuclear deal means for IT leaders</h2> CIOs are under pressure to quickly adopt AI, while keeping costs and sustainability in mind. AI demand could have companies -- both hyperscalers and enterprises -- investing $5.2 trillion into data centers by 2030, <a target="_blank" href="https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/the-cost-of-compute-a-7-trillion-dollar-race-to-scale-data-centers" rel="noopener">according to a McKinsey report</a>. Meta said its deal with Constellation will allow for 1,121 megawatts of emissions-free energy, reducing its environmental footprint. That could be a win for CIOs who have to show gains both in AI adoption and sustainability efforts. <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> Meta has found a creative way to get enough energy for this data center without having to worry about building a new power plant. </figure> <figcaption> Bob JohnsonAnalyst, Gartner </figcaption> </div> </blockquote> Bob Johnson, a vice president and analyst at Gartner, said the environmental advantages are a plus, but the real advantage is new energy availability during a spike in demand. "There isn't enough power to run all the data centers people need for AI," he said in an interview with Informa TechTarget. "Meta has found a creative way to get enough energy for this data center without having to worry about building a new power plant." Johnson said finding alternative energy sources is a matter of meeting demand. "The best way to get power right now is through natural gas," he said. "Desirability for data centers right now depends on how close you are to a natural gas line." He added that data center demand is overwhelming now because of the current state of AI training. In the future, the energy needs might decrease. "The question is: What will AI use end up looking like? Smaller models would imply smaller data centers. We'll see." For now, AI adopters are facing a power crunch. "We may be unable to keep up with power for the next decade or two," Johnson said. Pradeep Sanyal, AI and data leader at IT consultancy Capgemini and a former CIO, said IT leaders should keep a close eye on energy moves by hyperscalers. "It's tempting to dismiss Meta's nuclear deal as overkill for regular CIOs," Sanyal said in an email interview. "But AI's energy profile is already reshaping infrastructure economics faster than most expect. Even if you're not running massive AI clusters, your cloud vendors are. Rising energy demand means high costs, tighter [service-level agreements], and more scrutiny on where your compute power actually comes from." He added, "The key is intent. Whether you act now or later, energy strategy is becoming part of IT strategy. Meta's deal is a signal. The real question is not whether to follow, but when it stops being optional." Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget. </section> The tech giant's agreement with Constellation Energy will secure an alternative source for power-hungry AI ambitions -- likely saving a once-doomed nuclear energy plant. https://cdn.ttgtmedia.com/rms/onlineimages/storage_g1180684429.jpg https://www.techtarget.com/searchdatacenter/news/366625268/Meta-inks-20-year-nuclear-deal-to-power-data-center Wed, 04 Jun 2025 08:21:00 GMT Meta inks 20-year nuclear deal to power data center <iframe title="Exploring Nvidia’s approach to AI factories and AI supercomputers" allowtransparency="true" height="150" width="100%" style="border: none; min-width: min(100%, 430px); height: 150px;" scrolling="no" data-name="pb-iframe-player" src="https://www.podbean.com/player-v2/?i=6v5dr-18c19d7-pb&from=pb6admin&share=1&download=1&rtl=0&fonts=Arial&skin=f6f6f6&font-color=auto&logo_link=episode_page&btn-skin=2baf9e" loading="lazy"></iframe> As a leading AI hardware vendor, Nvidia has long been a leading force in GPUs, far outdistancing competitors like Intel and AMD. In recent years, though, the AI vendor has begun to focus on AI factories. An AI factory is a type of computing infrastructure -- really, a large-scale data center -- that manages the entire AI lifecycle. It requires a massive amount of storage, networking and computing resources. The AI factory handles everything in the <a href="https://www.techtarget.com/searchnetworking/tip/Building-networks-for-AI-workloads">AI workload</a>, from data ingestion and training to fine-tuning and inference. <section class="section main-article-chapter" data-menu-title="Nvidia and AI factories"> <h2 class="section-title">Nvidia and AI factories</h2> Nvidia CEO Jensen Huang is a <a href="https://www.computerweekly.com/news/366570992/Nvidia-CEO-sees-shift-in-datacentres-to-AI-generation-factories">major proponent of the AI factory</a> and has been working to include it as a significant part of the vendor's AI strategy. Nvidia made a big stride toward that by acquiring Mellanox in 2019, a networking company for <a href="https://www.techtarget.com/searchdatacenter/definition/high-performance-computing-HPC">high-performance computing</a>. "It