Get ONMINE

Stay Ahead, Stay ONMINE

Three takeaways about AI’s energy use and climate impacts

This week, we published Power Hungry, a package all about AI and energy. At the center of this package is the most comprehensive look yet at AI’s growing power demand, if I do say so myself.  This data-heavy story is the result of over six months of reporting by me and my colleague James O’Donnell (and the work of many others on our team). Over that time, with the help of leading researchers, we quantified the energy and emissions impacts of individual queries to AI models and tallied what it all adds up to, both right now and for the years ahead.  There’s a lot of data to dig through, and I hope you’ll take the time to explore the whole story. But in the meantime, here are three of my biggest takeaways from working on this project.  1. The energy demands of AI are anything but constant.  If you’ve heard estimates of AI’s toll, it’s probably a single number associated with a query, likely to OpenAI’s ChatGPT. One popular estimate is that writing an email with ChatGPT uses 500 milliliters (or roughly a bottle) of water. But as we started reporting, I was surprised to learn just how much the details of a query can affect its energy demand. No two queries are the same—for several reasons, including their complexity and the particulars of the model being queried. One key caveat here is that we don’t know much about “closed source” models—for these, companies hold back the details of how they work. (OpenAI’s ChatGPT and Google’s Gemini are examples.) Instead, we worked with researchers who measured the energy it takes to run open-source AI models, for which the source code is publicly available.  But using open-source models, it’s possible to directly measure the energy used to respond to a query rather than just guess. We worked with researchers who generated text, images, and video and measured the energy required for the chips the models are based on to perform the task.   Even just within the text responses, there was a pretty large range of energy needs. A complicated travel itinerary consumed nearly 10 times as much energy as a simple request for a few jokes, for example. An even bigger difference comes from the size of the model used. Larger models with more parameters used up to 70 times more energy than smaller ones for the same prompts.  As you might imagine, there’s also a big difference between text, images, or video. Videos generally took hundreds of times more energy to generate than text responses.  2. What’s powering the grid will greatly affect the climate toll of AI’s energy use.  As the resident climate reporter on this project, I was excited to take the expected energy toll and translate it into an expected emissions burden.  Powering a data center with a nuclear reactor or a whole bunch of solar panels and batteries will not affect our planet the same way as burning mountains of coal. To quantify this idea, we used a figure called carbon intensity, a measure of how dirty a unit of electricity is on a given grid.  We found that the same exact query, with the same exact energy demand, will have a very different climate impact depending on what the data center is powered by, and that depends on the location and the time of day. For example, querying a data center in West Virginia could cause nearly twice the emissions of querying one in California, according to calculations based on average data from 2024. This point shows why it matters where tech giants are building data centers, what the grid looks like in their chosen locations, and how that might change with more demand from the new infrastructure.  3. There is still so much that we don’t know when it comes to AI and energy.  Our reporting resulted in estimates that are some of the most specific and comprehensive out there. But ultimately, we still have no idea what many of the biggest, most influential models are adding up to in terms of energy and emissions. None of the companies we reached out to were willing to provide numbers during our reporting. Not one. Adding up our estimates can only go so far, in part because AI is increasingly everywhere. While today you might generally have to go to a dedicated site and type in questions, in the future AI could be stitched into the fabric of our interactions with technology. (See my colleague Will Douglas Heaven’s new story on Google’s I/O showcase: “By putting AI into everything, Google wants to make it invisible.”) AI could be one of the major forces that shape our society, our work, and our power grid. Knowing more about its consequences could be crucial to planning our future.  To dig into our reporting, give the main story a read. And if you’re looking for more details on how we came up with our numbers, you can check out this behind-the-scenes piece. There are also some great related stories in this package, including one from James Temple on the data center boom in the Nevada desert, one from David Rotman about how AI’s rise could entrench natural gas, and one from Will Douglas Heaven on a few technical innovations that could help make AI more efficient. Oh, and I also have a piece on why nuclear isn’t the easy answer some think it is.  Find them, and the rest of the stories in the package, here.  This article is from The Spark, MIT Technology Review’s weekly climate newsletter. To receive it in your inbox every Wednesday, sign up here.

