Latest news with #SemiAnalysis
CNBC
28-05-2025
- Business
- CNBC
The U.S.' AI love affair with the UAE isn't just about access — it's about dominance
DUBAI, United Arab Emirates — Deep in the oil-rich deserts of the Middle East, the United Arab Emirates is on a mission to establish supremacy in the field of artificial intelligence. Seven thousand miles across the planet, the United States, led by President Donald Trump, wants American firms to dominate the global AI race. While their goals may be separated by continents, their ambitions are strikingly aligned. The U.S. currently makes the world's most advanced semiconductor chips, while the UAE and neighboring Gulf countries have the abundant, cheap energy needed to power enormous AI data centers. The two countries have been allies for half a century, and Abu Dhabi embraced Trump during the U.S. president's visit this month with unprecedented fanfare and investment pledges, many of which focused on tech and AI. In the eyes of many investors, financial leaders, and political powers players from Silicon Valley and Washington to Abu Dhabi and Dubai, the two countries' ever-strengthening AI alliance — to which hundreds of billions of dollars have already been committed — is a match made in heaven. "Energy‑rich Gulf nations join the roster of trusted partners just as U.S. data‑center grids hit their physical limits," Myron Xie, an analyst at SemiAnalysis, told CNBC. At the same time, "the UAE gains access to advanced compute and talent, helping it pursue its own sovereign AI goals," Xie said. "The Middle East, flush with cheap energy and capital, is poised to become the next regional AI hub." In the UAE, the developments are part of a long-term strategy by the Gulf sheikhdom to position itself as a global leader in AI. This, the country's leadership holds, will enhance its geopolitical influence, diversify its economy beyond crude oil dependency, and assert itself as a technological powerhouse. The goal for Washington is clear: to ensure American companies lead the global AI race with China and spread American tech around the world. Trump's Middle East visit in mid-May — which featured stops in Riyadh, Doha, and Abu Dhabi — saw the announcement of over $200 billion in commercial deals between the U.S. and the UAE. This brought the total of investment agreements in the Gulf region, including those from Saudi Arabia and Qatar, to over $2 trillion. As part of the Abu Dhabi deals, OpenAI, Oracle, Nvidia and Cisco Systems announced that they will help build Stargate UAE AI campus launching in 2026. The Stargate Project is a $500 billion private sector AI-focused investment vehicle, announced by OpenAI in January in partnership with Abu Dhabi investment firm MGX and Japan's SoftBank. The companies said an initial 200-megawatt AI cluster should launch in Abu Dhabi next year. And the AI campus deal means the UAE gets access to many of Nvidia's latest chips, American technology and software. It's the kind of agreement that would have faced restrictions under the previous U.S. administration, but Trump has looked to change the way is approaching tech export restrictions. His administration plans to rescind a Biden era "AI diffusion rule," which imposed strict export controls on advanced AI chips even to U.S.-friendly nations. that doing away with these limits could open the door for the sensitive American technology to end up in the hands of rivals like China — a topic of ongoing debate among U.S. lawmakers and security professionals. Once known primarily as a partnership centered around oil exports and weapons purchases, the pillars of the U.S.-Gulf relationship are changing, says Mohammed Soliman, senior fellow at the Middle East Institute in Washington DC. "Compute, not crude, is going to be the central pillar of the U.S.-Gulf relationship," Soliman said. "Moving forward, it's no longer going to be only about energy policy; it is going to be about compute and how we and the Gulf are building an AI ecosystem that's able to service third markets, emerging markets." Compute, in the context of AI, refers to the computational resources, like hardware and processing power, needed to train and run AI models. "And this is a huge inflection point for the relationship [compared to] where we were a few years ago," Soliman said, speaking on a Middle East Institute podcast recorded on May 19. "Moving forward, the relationship is going to be much more impactful on technical questions around AI, data centers, and chips than ever before." Notably, the UAE has bet fully on a U.S.-led AI future — a particularly salient point within the context of U.S.-China competition. Emirati AI company G42, which has major partnerships with OpenAI, Nvidia and Microsoft, to name a few, has fully divested from Chinese companies — including an estimated $100 million stake in TikTok owner ByteDance — to avert U.S. Commerce Department sanctions and retain access to Nvidia chips and other U.S. technology that powers AI applications. "So far right now, we are racing to have the best large language model and ultimately to have AGI (artificial general intelligence)," said Baghdad Gherras, a UAE-based venture partner at Antler, which invests in early-stage AI ventures. AGI generally refers to artificial intelligence that is smarter than humans, though definitions vary. "For the UAE, if they want to be a leader in the AI race, the first thing that they have to secure is compute. If you don't have compute, you won't have a seat in terms of AI leadership," Gherras told CNBC. He added that the UAE "decided to re-shift the geo-economic focus from China to the U.S., because they understood that Nvidia makes by far the best chips for AI, but also the entire semiconductor supply chain is mostly in Taiwan." Still, Gherras noted, China "is catching up really fast, crazy fast." The UAE's development of its own large language model (LLM), Falcon AI, represents a major step for the region in AI development — but it also provides the foundation for the