Latest news with #dataCenters
CNET
23-07-2025
- Business
- CNET
Trump's AI Action Plan Is Here. These Are the Top 5 Themes
The Trump administration on Wednesday laid out the steps it plans to take to ensure "global AI dominance" for the US, with an AI Action Plan that calls for cutting regulations to speed up the development of artificial intelligence tools and the infrastructure to power them. Critics said the plan is a handout to tech and fossil fuel companies, slashing rules that could protect consumers, prevent pollution and fight climate change. The plan itself isn't all that binding. It includes dozens of policy recommendations for the executive branch to carry out but doesn't on its own make anything happen. The steps it lays out follow how the Trump administration has approached AI and technology over the past six months -- giving tech companies a largely free hand, focusing on beating China and prioritizing the construction of data centers, factories and fossil fuel power plants over environmental regulations. It is seizing on the moment created by the arrival of ChatGPT less than three years ago and the ensuing wave of generative AI efforts by Google, Meta and others. "An industrial revolution, an information revolution, and a renaissance -- all at once. This is the potential that AI presents," the AI Action Plan says. President Donald Trump is expected to speak on AI priorities later Wednesday during a forum with the hosts of the All-In Podcast. The administration and tech industry groups touted the plan as a framework for US success in a race against China. "President Trump's AI Action Plan presents a blueprint to usher in a new era of US AI dominance," Jason Oxman, president and CEO of the tech industry trade group ITI, said in a statement. Consumer groups said the plan focuses on deregulation and would hurt consumers by reducing the rules that could protect them. "Whether it's promoting the use of federal land for dirty data centers, giving the FTC orders to question past cases, or attempting to revive some version of the soundly defeated AI moratorium by tying federal funds to not having 'onerous regulation' according to the FCC, this is an unwelcome distraction at a critical time for government to get consumer protection right with increasing AI use and abuse," Ben Winters, director of AI and privacy at the Consumer Federation of America, said in a statement. Here's a look at the proposals in the plan. Slashing regulations for AI infrastructure The plan says AI growth will require infrastructure, including chip factories, data centers and more energy generation. And it blames environmental regulations for getting in the way. In response, it proposes exemptions for AI-related construction from certain environmental regulations, including those aimed at protecting clean water and air. It also suggests making federal lands available for data center construction and related power plants. To provide energy for all those data centers, the plan calls for steps to prevent the "premature decommissioning of critical power generation resources." This likely refers to keeping coal-fired power plants and other mostly fossil-fuel-driven infrastructure online for longer. The administration also called to prioritize the connection of new "reliable, dispatchable power sources" to the grid and specifically named nuclear fission and fusion and advanced geothermal generation. Earlier this month, the president signed a bill that would end many tax credits and incentives for renewable energy -- wind and solar -- years earlier than planned. Wind and solar make up the bulk of the new energy generation being added to the US grid right now. Fewer rules around AI technology Congress ended up not including a moratorium on state AI rules in the recently passed tax and spending bill but efforts to cut regulations around AI continue from the executive branch in the action plan. "AI is far too important to smother in bureaucracy at this early stage, whether at the state or Federal level," the plan says. The plan recommends that several federal agencies review whether existing or proposed rules would interfere with the development and deployment of AI. The feds would consider whether states' regulatory climate is favorable for AI when deciding to award funding. Federal Trade Commission investigations and orders would be reviewed to determine that they don't "advance theories of liability that unduly burden AI innovation." Those rule changes could undermine efforts to protect consumers from problems caused by AI, critics said. "Companies -- including AI companies -- have a legal obligation to protect their products from being used for harm," Justin Brookman, director of tech policy at Consumer Reports, said in a statement. "When a company makes design choices that increase the risk their product will be used for harm, or when the risks are particularly serious, companies should bear legal responsibility." Ideology and large language models