This week, we published Power Hungry, a package all about AI and energy. At the center of this package is the most comprehensive look yet at AI’s growing power demand, if I do say so myself. 

This data-heavy story is the result of over six months of reporting by me and my colleague James O’Donnell (and the work of many others on our team). Over that time, with the help of leading researchers, we quantified the energy and emissions impacts of individual queries to AI models and tallied what it all adds up to, both right now and for the years ahead. 

There’s a lot of data to dig through, and I hope you’ll take the time to explore the whole story. But in the meantime, here are three of my biggest takeaways from working on this project. 

1. The energy demands of AI are anything but constant. 

If you’ve heard estimates of AI’s toll, it’s probably a single number associated with a query, likely to OpenAI’s ChatGPT. One popular estimate is that writing an email with ChatGPT uses 500 milliliters (or roughly a bottle) of water. But as we started reporting, I was surprised to learn just how much the details of a query can affect its energy demand. No two queries are the same—for several reasons, including their complexity and the particulars of the model being queried.

One key caveat here is that we don’t know much about “closed source” models—for these, companies hold back the details of how they work. (OpenAI’s ChatGPT and Google’s Gemini are examples.) Instead, we worked with researchers who measured the energy it takes to run open-source AI models, for which the source code is publicly available. 

But using open-source models, it’s possible to directly measure the energy used to respond to a query rather than just guess. We worked with researchers who generated text, images, and video and measured the energy required for the chips the models are based on to perform the task.  

Even just within the text responses, there was a pretty large range of energy needs. A complicated travel itinerary consumed nearly 10 times as much energy as a simple request for a few jokes, for example. An even bigger difference comes from the size of the model used. Larger models with more parameters used up to 70 times more energy than smaller ones for the same prompts. 

As you might imagine, there’s also a big difference between text, images, or video. Videos generally took hundreds of times more energy to generate than text responses. 

2. What’s powering the grid will greatly affect the climate toll of AI’s energy use. 

As the resident climate reporter on this project, I was excited to take the expected energy toll and translate it into an expected emissions burden. 

Powering a data center with a nuclear reactor or a whole bunch of solar panels and batteries will not affect our planet the same way as burning mountains of coal. To quantify this idea, we used a figure called carbon intensity, a measure of how dirty a unit of electricity is on a given grid. 

We found that the same exact query, with the same exact energy demand, will have a very different climate impact depending on what the data center is powered by, and that depends on the location and the time of day. For example, querying a data center in West Virginia could cause nearly twice the emissions of querying one in California, according to calculations based on average data from 2024.

This point shows why it matters where tech giants are building data centers, what the grid looks like in their chosen locations, and how that might change with more demand from the new infrastructure. 

3. There is still so much that we don’t know when it comes to AI and energy. 

Our reporting resulted in estimates that are some of the most specific and comprehensive out there. But ultimately, we still have no idea what many of the biggest, most influential models are adding up to in terms of energy and emissions. None of the companies we reached out to were willing to provide numbers during our reporting. Not one.

Adding up our estimates can only go so far, in part because AI is increasingly everywhere. While today you might generally have to go to a dedicated site and type in questions, in the future AI could be stitched into the fabric of our interactions with technology. (See my colleague Will Douglas Heaven’s new story on Google’s I/O showcase: “By putting AI into everything, Google wants to make it invisible.”)

AI could be one of the major forces that shape our society, our work, and our power grid. Knowing more about its consequences could be crucial to planning our future. 

To dig into our reporting, give the main story a read. And if you’re looking for more details on how we came up with our numbers, you can check out this behind-the-scenes piece.

There are also some great related stories in this package, including one from James Temple on the data center boom in the Nevada desert, one from David Rotman about how AI’s rise could entrench natural gas, and one from Will Douglas Heaven on a few technical innovations that could help make AI more efficient. Oh, and I also have a piece on why nuclear isn’t the easy answer some think it is. 

Find them, and the rest of the stories in the package, here. 

This article is from The Spark, MIT Technology Review’s weekly climate newsletter. To receive it in your inbox every Wednesday, sign up here.

Shape
Shape
Stay Ahead

Explore More Insights

More Articles

Stay ahead with more perspectives on cutting-edge power, infrastructure, energy,  bitcoin and AI solutions. Explore these articles to uncover strategies and insights shaping the future of industries.