turns out that high-performance computing and AI are almost the same thing, and that's making a bunch of computers behave as a single computer," said <a target="_blank" href="https://blogs.nvidia.com/blog/author/kdeierling/" rel="noopener">Kevin Deierling</a>, senior vice president of networking at Nvidia, on the latest Targeting AI podcast episode. Deierling was senior vice president of marketing at Mellanox before Nvidia bought it. "The realization that I think Jensen had is that the computer is no longer a box that has a processor inside or even a GPU inside, but it's the entire data center," he continued. "These AI factories are indeed data centers with racks and racks and racks of GPU-based storage, but it's more than just a single computer with some cheap metal wrapped around it. We integrate within the rack. We've squeezed more and more of these GPUs together, and we scale them up." <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> It turns out that high-performance computing and AI are almost the same thing, and that's making a bunch of computers behave as a single computer. </figure> <figcaption> Kevin DeierlingSenior vice president of networking, Nvidia </figcaption> </div> </blockquote> For enterprises, the key to using AI factories is scaling the use and fine-tuning of AI models. "You want to do fine-tuning of a model, and you want to use technologies like retrieval-augmented generation inside of the enterprise. You're going to take advantage of that AI data factory building things at scale," Deierling said. </section> <section class="section main-article-chapter" data-menu-title="Some challenges"> <h2 class="section-title">Some challenges</h2> However, AI factories require a lot of energy. To reduce power demands, Nvidia is putting the GPUs in the factory closer and closer together. The vendor has also implemented <a href="https://www.techtarget.com/searchdatacenter/feature/Liquid-coolings-moment-comes-courtesy-of-AI">liquid cooling</a>. Even with its attempt to manage the energy consumption of the AI factories, Nvidia still aims to do what it is known for. "We have to continue to do what we've been doing, which is deliver higher performance per watt to keep this something that we can build and do sustainably," Deierling said. Other than Nvidia, Dell is also a player in the AI factory market. Esther Shittu is an Informa TechTarget news writer and podcast host covering AI software and systems. Shaun Sutner is senior news director for Informa TechTarget's information management team, driving coverage of AI, analytics and data management technologies, and big tech and federal regulation. Together, they host the Targeting AI podcast series. </section> The AI hardware and software provider is focusing on building massive data centers that do everything in the AI lifecycle, from data ingestion to inferencing and fine-tuning. https://cdn.ttgtmedia.com/rms/onlineimages/ai_a264431831.jpg https://www.techtarget.com/searchenterpriseai/podcast/Exploring-Nvidias-approach-to-AI-factories Tue, 03 Jun 2025 08:00:00 GMT Exploring Nvidia's approach to AI factories For IT professionals, conference season provides an excellent opportunity to gauge the state of technology options and measure their own organization's progress compared to their peers. With that in mind, Dell Technologies World 2025 presented two important lessons that IT leaders must know as we enter the second half of 2025. <section class="section main-article-chapter" data-menu-title="Lesson 1. The generative AI era is underway and your company is likely already behind"> <h2 class="section-title">Lesson 1. The generative AI era is underway and your company is likely already behind</h2> During the second-day keynote, Jeff Clarke, COO of Dell Technologies, shared some insights on the company's own internal adoption of <a href="https://www.techtarget.com/searchenterpriseai/definition/generative-AI">generative AI</a> technology to help improve productivity among sales, technical support and engineering. With the assistance of AI, according to Clarke, "individual contributors become as productive as a two-person team." Dell's internal AI adoption can serve as a blueprint for other enterprises. Our <a href="https://research.esg-global.com/reportaction/515201989/Toc">research</a> at Enterprise Strategy Group, now part of Omdia, found that the three most commonly identified primary business drivers for AI -- improving operational efficiency (cited by 48% of decision-makers surveyed), customer experience (44%) and innovation (42%) -- align with the priorities that Dell set forth internally. For AI, best practices start with quality data and the use case, but often those initial steps can create complexity, stalling efforts. Businesses can struggle to implement a proper data management strategy and lack definition in their internal processes, making it difficult to define and prioritize use cases for AI. When I asked Clarke what advice he could provide businesses looking to