The plan proposes some steps around ensuring AI "protects free speech and American values," further steps in the Trump administration's efforts to roll back federal policies around what it refers to as "diversity, equity and inclusion," along with references to the problems of misinformation and climate change. It calls for eliminating references to those items in the National Institute of Standards and Technology's AI Risk Management Framework. Federal agencies would only be allowed to contract with AI developers who "ensure that their systems are objective and free from top-down ideological bias." The Trump administration has recently announced contracts of up to $200 million each to developers Anthropic, Google, OpenAI and xAI. Grok, the model from Elon Musk's xAI, has recently come under fire for spouting antisemitism and hate speech. Dealing with workforce challenges The plan acknowledges that AI will "transform how work gets done across all industries and occupations, demanding a serious workforce response to help workers navigate that transition" and recommends actions by federal agencies including the Department of Labor intended to mitigate the harms of AI-driven job displacement. The plan calls for the Bureau of Labor Statistics, Census Bureau and Bureau of Economic Analysis to monitor how AI affects the labor market using data already collected. An AI Workforce Research Hub under the Department of Labor would lead monitoring and issue policy recommendations. Most of the actual plans to help workers displaced by AI involve retraining those workers for other jobs or to help states do the same. Other jobs-related recommendations are aimed at boosting the kinds of jobs needed for all those data centers and chip manufacturing plants -- like electricians and HVAC technicians. These plans and others to encourage AI literacy and AI use in education drew praise from the Software & Information Industry Association, a tech industry trade group. "These are key components for building trust and ensuring all communities can participate in and benefit from AI's potential," Paul Lekas, SIIA's senior vice president of global public policy, said in a statement. More AI in government The plan envisions more use of AI by the federal government. A talent exchange program would allow employees with experience or talent in AI to be detailed to other agencies in need. The General Services Administration would create a toolbox of AI models that would help agencies see models to choose from and use cases in other parts of the government. Every government agency would also be required to ensure employees who could use AI in their jobs have access to and training for AI tools. Many recommendations focus specifically on the Department of Defense, including creating a virtual proving ground for AI and autonomous systems. AI companies have already been signing contracts with the DOD to develop AI tools for the military.
Yahoo
23-07-2025
- Business
- Yahoo
Better Cloud AI Stock: CoreWeave vs. DigitalOcean
Key Points CoreWeave's transformation from a crypto miner to a cloud GPU leader is paying off. DigitalOcean is expanding its cloud platform at a slower and steadier rate. The hare might beat the tortoise this time. 10 stocks we like better than CoreWeave › CoreWeave (NASDAQ: CRWV) and DigitalOcean (NYSE: DOCN) both help companies process artificial (AI) tasks with their cloud-based graphics processing units (GPUs). CoreWeave, previously a cryptocurrency mining company, mainly serves larger companies. DigitalOcean splits its servers into "droplets" for smaller businesses and developers. Each should be in a good position to profit from the explosive growth of the AI market. However, investors are clearly more bullish on CoreWeave, which went public at $40 in March but now trades at around $125. DigitalOcean trades at $29, which is nearly 40% below its initial public offering price of $47 from March 2021. Let's see which is the better cloud AI stock. The differences between CoreWeave and DigitalOcean CoreWeave was once an Ethereum (CRYPTO: ETH) miner, but it abandoned that business model in 2018 and started using its GPUs to remotely process AI tasks. In 2022, it spent about $100 million to install Nvidia's (NASDAQ: NVDA) H100 GPUs in its data centers, and it used those GPUs as collateral to secure more funding to build additional data centers. It subsequently attracted investments from Nvidia, Cisco, and other tech giants. Today, CoreWeave operates 33 data centers across the U.S. and Europe -- up from just three centers at the end of 2022. Its top customers include Microsoft (NASDAQ: MSFT) and OpenAI. DigitalOcean's cloud infrastructure platform, which provides remote storage and computing power, is similar to Amazon Web Services and Microsoft Azure. But unlike those leading cloud platforms, which mainly serve large enterprise clients, DigitalOcean carves up its cloud servers into thinner and more affordable slices for smaller businesses. In 