Shape

Two Elliott Nominees Set to Join Phillips 66 Board

Two names put forward by Elliott Investment Management LP and another two endorsed by Phillips 66 are expected to have won at the refiner’s directorial election during its annual meeting of shareholders on Wednesday. “Based on the preliminary results, the elected Phillips 66 directors are expected to be Robert W.

Read More »

USA Crude Oil Inventories Rise Week on Week

U.S. commercial crude oil inventories, excluding those in the Strategic Petroleum Reserve (SPR), increased by 1.3 million barrels from the week ending May 9 to the week ending May 16, the U.S. Energy Information Administration (EIA) highlighted in its latest weekly petroleum status report. That EIA report was released on May 21 and included data for the week ending May 16. It showed that crude oil stocks, not including the SPR, stood at 443.2 million barrels on May 16, 441.8 million barrels on May 9, and 458.8 million barrels on May 17, 2024. The EIA report highlighted that data may not add up to totals due to independent rounding. Crude oil in the SPR stood at 400.5 million barrels on May 16, 399.7 million barrels on May 9, and 368.8 million barrels on May 17, 2024, the report outlined. Total petroleum stocks – including crude oil, total motor gasoline, fuel ethanol, kerosene type jet fuel, distillate fuel oil, residual fuel oil, propane/propylene, and other oils – stood at 1.623 billion barrels on May 16, the report showed. Total petroleum stocks were up 5.8 million barrels week on week and up 4.3 million barrels year on year, the report revealed. “At 443.2 million barrels, U.S. crude oil inventories are about six percent below the five year average for this time of year,” the EIA stated in its latest weekly petroleum status report. “Total motor gasoline inventories increased by 0.8 million barrels from last week and are about two percent below the five year average for this time of year. Both finished gasoline inventories and blending components inventories increased last week,” it added. “Distillate fuel inventories increased by 0.6 million barrels last week and are about 16 percent below the five year average for this time of year. Propane/propylene inventories increased by 2.7

Read More »

NYC power bills going up this summer, ConEd warns

Dive Brief: Commercial and small business customers in New York City will see summer power bills rise by about 10% and 8%, respectively, Consolidated Edison warned Wednesday. Residential customers will see smaller increases, the utility said. The price increases are part of a national trend amid rising demand. “Retail electricity prices have increased faster than the rate of inflation since 2022, and we expect them to continue increasing through 2026,” the U.S. Energy Information Administration said May 14. Utilities are making large investments to meet growing demand for electricity. ConEdison has spent about $2.35 billion on its distribution system since last summer, the utility said, including new substation equipment, underground cables and transformers. Dive Insight: U.S. energy prices rapidly increased from 2020 to 2022 as a result of the global pandemic and supply chain interruptions tied to Russia’s invasion of Ukraine. Fuel prices have declined in recent years, but electricity price increases have continued unabated, EIA said. “We expect the nominal U.S. average electricity price to increase by 13% from 2022 to 2025,” EIA said. “Parts of the country with relatively high electricity prices may experience greater price increases than those with relatively low electricity prices.” The Pacific Coast has seen retail electricity prices rise 26% since 2022, the federal government’s energy data arm said. New England and mid-Atlantic prices have risen 19%. Optional Caption Retrieved from U.S. Energy Information Administration. Retail bills include the cost of generation, transmission and distribution, along with taxes and fees. ConEd has been investing billions in its distribution system, and also raised delivery charges in January under a rate plan the New York State Public Service Commission approved in 2023, the utility said. A typical residential customer in New York City can expect average monthly bills to be 2.7% higher from June to September, relative

Read More »

Nuclear industry ‘facing some headwinds’ amid Congressional budget fight: NEI CEO