project potential returns on investment for AI to assist with use case prioritization, his response was to "get busy." This is an important takeaway for competitive success in AI. It will likely not come in a one-size-fits-all package. Success in AI requires your organization to conduct internal analyses to understand your data and processes, and then identify and measure the areas of greatest potential. In addition, the faster your organization starts, the faster it will develop the internal learnings necessary for success. These steps are essential prior to making a significant infrastructure investment, and help is available. For example, Dell Technologies offers services to help organizations identify the right AI use cases, prepare and clean data for AI, augment models, and provide AI infrastructure options. Dell is also not alone in providing a portfolio of AI services to help. At the event, partners such as Deloitte and Accenture were present to highlight their own portfolios of AI and data services to help organizations get started with AI. When it comes to infrastructure, the majority (52%) of new AI initiatives are deployed on public cloud services. However, there is growing interest in deploying AI initiatives on-premises, fueled by concerns tied to data privacy, cost and control of the architecture. In fact, 78% of organizations identify that they would prefer to run AI applications on-premises. To serve the demand for on-premises AI initiatives, Dell offers its <a href="https://www.techtarget.com/searchenterpriseai/opinion/Dell-AI-Factory-takes-the-spotlight-at-Dell-Technologies-World">Dell AI Factory portfolio</a>, announced last year, partnering with multiple accelerator options from Nvidia, Intel and AMD. At Dell Tech World last week, Dell announced that it has over 3,000 Dell AI Factory customers. Outside of Dell, earlier this year, HPE introduced Nvidia AI Computing by HPE, and Cisco introduced Cisco Secure AI Factory with Nvidia. Nutanix also offers its GPT-in-a-Box 2.0 and <a href="https://www.techtarget.com/searchcloudcomputing/news/366623684/Nutanix-expands-storage-Kubernetes-and-AI-platforms-at-Next">Nutanix Enterprise AI</a> to help simplify AI infrastructure deployment. In addition, nearly every enterprise storage provider offers validated or integrated AI systems, including Hitachi Vantara, IBM, Infinidat, NetApp, Pure Storage, Vast Data and WekaIO. </section> <section class="section main-article-chapter" data-menu-title="Lesson 2. A growing number of options are available to improve flexibility for hypervisor choice and migration"> <h2 class="section-title">Lesson 2. A growing number of options are available to improve flexibility for hypervisor choice and migration</h2> Recent shifts in the structure of hypervisor licenses have fueled an increased desire to explore alternatives and renewed concerns over lock-in. With that, Dell <a href="https://www.techtarget.com/searchdatacenter/news/366623831/Dell-makes-private-cloud-optionality-a-priority">launched Dell Private Cloud</a> and Dell Automation Platform, which offer validated blueprints for systems from Broadcom/VMware, Nutanix and Red Hat, simplifying the ability to provision multiple private cloud options. According to our research, 89% of organizations say the ability to use or evaluate multiple hypervisor and orchestration options is strategic. For years, businesses have focused on consistency to simplify operations. <a href="https://www.techtarget.com/searchcloudcomputing/news/366591935/Broadcom-faces-challenges-with-latest-VMware-releases">Recent cost increases</a> have shifted the priority to flexibility and ensuring that the business has multiple options. Dell Private Cloud, however, is just one option helping to simplify hypervisor flexibility. HPE recently announced <a href="https://www.techtarget.com/searchcloudcomputing/news/366623936/HPE-adds-Morpheus-Data-to-KVM-hypervisor-for-enterprises">HPE Private Cloud Business Edition with HPE Morpheus</a> VM Essentials, which leverages technology from HPE's recent acquisition of Morpheus Data to simplify multi-hypervisor support and self-service cloud consumption. In addition, Nutanix has added support for external third-party storage from Dell and Pure Storage to help simplify the adoption of its <a href="https://www.techtarget.com/searchitoperations/tip/Compare-Nutanix-AHV-vs-VMware-ESXi-in-the-hypervisor-battle">AHV hypervisor</a> technology into existing production environments. Whether your business is <a href="https://www.techtarget.com/searchdatacenter/news/366624166/Dell-Technologies-customers-building-real-world-AI-apps">focused on AI</a>, exploring private cloud alternatives or all of the above, you have far more options today than you did 12 or even six months ago. As this space evolves, the options and capabilities will likely only increase moving forward. Scott Sinclair is practice director with Enterprise Strategy Group, now part of Omdia, covering the storage industry. Enterprise Strategy