country's geopolitical and economic ambitions to dominate the AI market within the next decade. Such a position would also enhance the Emirates' diplomatic leverage, allowing it to play a more influential role in global tech governance and policy discussions. "If those ambitions become reality, you might see the Gulf acting as a region that offers compute as a service for the rest of emerging markets," the Middle East Institute's Soliman said. "Think about the Gulf as a place that houses large language models in Swahili, in Hindi, in these languages, and they are able to offer housing data, training data, inference for all these economies, because they have the infrastructure," he added. "So they become their AI leader for the emerging markets." "And this is a tremendous, tremendous level of influence, tremendous level of development," Soliman emphasized. "Where they used to serve as energy producers, to become a back-end for AI applications — this is really, really massive." Part of the U.S.'s push in the UAE and the broader region comes down to a desire for American technology to establish supremacy globally and push back the advances of China. On the one hand, U.S. export curbs have restricted access for companies like Nvidia to sell advanced technology to China. It has also stopped China access some technology to advance its own development in areas like semiconductors and AI. At the same time, Washington is opening up new markets, like the Middle East, to its biggest tech companies. "The move has a political angle, as it bolsters the U.S. compute supply chains while constraining China. It grants the U.S. an edge in the AI arms race, positioning the country for continued leadership," David Meier, economist at Julius Baer, said in note earlier this month. Beijing and Chinese companies have been trying to access new markets to push their technology across the globe, especially in areas like AI. But the U.S. has been working to entrench itself first and strike partnerships with governments to do so. "The race is on to diffuse U.S.-based AI into every part of the world," Daniel Newman, CEO of Futurum Group, told CNBC on Tuesday. American companies have taken up the call. OpenAI, which struck a deal with the UAE last week to build AI infrastructure and roll out ChatGPT nationwide, has positioned itself as a countermeasure to China and as the business able to deliver U.S. artificial intelligence to countries around the world. In February, OpenAI's chief global affairs officer Chris Lehane told CNBC that the company sees a world in which there are two major AI models — one led by China's Communist Party and a U.S.-led "small 'd' democratic" AI. "If you're a country and you're looking to build your own AI ecosystem, your own AI hub, you're building developers in your country which are going to be some version of the companies of the future, I think you would prefer to be seeing that built on a democratic AI system because it is going to facilitate your country being able to use this technology for your own nation building purposes," Lehane said.
Globe and Mail
22-05-2025
- Business
- Globe and Mail
Will $10,000 Invested in CoreWeave Stock (an Nvidia-Backed AI Company) Be Worth $1 Million in a Decade?
CoreWeave (NASDAQ: CRWV) is the hottest new artificial intelligence stock on the market. The company held its initial public offering (IPO) in March 2025. The stock started trading at $39 per share but has since risen 130% to $90 per share. Those gains are partially due to strong first-quarter financial results, but the company also benefited when Nvidia disclosed its ownership of 24 million shares last week. Its stake is currently worth more than $2 billion and represents about 5% of outstanding shares. Where to invest $1,000 right now? Our analyst team just revealed what they believe are the 10 best stocks to buy right now. Continue » Artificial intelligence promises to created substantial wealth in the coming years. But can CoreWeave stock turn $10,000 into $1 million by 2035? Here's what investors should know. CoreWeave is a leading provider of artificial intelligence infrastructure CoreWeave provides cloud infrastructure and software services. Whereas hyperscalers like Amazon and Microsoft offer similar solutions, CoreWeave is differentiated because its data centers are purpose-built for artificial intelligence (AI) and other accelerated computing workloads. The company has been very successful with that strategy. Despite competing against much larger cloud providers, CoreWeave was recently ranked as the best GPU cloud by research company SemiAnalysis. CoreWeave also reported impressive financial results in its first quarter post-IPO. Revenue increased 420% to $981 million and adjusted operating income increased 550% to $162 million. CoreWeave also announced its acquisition of AI developer platform Weights & Biases. That move extends the utility of its platform. To elaborate, CoreWeave provides the underlying infrastructure needed for AI workloads, while Weights & Biases has the tools developers need to train, evaluate, and monitor AI models. Importantly, CoreWeave currently trades at 16 times sales, which is neither outrageously expensive nor cheap. For context, cloud services provider Cloudflare trades at 31 times sales, while AI server manufacturer Super Micro Computer trades at 1.2 times sales. The sales multiple a stock commands is usually a function of its earnings capacity. But it is difficult to know how profitable CoreWeave will be in the future because the company is still losing money on a GAAP basis. Investors should be aware of that risk. Statistically, CoreWeave stock is unlikely to turn $10,000 into $1 million by 2035 Turning $10,000 into $1 million requires a 100-fold return, which is the same as a 9,900% gain. Not many stocks ever generate that much upside, and even fewer do so in a decade. In fact, only one stock in the S&P 500 would have made the cut the past decade. Nvidia shares have advanced 25,400% since May 2015. That 