Dive Brief: House Republicans’ proposed wind-down of clean energy tax credits and key Department of Energy programs would reverse the nuclear industry’s recent momentum and threaten the Trump administration’s “energy dominance” goals, Nuclear Energy Institute CEO Maria Korsnick said on Tuesday. In an address at NEI’s annual Nuclear Energy Policy Forum in Washington, D.C., Korsnick emphasized the need for further reforms at the Nuclear Regulatory Commission, federal support for an expanded U.S. nuclear supply chain and public-private workforce investments if the industry is to meet expected future power demand. “We’re facing some headwinds, and the actions we take today will define what the future looks like,” Korsnick said. “We will be out there pounding the pavement every step of the way.” Dive Insight: During her 25-minute address and a brief question-and-answer session afterward, Korsnick reiterated concerns raised by nuclear industry stakeholders in an April 30 letter to House and Senate leaders. NEI organized the campaign and collected more than 120 signatures from utilities, independent power producers, reactor developers, nuclear services firms and others. The letter urged Congress to preserve four Inflation Reduction Act tax credits that benefit the nuclear industry, including technology-neutral clean energy investment and production credits and a separate production tax credit for existing nuclear reactors. But today, the U.S. House of Representatives passed a sweeping budget package that terminated the technology-neutral credits for new and expanded nuclear reactor projects that begin construction after 2028, four years earlier than current law requires. The bill also imposed new domestic sourcing and ownership requirements that some experts say are broadly unworkable, though it’s unclear whether those will affect existing or planned nuclear power plants. The changes surprised some nuclear advocates, given the Trump administration’s stated support for nuclear energy and the president’s influence with the GOP House caucus.  “We aren’t

Read More »

“Whatever works for industry, works for Tees Valley” – Chris Rowell on hydrogen

Hydrogen will play a pivotal role in the Tees Valley’s plans to decarbonise local industry. Speaking to Energy Voice, Tees Valley Combined Authority head of net zero Chris Rowell said: “There have been people who’ve pushed a particular technology or ideology. We’ve always said: whatever works for industry, that creates jobs and decarbonises the UK’s manufacturing base, works for Tees Valley.” “The hydrogen picture actually works really well for us.” In its 2022 Net Zero Strategy, the region laid out its ambitions to create the UK’s first decarbonised heavy industrial cluster by 2040. Central to its strategy is building out 4 GW of hydrogen production by 2030 and developing a National Hydrogen Transport Hub. Among the projects in the area is EDF’s Tees Green Hydrogen project, which secured UK government funding as part of the net zero hydrogen fund (NZHF) and the first hydrogen allocation round (HAR1). The project will use renewable electricity from the Teesside offshore wind farm and a planned solar farm to power hydrogen electrolysers. In addition, the HAR2 auction saw MorGen Energy secure a spot for its Teesside Green Hydrogen project. “You’ve got ready-made use cases for hydrogen and we can already see hydrogen producers who’ve chosen the Tees Valley interacting with sustainable fuels projects and industrial use cases as well,” said Rowell. “And that’s before you look at where to send hydrogen and the rest of the UK.” By making the Tees Valley a national Hydrogen Transport Hub, the region will have a central role in driving the UK’s energy transition. “We find all of the UK’s industrial clusters and manufacturing centres say we would get more inward investment if we have cheaper energy,” Rowell noted. “We are exploring options in partnership with Northern Gas Networks through the East Coast Hydrogen Initiative to look at

Read More »

Oil and gas workers face ‘unjust transition’ with no plan, commission warns

Scotland is on course for an “unjust transition”, a report has found, with the government accused of having no plan for oil and gas workers. The Just Transition Commission said urgent action is needed to ensure the transition from fossil fuels to renewables happens fairly. “Without urgent and ambitious action, investment and government leadership, Scotland’s offshore transition will not take place fairly, with harmful effects on workers, communities, employers and the regional economy of the north east that could otherwise be avoided,” the report said. The independent advisory body warned an unjust transition is possible despite it being known for decades that the North Sea oil and gas sector would decline. Oil and gas workers in Aberdeen told the commission they fear a “cliff edge” for their livelihoods. The report said: “In the context of global economic volatility, the pace and sequencing of the transition will be unjust if determined mainly by turbulent commodity prices. “The fragmented nature of both the fossil fuel and renewables industries makes effective planning more challenging, but also more critical. “To avoid harms to workers and communities and support new industry, governments must now take a bold, innovative approach that maximises leverage to set standards, establish pathways, create jobs, and manage shocks.” The commission said more needs to be done to support jobs in the offshore renewable energy sector, including wind, decommissioning and green hydrogen – areas it said are expected to see “rapid” growth. The expert group said: “Renewables have a key role to play in delivering a just transition provided robust minimum standards are achieved across the industry for pay, conditions, health and safety regulation and union recognition.” It called for a “clear plan” to be developed for building up Scotland’s renewables supply chain that could help mitigate the job losses seen in