Group is part of Omdia. Its analysts have business relationships with technology providers. </section> Two key insights emerged for IT professionals around enterprise AI integration and private cloud infrastructure. https://cdn.ttgtmedia.com/rms/onlineimages/cloud_g1183958722.jpg https://www.techtarget.com/searchdatacenter/opinion/AI-and-private-cloud-2-lessons-from-Dell-Tech-World-2025 Wed, 28 May 2025 10:42:00 GMT AI and private cloud: 2 lessons from Dell Tech World 2025 LAS VEGAS -- AI implementations beyond mere chatbot interactions with large language models have permeated the data centers of Dell Technologies World attendees this year. These implementations have moved into production, as these organizations attempt to sustain <a href="https://www.techtarget.com/searchenterpriseai/feature/Real-world-GenAI-applications-across-leading-industries">real-world uses</a> beyond customer support or querying a knowledge base. Enterprises that have yet to fully embrace AI technology or processes are going to be left behind in the market, said Jeff Clarke, Dell Technologies vice chairman and chief operating officer, during the Day 2 keynote Wednesday. "This is the most disruptive technology I've seen in my career," he said. "The threat [to your organization] is existential if you don't respond." At the keynote, the vendor pitched its <a href="https://www.techtarget.com/searchdatacenter/news/366624232/Dell-Tech-CEO-expects-AI-ubiquity-in-the-enterprise">Dell AI Factory</a> offering and recent storage hardware refreshes as <a href="https://www.techtarget.com/searchdatacenter/news/366623831/Dell-makes-private-cloud-optionality-a-priority">data center upgrades</a> that will empower AI development for hybrid-cloud uses. Dell executives at the keynote also confirmed the new parallel file system, codenamed <a href="https://www.techtarget.com/searchstorage/feature/With-Project-Lightning-Dell-to-strike-out-in-new-directions">Project Lighting for the past year</a>, would launch to general availability before year's end. Project Lighting is a software-defined file system built for Dell PowerEdge server hardware, specifically tested on the <a href="https://www.dell.com/en-us/shop/servers-storage-and-networking/new-poweredge-r660-rack-server/spd/poweredge-r660/pe_r660_15878_vi_vp?gacd=9650523-1130-5761040-266691960-0&dgc=ST&SA360CID=71700000062488627&gclsrc=aw.ds&gad_source=1&gad_campaignid=973870696&gclid=Cj0KCQjw0LDBBhCnARIsAMpYlApog6SmYZltRWG23C1-dKigU2e8a62vsXcDfg71496pr91wq5VI6LoaAoHUEALw_wcB">R660 model</a>. Currently in limited testing, it's similar in concept to offerings from competitors such as DataDirect Networks, Weka and Lustre. Companies will upgrade for AI workloads, CEO Michael Dell said during a media and analyst Q&A following the keynote. The competitive benefits of AI will make a hardware refresh inevitable for many organizations, Dell said, although upgrades might not happen overnight for many customers, especially as <a href="https://www.techtarget.com/searchcio/news/366618727/US-tariffs-may-stymy-executives-product-decisions">external</a> <a href="https://www.techtarget.com/searchenterprisedesktop/tip/When-is-Windows-10-end-of-life-How-to-extend-support">factors</a> may complicate purchasing decisions. "My general view is that the importance of this technology is greater than all those problems,” he said. “I heard somebody say recently, <a href="https://www.techtarget.com/whatis/definition/context-window">tokens</a> are bigger than tariffs. That would sort of summarize our view of it.” Even when backed by high-end hardware or software, AI initiatives shouldn't be implemented haphazardly, Clarke said, echoing the sentiments of organizations already using AI in production. Instead, customers should start small, find what works and scale quickly to business needs to avoid wasting time, hardware and resources, he said. "If you apply AI to a shitty process, you just get a shitty answer faster," Clarke said. <section class="section main-article-chapter" data-menu-title="AI implementations"> <h2 class="section-title">AI implementations</h2> Dauntless XR, a Texas-based company specializing in augmented and virtual reality software, is using AI for object recognition and creating digital twins of atmospheres. <div class="imagecaption alignLeft"> <img src="https://cdn.ttgtmedia.com/rms/onlineimages/elliot_lori-lee.jpg" alt="Lori-Lee Elliott, co-founder and CEO, Dauntless X">Lori-Lee Elliott </div> Lori-Lee Elliott, co-founder and CEO of Dauntless XR, said these implementations are also complemented by a more traditional use with content generation using company data. For her, AI use needs to move beyond the chat room visualization and into uses that directly support an organization's business and workers seamlessly. "When a lot of people think of AI, they think of LLMs," Elliott said. "The end state of [our AI initiatives] -- and we're not quite there yet -- [is Dauntless XR having] a single platform where all of this data lives eventually." Recent