Power & Sites · United States

OpenAI makes new data center commitments, AI policy recommendations

Credited to SiliconANGLE · siliconangle.com · 16h ago

OpenAI Group PBC today made a set of commitments designed to ensure that its data centers benefit the communities where they’re built.

It also proposed that lawmakers make some of the commitments mandatory for artificial intelligence companies.

The move comes two days after Democrats won both the House and the Senate on a campaign backing data center construction curbs. Additionally, voters approved dozens of ballot measures designed to scale back or limit AI infrastructure projects.

OpenAI today reiterated a commitment it made in January to fully pay for the electricity and grid upgrades that its data centers require. That covers items such as the construction of new overhead power lines. Additionally, the company pledged to provide “financial assurances” that will protect other grid users from potential project cancellations and delays.

OpenAI will also take steps to reduce grid strain. In particular, it will scale back its data centers’ power draw at times when energy demand is particularly high. That approach, which is known as a flexible data center model, has also drawn the interest of other industry players as Nvidia Corp. and Google LLC.

Some data center builders sign non-disclosure agreements with local officials to keep construction details under wraps. OpenAI today pledged not to use NDAs. It will instead only seek “narrow protections covering legitimate security information, trade secrets, or proprietary or competitively sensitive information.”

OpenAI will publicly share construction plans and noise projections before construction of a new facility begins. Additionally, it will provide information about the jobs that each data center is expected to create. The goal is to help local unions and training providers prepare residents for the positions.

OpenAI plans to tailor its commitments to each data center. It will publish the customized project terms in a document it calls a local compact. Once construction is underway, the company will provide annual updates and designate a local contact to field residents’ questions.

OpenAI hopes to codify some of the best practices that it published today into law. The company released a policy paper that calls on lawmakers to limit data center developers’ use of NDAs and require them to cover their electricity costs. Furthermore, OpenAI argues that AI companies should be offered incentives to lower their power usage at times of peak grid stress.

Some of the best practices that the ChatGPT developer is prompting are already being adopted by other industry players. Google, for example, today pledged to fully cover the electricity costs of a new $17 billion data center that it plans to build in Wyoming. It will also finance related grid upgrades. Anthropic PBC made a similar commitment in February.

Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.

Are you an AWS customer? Support SiliconANGLE financially by buying your AWS services from our Marketplace portal page and links: https://siliconangle.com/aws-marketplace/

Founded by tech visionaries John Furrier and Dave Vellante, SiliconANGLE Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.

Cloudflare shares surge as revenue growth accelerates to 39% and outlook rises

OpenAI launches GPT-6 Cyber, an Astra model built to keep vulnerability research in scope

Atlassian shares fall 17% as slower subscription growth forecast overshadows earnings beat

Google Cloud storage moves to the front of the AI stack as enterprises hit the GPU utilization ceiling

Humanoid robot maker Agility Robotics debuts on the Nasdaq after SPAC merger

Rising demand puts Locaweb's data center strategy to the test

SECURITY - BY DUNCAN RILEY. 58 MINUTES AGO

EMERGING TECH - BY MARIA DEUTSCHER. 2 HOURS AGO

Join our community

Power & SitesUnited States

Microsoft tightens oversight of neocloud data center projects, Bloomberg News reports

Credited to Investing.com · investing.com22h ago

Useful

View the articleInvesting.com

Useful

View the articleInvesting.com

Front

Power & Sites · United States

University of Utah retains operational sovereignty as Redtail expands research access

Credited to SiliconANGLE · siliconangle.com · 22h ago

Operational sovereignty gives universities authority over how shared computing resources are allocated, managed and secured.

The University of Utah’s Redtail AI factory supercomputer brings advanced computing capabilities to a statewide community. Supporting researchers, students, administrators and external partners through that infrastructure requires more than computing capacity, according to Manish Parashar (pictured), chief AI officer of the University of Utah.

“In research, we are always trying to balance research security with doing open research for societal good,” he said. “Depending on where you are, you might want to lean in one direction or the other. Having that capability to turn the knob becomes so important.”