Read More »

Galveston LNG Bunker Port Gets Remaining Construction Permits

Pilot LNG LLC and Seapath Group have received the green light from the United States Army Corps of Engineers (USACE) and Coast Guard for the Galveston LNG Bunker Port (GLBP), meaning the project in the Texas City industrial area is now fully permitted for construction. Galveston LNG Bunker Port LLC expects to make an FID (final investment decision) mid-2025. Start-up is expected in the second half of 2027. “After several years of challenging and complex work bringing together the engineering, permitting, and third-party supplies for gas and power to the project, we are now comfortably ahead in the marketplace to be the first dedicated LNG marine fuels supplier in the U.S. Gulf”, Seapath president Josh Lubarsky said in an online statement. “We have made a significant financial commitment to this project and, over the course of the last several years, have positioned GLBP to be the foremost clean fuel supply hub in the Galveston Bay/Gulf region”, Lubarsky said. Lubarsky added construction could start “in the coming months”. Located on Shoal Point in Galveston County, Texas, the project will serve the greater Houston-Galveston port complex. It is being undertaken in two phases with a total capacity of 720,000 gallons per day. It will have two three-million-gallon storage tanks. GLBP said, “GLBP will supply LNG by fuel barge to the rapidly expanding fleet of LNG-fueled vessels in the greater Houston-Galveston region. It is optimally located to serve major ports, including Port Houston, the Port of Galveston and the Port of Texas City”. USACE issued its Section 408 and 404/10 authorizations under the Clean Water Act. The Coast Guard issued its Captain of the Port Letter of Recommendation for the project’s Waterway Suitability Assessment pursuant to Regulation 33 CFR 127, GLBP said. In January GLBP said it had received its Texas Railroad Commission

Read More »

AWS clamping down on cloud capacity swapping; here’s what IT buyers need to know

As of June 1, AWS will no longer allow sub-account transfers or new commitments to be pooled and reallocated across customers. Barrow says the shift is happening because AWS is investing billions in new data centers to meet demand from AI and hyperscale workloads. “That infrastructure requires long-term planning and capital discipline,” he said. Phil Brunkard, executive counselor at Info-Tech Research Group UK, emphasized that AWS isn’t killing RIs or SPs, “it’s just closing a loophole.” “This stops MSPs from bulk‑buying a giant commitment, carving it up across dozens of tenants, and effectively reselling discounted EC2 hours,” he said. “Basically, AWS just tilted the field toward direct negotiations and cleaner billing.” What IT buyers should do now For enterprises that sourced discounted cloud resources through a broker or value-added reseller (VAR), the arbitrage window shuts, Brunkard noted. Enterprises should expect a “modest price bump” on steady‑state workloads and a “brief scramble” to unwind pooled commitments.  If original discounts were broker‑sourced, “budget for a small uptick,” he said. On the other hand, companies that buy their own RIs or SPs, or negotiate volume deals through AWS’s Enterprise Discount Program (EDP), shouldn’t be impacted, he said. Nothing changes except that pricing is now baselined.

Read More »

DriveNets extends AI networking fabric with multi-site capabilities for distributed GPU clusters

“We use the same physical architecture as anyone with top of rack and then leaf and spine switch,” Dudy Cohen, vice president of product marketing at DriveNets, told Network World. “But what happens between our top of rack, which is the switch that connects NICs (network interface cards) into the servers and the rest of the network is not based on Clos Ethernet architecture, rather on a very specific cell-based protocol. [It’s] the same protocol, by the way, that is used in the backplane of the chassis.” Cohen explained that any data packet that comes into an ingress switch from the NIC is cut into evenly sized cells, sprayed across the entire fabric and then reassembled on the other side. This approach distinguishes DriveNets from other solutions that might require specialized components such as Nvidia BlueField DPUs (data processing units) at the endpoints. “The fabric links between the top of rack and the spine are perfectly load balanced,” he said. “We do not use any hashing mechanism… and this is why we can contain all the congestion avoidance within the fabric and do not need any external assistance.” Multi-site implementation for distributed GPU clusters The multi-site capability allows organizations to overcome power constraints in a single data center by spreading GPU clusters across locations. This isn’t designed as a backup or failover mechanism. Lasser-Raab emphasized that it’s a single cluster in two locations that are up to 80 kilometers apart, which allows for connection to different power grids. The physical implementation typically uses high-bandwidth connections between sites. Cohen explained that there is either dark fiber or some DWDM (Dense Wavelength Division Multiplexing) fibre optic connectivity between the sites. Typically the connections are bundles of four 800 gigabit ethernet, acting as a single 3.2 terabit per second connection.