AI marketing hype around Generative AI has buried the decades of prior research and real-world uses of AI, said Chris Sullivan, director of research and academic computing at the Oregon State University College of Earth, Ocean, and Atmospheric Sciences. His department's research into AI involves demanding <a href="https://www.techtarget.com/searchdatacenter/definition/high-performance-computing-HPC">high-performance computing (HPC)</a> uses, such as mapping ocean wildlife changes over centuries, photographic research and monitoring of microscopic krill and geographic information systems. The underlying mathematics and science behind today's AI have existed for decades, but recent advances in connectivity and memory have made the implementation far more feasible and usable to the public. <div class="imagecaption alignRight"> <img src="https://cdn.ttgtmedia.com/rms/onlineimages/sullivan_chris.jpg" alt="Chris Sullivan, Oregon State University College of Earth, Ocean, and Atmospheric Sciences">Chris Sullivan </div> "The number of times I've heard someone say AI and have no clue about what they're doing is stunning," Sullivan said. "People don't understand the term AI and misuse it so badly that it's become an umbrella that's flipped upside down and catches everything." The actual bottleneck most organizations will encounter in their AI ambitions comes down to data, he said. Many organizations treat data as a black box and expect off-the-shelf LLMs to help make sense of the digital detritus that has built up over the years. Many organizations might not even realize their data is ultimately lacking for AI uses, Sullivan said. "What we really need to do is start the AI down the pathway of data management and put agents in that place to help us understand what's there," he said. "We need [AI data] to be presented in a way that people can leverage it." </section> <section class="section main-article-chapter" data-menu-title="AI for the future"> <h2 class="section-title">AI for the future</h2> AI deployments aren't just the domain of massive organizations, but they could also prove useful for local charities. A hardware and software donation from Dell Technologies has brought virtual job interview training to Hopeworks, an American nonprofit serving economically disadvantaged individuals. <div class="imagecaption alignLeft"> <img src="https://cdn.ttgtmedia.com/rms/onlineimages/rhoton_dan.jpg " alt="Dan Rhoton, CEO, Hopeworks">Dan Rhoton </div> Hopeworks provides <a href="https://www.techtarget.com/searchcloudcomputing/tip/How-to-use-AI-to-bridge-the-cloud-skills-gap">technology training</a> and jobs programs for young adults and other needy individuals in the communities of Camden, N.J. and Philadelphia, said Dan Rhoton, CEO of Hopeworks. Clients are offered training in modern technology jobs such as web development or data analysis, alongside other career development skills like interviews, resume building and networking. Dell donated PowerEdge servers and other associated hardware, alongside the technology stack of the Dell AI Factory, to create job interview simulations using Dell Digital Assistants, Rhoton said. Clients can provide their resumes and job descriptions to more accurately calibrate the interviews they will face in the future, he said. "They're often interviewing for jobs where they can't get any advice," Rhoton said. "[With AI] they've got a partner that's not going to judge them and they can go into that next meeting confident and ready to contribute." The use of <a href="https://www.techtarget.com/searchapparchitecture/tip/366539577/The-past-present-and-future-of-AI-coding-tools">copilots and other AI tools</a> has been a boon for HopeWorks clients, enabling them to quickly learn new coding skills and see completed examples in a matter of days over months, he said. "AI is the best thing to happen to our young adults for job prospects," Rhoton said. "Back in the day, they were well-trained, but they were slow since they were new and inexperienced. What we're seeing is AI is not taking anyone's job, but they're able to do more in partnership [with AI]." Editor's note: This story was updated to clarify some of Michael Dell's comments. <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/1U83yhGY_pI?autoplay=0&modestbranding=1&rel=0&widget_referrer=null&enablejsapi=1&origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> Tim McCarthy is a news writer for Informa TechTarget covering cloud and data storage. </section> Enterprises and nonprofits at Dell Tech World 2025 are already deploying advanced AI, as Dell Technologies readies a new wave of products to meet future demand. https://cdn.ttgtmedia.com/rms/onlineimages/ai_a252657224.jpg https://www.techtarget.com/searchdatacenter/news/366624166/Dell-Technologies-customers-building-real-world-AI-apps Thu, 22 May 2025 11:30:00 GMT Dell Technologies customers building real-world AI apps Search Data Center Resources and Information from TechTarget 60 webmaster@techtarget.com