Parashar spoke with theCUBE Research’s Dave Vellante and Amit Eyal Govrin for the “Sovereign AI: From Strategy to Reality” video series, during an exclusive interview on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed university control of computing infrastructure, research governance and the investment needed to support Utah’s expanding ambitions. (* Disclosure below.)

Operational sovereignty starts with university expertise

Hewlett Packard Enterprise Co. helped architect and deploy Redtail as a turnkey system. The university’s Center for High Performance Computing operates it, drawing on expertise that also supports separate regulated research environments, Parashar explained.

“They have admin access, they’re doing the allocations, they are doing the onboarding,” he said. “They’re doing the management and monitoring … the optimization [and] … the right sizing.”

The interview examined Utah’s approach through theCUBE Research’s Five Pillars of Sovereign AI framework: Territorial, Legal, Technological, Operational and Financial. The university retains operational and legal responsibilities, while financing rests on partnerships with the state and donors, according to Parashar.

“Our overall ecosystem is a hybrid,” he said. “There’s things that are on-prem, there are things that are cloud-hosted, and we always will look at it as a hybrid system. Not everything has to be territorially in a single data center or in the state.”

Redtail also gives the university a way to accommodate growing demand for protected computing by freeing up capacity in existing facilities. Current environments support sensitive and regulated research, including healthcare projects, and expanding their capacity is a priority, Parashar noted.

“How do we move into a facility that has not only the capacity, but is also compliant with emerging controls that are coming out? We want to grow that out, and that’s our focus right now,” he said. “And what happens is with the Redtail, a lot of our open research moves to that and that frees up capacity in our data center and other places where we can grow out this regulated environment.”

Statewide ambitions require sustained investment

The university’s One-U Responsible AI initiative connects research with broader societal needs. Its work spans entrepreneurship, public-sector collaboration and regional challenges such as air quality, Parashar explained.

“Being open there becomes very, very important because you want to bring the best ideas, the best data — and most often, it is citizen data that’s coming in — and be able to then see how we can most effectively use AI to address this very regional challenge,” he said. “We don’t want to lose ideas; we don’t want to lose the data that’s coming in. We want to assimilate it and make sure the outcome and the warnings that come out of that — the notifications — are as broadly disseminated as possible.”

Financial sustainability must accommodate rising demand as well as technology refreshes. Expansion planning began before the Redtail contract was signed, and the university is approaching federal agencies, including the Department of Energy and the National Science Foundation, about investing in the existing ecosystem, Parashar noted.

“We are talking to the state and showing them value — very quickly — and convincing them that investing in this is good for the state [and] it’s good for its citizens,” he said. “We are [also] talking to additional philanthropic donors.”

The university’s statewide mission extends beyond providing access to computing resources. Training, workforce development and shared expertise are also part of the ecosystem that operational sovereignty must support, according to Parashar.

“It’s not that we need everything to be tightly controlled. We don’t want everything to be completely open,” he said. “We want to make sure we have the ability to right-size it depending on what we are working on.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the “Sovereign AI: From Strategy to Reality” video series:

(* Disclosure: TheCUBE is a paid media partner for the “Sovereign AI: From Strategy to Reality” interview series. Neither HPE, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Are you an AWS customer? Support SiliconANGLE financially by buying your AWS services from our Marketplace portal page and links: https://siliconangle.com/aws-marketplace/

Oracle wants to pull stalled AI agents out of demo mode and into production

Same model, better answer: Enterprise data is what turns generic AI into a business tool

What to expect during KubeCon + CloudNativeCon NA: Join theCUBE Nov. 10-12

Oracle's multicloud approach: Put the database wherever AI pipelines already run

Aina's smart AI ring promises a new way to interact with AI agents that anticipate user's needs

Box taps Nvidia's AI models to build a rich context layer for autonomous enterprise agents

AI - BY JONATHAN ANTHONY. 55 MINUTES AGO

BIG DATA - BY MIKE WHEATLEY. 4 HOURS AGO

Join our community