Read More »

Intel eyes exit from NEX unit as focus shifts to core chip business

“That’s something we’re going to expand and build on,” Tan said, according to the report, pointing to Intel’s commanding 68% share of the PC chip market and 55% share in data centers. By contrast, the NEX unit — responsible for silicon and software that power telecom gear, 5G infrastructure, and edge computing — has struggled to deliver the kind of strategic advantage Intel needs. According to the report, Tan and his team view it as non-essential to Intel’s turnaround plans. The report described the telecom side of the business as increasingly disconnected from Intel’s long-term objectives, while also pointing to fierce competition from companies like Broadcom that dominate key portions of the networking silicon market and leave little room for Intel to gain a meaningful share. Financial weight, strategic doubts Despite generating $5.8 billion in revenue in 2024, the NEX business was folded into Intel’s broader Data Center and Client Computing groups earlier this year. The move was seen internally as a signal that NEX had lost its independent strategic relevance and also reflects Tan’s ruthless prioritization.  To some in the industry, the review comes as little surprise. Over the past year, Intel has already shed non-core assets. In April, it sold a majority stake in Altera, its FPGA business, to private equity firm Silver Lake for $4.46 billion, shelving earlier plans for a public listing. This followed the 2022 spinoff of Mobileye, its autonomous driving arm. With a $19 billion loss in 2024 and revenue falling to $53.1 billion, the chipmaker also aims to streamline management, cut $10 billion in costs, and bet on AI chips and foundry services, competing with Nvidia, AMD, and TSMC.

Read More »

Tariff uncertainty weighs on networking vendors

“Our guide assumes current tariffs and exemptions remain in place through the quarter. These include the following: China at 30%, partially offset by an exemption for semiconductors and certain electronic components; Mexico and Canada at 25% for the components and products that are not eligible for the current exemptions,” Cisco CFO Scott Herron told Wall Street analysts in the company’s quarterly earnings report on May 14. At this time, Cisco expects little impact from tariffs on steel and aluminum and retaliatory tariffs, Herron said. “We’ll continue to leverage our world-class supply chain team to help mitigate the impact,” he said, adding that “the flexibility and agility we have built into our operations over the last few years, the size and scale of our supply chain, provides us some unique advantages as we support our customers globally.” “Once the tariff scenario stabilizes, there [are] steps that we can take to mitigate it, as you’ve seen us do with China from the first Trump administration. And only after that would we consider price [increases],” Herron said. Similarly, Extreme Networks noted the changing tariff conditions during its earnings call on April 30. “The tariff situation is very dynamic, I think, as everybody knows and can appreciate, and it’s kind of hard to call. Yes, there was concern initially given the magnitude of tariffs,” said Extreme Networks CEO Ed Meyercord on the earnings call. “The larger question is, will all of the changes globally in trade and tariff policy have an impact on demand? And that’s hard to call at this point. And we’re going to hold as far as providing guidance or judgment on that until we have finality come July.” Financial news Meanwhile, AI is fueling high expectations and influencing investments in enterprise campus and data center environments.