2023, it added cloud-based GPUs to its platform via its acquisition of Paperspace. DigitalOcean has been expanding much more slowly than CoreWeave: It currently operates 15 data centers across nine geographic regions, up from 14 centers at the end of 2022. Which company is growing faster? From 2022 to 2024, CoreWeave's annual revenue grew at a staggering compound annual growth rate (CAGR) of 990%, from $16 million to $1.9 billion. DigitalOcean's revenue rose at a more modest (but still respectable) CAGR of 16%, from $576 million in 2022 to $781 million in 2024. CoreWeave grew much faster than DigitalOcean for three reasons. First, it focused only on providing cloud-based GPUs for demanding AI tasks instead of a broader range of storage and computing services. DigitalOcean's acquisition of Paperspace gave it a foothold in the AI market, but its non-AI cloud services aren't growing as rapidly. Second, CoreWeave locked in huge customers, like Microsoft and OpenAI, that could afford to quickly ramp up their spending on its cloud-based GPU services. DigitalOcean served smaller developers and small-to-medium-size businesses -- which paid less money to deploy their apps and sandboxes. Third, CoreWeave has taken on lots of debt and racked up steep losses to buy more GPUs and open more data centers. DigitalOcean has been prioritizing its profit growth over its near-term expansion, and its net income has stayed in the black over the past two years. Which stock has more upside potential? From 2024 to 2027, analysts expect CoreWeave's revenue to grow at a CAGR of 106% to $16.7 billion as it turns profitable in the final year. They expect DigitalOcean's revenue to increase at a CAGR of 14% to $1.2 billion as its net income rises at a CAGR of 29% to $179 million. CoreWeave's projected growth trajectory looks incredible, but that expansion will likely be driven by a lot of debt and secondary offerings. Yet with a market cap of $63.5 billion, it doesn't seem that pricey relative to its growth potential at 13 times this year's sales. DigitalOcean, with a market cap of $2.7 billion, might seem a lot cheaper at 3 times this year's sales. But it's trading at that discount because it's growing at a much slower rate. Its conservative AI strategy also isn't attracting as much attention as CoreWeave's all-in expansion. So for now, CoreWeave still looks like a better play on the cloud and AI markets than DigitalOcean. Its business strategy is risky and aggressive, but it could generate much bigger long-term returns for its investors than DigitalOcean's less ambitious approach. Should you invest $1,000 in CoreWeave right now? Before you buy stock in CoreWeave, consider this: The Motley Fool Stock Advisor analyst team just identified what they believe are the for investors to buy now… and CoreWeave wasn't one of them. The 10 stocks that made the cut could produce monster returns in the coming years. Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you'd have $665,092!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you'd have $1,050,477!* Now, it's worth noting Stock Advisor's total average return is 1,055% — a market-crushing outperformance compared to 180% for the S&P 500. Don't miss out on the latest top 10 list, available when you join Stock Advisor. See the 10 stocks » *Stock Advisor returns as of July 21, 2025 John Mackey, former CEO of Whole Foods Market, an Amazon subsidiary, is a member of The Motley Fool's board of directors. Leo Sun has positions in Amazon. The Motley Fool has positions in and recommends Amazon, Cisco Systems, DigitalOcean, Ethereum, Microsoft, and Nvidia. The Motley Fool recommends the following options: long January 2026 $395 calls on Microsoft and short January 2026 $405 calls on Microsoft. The Motley Fool has a disclosure policy. Better Cloud AI Stock: CoreWeave vs. DigitalOcean was originally published by The Motley Fool
Arabian Business
15-07-2025
- Business
- Arabian Business
Meta to invest ‘hundreds of billions' of dollars in advanced AI
Meta Platforms plans to invest hundreds of billions of dollars in computing infrastructure to advance its superintelligence ambitions in artificial intelligence (AI), according to Chief Executive Mark Zuckerberg. Writing on social media platform Threads, Zuckerberg said Meta would build one of the most 'elite and talent-dense teams in the industry' and had the capital from its business to support the plans. 'We're also going to invest hundreds of billions of dollars into compute to build superintelligence,' he said. Meta investing hundreds of billions in AI Superintelligence refers to a hypothetical AI system that is capable of outperforming the human brain. Zuckerberg unveiled plans for massive data centres to support the effort. One facility, called Prometheus, is expected to go online in 2026. Another, Hyperion, could eventually consume up to 5 gigawatts of power – enough to supply electricity to more than four million average US households, according to experts.