Read More »

Liquid cooling becoming essential as AI servers proliferate

“Facility water loops sometimes have good water quality, sometimes bad,” says My Troung, CTO at ZutaCore, a liquid cooling company. “Sometimes you have organics you don’t want to have inside the technical loop.” So there’s one set of pipes that goes around the data center, collecting the heat from the server racks, and another set of smaller pipes that lives inside individual racks or servers. “That inner loop is some sort of technical fluid, and the two loops exchange heat across a heat exchanger,” says Troung. The most common approach today, he says, is to use a single-phase liquid — one that stays in liquid form and never evaporates into a gas — such as water or propylene glycol. But it’s not the most efficient option. Evaporation is a great way to dissipate heat. That’s what our bodies do when we sweat. When water goes from a liquid to a gas it’s called a phase change, and it uses up energy and makes everything around it slightly cooler. Of course, few servers run hot enough to boil water — but they can boil other liquids. “Two phase is the most efficient cooling technology,” says Xianming (Simon) Dai, a professor at University of Texas at Dallas. And it might be here sooner than you think. In a keynote address in March at Nvidia GTC, Nvidia CEO Jensen Huang unveiled the Rubin Ultra NVL576, due in the second half of 2027 — with 600 kilowatts per rack. “With the 600 kilowatt racks that Nvidia is announcing, the industry will have to shift very soon from single-phase approaches to two-phase,” says ZutaCore’s Troung. Another highly-efficient cooling approach is immersion cooling. According to a Castrol survey released in March, 90% of 600 data center industry leaders say that they are considering switching to immersion

Read More »

Cisco taps OpenAI’s Codex for AI-driven network coding

“If you want to ask Codex a question about your codebase, click “Ask”. Each task is processed independently in a separate, isolated environment preloaded with your codebase. Codex can read and edit files, as well as run commands including test harnesses, linters, and type checkers. Task completion typically takes between 1 and 30 minutes, depending on complexity, and you can monitor Codex’s progress in real time,” according to OpenAI. “Once Codex completes a task, it commits its changes in its environment. Codex provides verifiable evidence of its actions through citations of terminal logs and test outputs, allowing you to trace each step taken during task completion,” OpenAI wrote. “You can then review the results, request further revisions, open a GitHub pull request, or directly integrate the changes into your local environment. In the product, you can configure the Codex environment to match your real development environment as closely as possible.” OpenAI is releasing Codex as a research preview: “We prioritized security and transparency when designing Codex so users can verify its outputs – a safeguard that grows increasingly more important as AI models handle more complex coding tasks independently and safety considerations evolve. Users can check Codex’s work through citations, terminal logs and test results,” OpenAI wrote.  Internally, technical teams at OpenAI have started using Codex. “It is most often used by OpenAI engineers to offload repetitive, well-scoped tasks, like refactoring, renaming, and writing tests, that would otherwise break focus. It’s equally useful for scaffolding new features, wiring components, fixing bugs, and drafting documentation,” OpenAI stated. Cisco’s view of agentic AI Patel stated that Codex is part of the developing AI agent world, where Cisco envisions billions of AI agents will work together to transform and redefine the architectural assumptions the industry has relied on. Agents will communicate within and

Read More »

Microsoft will invest $80B in AI data centers in fiscal 2025

And Microsoft isn’t the only one that is ramping up its investments into AI-enabled data centers. Rival cloud service providers are all investing in either upgrading or opening new data centers to capture a larger chunk of business from developers and users of large language models (LLMs).  In a report published in October 2024, Bloomberg Intelligence estimated that demand for generative AI would push Microsoft, AWS, Google, Oracle, Meta, and Apple would between them devote $200 billion to capex in 2025, up from $110 billion in 2023. Microsoft is one of the biggest spenders, followed closely by Google and AWS, Bloomberg Intelligence said. Its estimate of Microsoft’s capital spending on AI, at $62.4 billion for calendar 2025, is lower than Smith’s claim that the company will invest $80 billion in the fiscal year to June 30, 2025. Both figures, though, are way higher than Microsoft’s 2020 capital expenditure of “just” $17.6 billion. The majority of the increased spending is tied to cloud services and the expansion of AI infrastructure needed to provide compute capacity for OpenAI workloads. Separately, last October Amazon CEO Andy Jassy said his company planned total capex spend of $75 billion in 2024 and even more in 2025, with much of it going to AWS, its cloud computing division.

Read More »

John Deere unveils more autonomous farm machines to address skill labor shortage

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Self-driving tractors might be the path to self-driving cars. John Deere has revealed a new line of autonomous machines and tech across agriculture, construction and commercial landscaping. The Moline, Illinois-based John Deere has been in business for 187 years, yet it’s been a regular as a non-tech company showing off technology at the big tech trade show in Las Vegas and is back at CES 2025 with more autonomous tractors and other vehicles. This is not something we usually cover, but John Deere has a lot of data that is interesting in the big picture of tech. The message from the company is that there aren’t enough skilled farm laborers to do the work that its customers need. It’s been a challenge for most of the last two decades, said Jahmy Hindman, CTO at John Deere, in a briefing. Much of the tech will come this fall and after that. He noted that the average farmer in the U.S. is over 58 and works 12 to 18 hours a day to grow food for us. And he said the American Farm Bureau Federation estimates there are roughly 2.4 million farm jobs that need to be filled annually; and the agricultural work force continues to shrink. (This is my hint to the anti-immigration crowd). John Deere’s autonomous 9RX Tractor. Farmers can oversee it using an app. While each of these industries experiences their own set of challenges, a commonality across all is skilled labor availability. In construction, about 80% percent of contractors struggle to find skilled labor. And in commercial landscaping, 86% of landscaping business owners can’t find labor to fill open positions, he said. “They have to figure out how to do

Read More »

2025 playbook for enterprise AI success, from agents to evals

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More 2025 is poised to be a pivotal year for enterprise AI. The past year has seen rapid innovation, and this year will see the same. This has made it more critical than ever to revisit your AI strategy to stay competitive and create value for your customers. From scaling AI agents to optimizing costs, here are the five critical areas enterprises should prioritize for their AI strategy this year. 1. Agents: the next generation of automation AI agents are no longer theoretical. In 2025, they’re indispensable tools for enterprises looking to streamline operations and enhance customer interactions. Unlike traditional software, agents powered by large language models (LLMs) can make nuanced decisions, navigate complex multi-step tasks, and integrate seamlessly with tools and APIs. At the start of 2024, agents were not ready for prime time, making frustrating mistakes like hallucinating URLs. They started getting better as frontier large language models themselves improved. “Let me put it this way,” said Sam Witteveen, cofounder of Red Dragon, a company that develops agents for companies, and that recently reviewed the 48 agents it built last year. “Interestingly, the ones that we built at the start of the year, a lot of those worked way better at the end of the year just because the models got better.” Witteveen shared this in the video podcast we filmed to discuss these five big trends in detail. Models are getting better and hallucinating less, and they’re also being trained to do agentic tasks. Another feature that the model providers are researching is a way to use the LLM as a judge, and as models get cheaper (something we’ll cover below), companies can use three or more models to

Read More »

OpenAI’s red teaming innovations define new essentials for security leaders in the AI era

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More OpenAI has taken a more aggressive approach to red teaming than its AI competitors, demonstrating its security teams’ advanced capabilities in two areas: multi-step reinforcement and external red teaming. OpenAI recently released two papers that set a new competitive standard for improving the quality, reliability and safety of AI models in these two techniques and more. The first paper, “OpenAI’s Approach to External Red Teaming for AI Models and Systems,” reports that specialized teams outside the company have proven effective in uncovering vulnerabilities that might otherwise have made it into a released model because in-house testing techniques may have missed them. In the second paper, “Diverse and Effective Red Teaming with Auto-Generated Rewards and Multi-Step Reinforcement Learning,” OpenAI introduces an automated framework that relies on iterative reinforcement learning to generate a broad spectrum of novel, wide-ranging attacks. Going all-in on red teaming pays practical, competitive dividends It’s encouraging to see competitive intensity in red teaming growing among AI companies. When Anthropic released its AI red team guidelines in June of last year, it joined AI providers including Google, Microsoft, Nvidia, OpenAI, and even the U.S.’s National Institute of Standards and Technology (NIST), which all had released red teaming frameworks. Investing heavily in red teaming yields tangible benefits for security leaders in any organization. OpenAI’s paper on external red teaming provides a detailed analysis of how the company strives to create specialized external teams that include cybersecurity and subject matter experts. The goal is to see if knowledgeable external teams can defeat models’ security perimeters and find gaps in their security, biases and controls that prompt-based testing couldn’t find. What makes OpenAI’s recent papers noteworthy is how well they define using human-in-the-middle

Read More »
More Articles