Bloomberg
10-07-2025
- Business
- Bloomberg
Oracle Said to Move Ahead With Cloud Services Plan in Indonesia
Oracle Corp. will partner with DayOne Data Centers Singapore Pte to establish its first cloud services center in Indonesia, people familiar with the matter said, boosting its partnership with a key regional operator that counts TikTok owner ByteDance Ltd as its largest customer. The American tech giant will lease DayOne's data centers located at Nongsa Digital Park on the Indonesian island of Batam, according to the people, who asked not to be identified discussing information that's private. Oracle will be the sole tenant at DayOne plots that could support facilities with at least 120 megawatts of power, they said.
Yahoo
09-07-2025
- Business
- Yahoo
How to power the AI economy while supporting local communities
American leadership in artificial intelligence promises massive economic opportunities and transformation across industries: AI systems are poised to drive breakthroughs in drug discovery, enable critical infrastructure, and revolutionize sectors from healthcare to finance. To drive these shifts, AI needs a massive amount of energy. By 2028, AI data centers are expected to account for 12 percent of nationwide electricity demand. Data centers' median size will more than double over the next decade, according to a survey of US data center operators from power provider Bloom Energy. Some will require more than a gigawatt of power – enough for over 800,000 average American homes. Many Americans are concerned about how AI's hunger for energy will impact them. Communities around the country are pushing back against the building of new data centers nearby, despite their potential to create jobs and boost local economies. Chief among communities' concerns are rising energy costs and how new data centers could drive them up further. Growing pressure on the grid also increases the risk of power outages, and communities are alarmed about data centers' potential pollution, noise, and water and land use. Powering the AI economy without local communities bearing the brunt of higher energy bills, power outages, and pollution is a challenge, to be sure. But specific types of onsite power can create a 'yes, and' to fuel AI's growth and support communities. Onsite power means physically producing electricity at the location where it's used. For example, fuel cells are a type of onsite power that convert natural gas, hydrogen, or biogas into electricity without using combustion. Other power sources that data centers can deploy onsite include natural gas engines and turbines. When data centers use certain kinds of onsite power, it creates several advantages for both local communities and AI's growth. Lower impact on consumer energy costs. Some facilities use onsite power to go 'off-grid,' meaning they don't connect to the electricity grid – and therefore that their energy costs don't affect local residents' utility bills. Others use onsite power as a primary source while maintaining a connection to the grid. Utilities can also work directly with onsite power providers to produce energy for data centers in their service areas and 'fence off' that coverage, preventing data centers' costs from being passed on to consumers. Approaches like these help keep consumer utility bills stable while providing reliable power for data centers to fuel AI. Less pollution. Some onsite power technologies like fuel cells produce significantly lower emissions than traditional energy sources because they don't use combustion. Fuel cells also reduce smog-forming local air pollution by more than 99% compared to traditional energy generation technologies, and use a tiny fraction of the water that legacy combustion generators need. Smaller footprint. High power density—producing relatively large amounts of power in a given space—is a major benefit of some onsite power sources. Gas turbines and reciprocating engines provide up to 50 megawatts of power per acre, while fuel cells can deliver double that – up to 100 megawatts on less than an acre. This small footprint can help ease communities' concerns about data centers' visual impacts and the loss of open space. Quieter. Excess noise is a significant concern among many communities near data centers. Onsite power sources like fuel cells are much quieter than other energy sources, operating at ~70 dBA – about the sound level of a normal conversation. Interest is rising in onsite power generation for data centers: by 2030, nearly 40% expect to use some onsite power and 27% say it will provide all of their electricity. This will help data centers deliver on their promised local economic benefits—and AI's broad transformative potential—without the adverse effects of higher energy costs, power outages, and more pollution and noise in American communities. This story was produced by Bloom Energy and reviewed and distributed by Stacker.