255-fold return would have turned $10,000 into $2.5 million. However, the odds improve as the holding period lengthens. Three other stocks currently in the S&P 500 generated 100-fold returns in the past 15 years: Texas Pacific Land gained 15,200%. Axon Enterprise gained 15,100%. Broadcom gained 11,400%. Similarly, if we lengthen the time horizon to two decades, another seven stocks join the list: Netflix gained 51,700%. Booking Holdings gained 21,900%. Apple gained 15,400%. Amazon gained 11,400%. Intuitive Surgical gained 10,600%. Monolithic Power Systems gained 10,500%. Deckers Outdoor gained 10,000%. Here's the bottom line: CoreWeave is currently a $42 billion company, which means its market value would need to reach $4.2 trillion for the stock to generate 100-fold returns. History says that outcome is unlikely over any period, especially the next decade. But it's certainly possible, and the odds improve as the time horizon expands. I think CoreWeave could generate substantial upside in the next decade. And who knows: 100-fold returns may be possible during the next two decades. Regardless, investors should watch this up-and-coming AI company closely. The stock is not cheap, but I think it makes sense to buy a small position today. 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 10 best stocks 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 $642,582!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you'd have $829,879!* Now, it's worth noting Stock Advisor 's total average return is975% — a market-crushing outperformance compared to172%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 May 19, 2025 John Mackey, former CEO of Whole Foods Market, an Amazon subsidiary, is a member of The Motley Fool's board of directors. Trevor Jennewine has positions in Amazon and Axon Enterprise. The Motley Fool has positions in and recommends Amazon, Apple, Axon Enterprise, Booking Holdings, Cloudflare, Deckers Outdoor, Intuitive Surgical, Microsoft, and Netflix. The Motley Fool recommends Broadcom and Monolithic Power Systems and 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.
Yahoo
19-04-2025
- Business
- Yahoo
'Nuclear-level product': Huawei launches new AI solution that rivals Nvidia's 72-GPU NVL72 but it is far less efficient
When you buy through links on our articles, Future and its syndication partners may earn a commission. Huawei launches CloudMatrix 384 Supernode to rival Nvidia's NVL72 AI CloudMatrix 384 delivers nearly double the compute power with superior memory and bandwidth Consumes almost four times the power but system efficiency is less critical in China Huawei has been positioning itself as the Chinese Nvidia for some time, and now, the South China Morning Post reports the company has launched a new AI infrastructure architecture set to rival the US chip giant's NVL72 system. Nvidia's NVL72 links 72 GPUs using NVLink technology, allowing them to function as a single, powerful GPU. Built for trillion-parameter AI models, it delivers real-time inference at speeds up to 30 times faster than previous systems by avoiding traditional data transfer bottlenecks. SCMP said Huawei's rival to this, the CloudMatrix 384 Supernode, has been described as a "nuclear-level product" by unnamed Huawei sources. According to reports, it uses 384 Ascend 910C chips to deliver 300 petaflops of dense BF16 compute, that's almost double the 180 petaflops offered by Nvidia's NVL72. The CloudMatrix 384 Supernode has so far been deployed at Huawei's data centers in Wuhu, a city in central Anhui province, the report says. Writing about the new product, SemiAnalysis confirms this rack-scale solution competes directly with Nvidia's GB200 NVL72 and, in some metrics, outperforms it. The site says that despite sanctions, China's domestic semiconductor capabilities are growing and Huawei's strength lies in system-level engineering, including networking, optics, and software. While the Ascend chips rely heavily on foreign supply chains, such as HBM from Samsung and wafers from TSMC, Huawei has managed to skirt export controls through complex sourcing strategies. The CloudMatrix 384 doesn't only outperform the NVL72 in terms of compute, it also offers 3.6x aggregate memory capacity and 2.1x more memory bandwidth. However, something Huawei is likely less keen to shout about is that it consumes nearly four times the power. SemiAnalysis points out that in China this is less of an issue than you might think. Abundant power generation means efficiency is less of a constraint compared to in the West, and China is expanding its energy grid rapidly, supporting such power-hungry AI infrastructure, even when supply outstrips demand. US sanctions allow Huawei 'case-by-case' access to chip-making equipment Huawei is pairing its supercharged SSD with a 60-year old technology Huawei's blueprint for a post-Trump, non-US centric technology world
South China Morning Post
13-03-2025
- Business
- South China Morning Post
China manufacturing, AI pose an ‘existential threat' to US in robotics sector
Advertisement Automation enabled by intelligent robots will allow for a massive expansion in production capacity across industries, and China is currently 'the only country that is positioned' to achieve a high level of automation, according to a report published this week by US-based independent research company SemiAnalysis. 'China's robotics localisation effort is well under way,' the report said. It pointed out that Chinese manufacturers' share in the world's largest robotics market was approaching 50 per cent, up from 30 per cent in 2020. 'While Chinese manufacturers are currently on par with Western giants in the low-end market, our supply chain review leads us to believe that local firms are beginning to take over the higher-end market segments.' The rise of Unitree Robotics exemplifies that shift. The Hangzhou -based start-up makes 'the only viable humanoid robot on the market', according to the report. It pointed out that Unitree's G1 robot was also 'entirely decoupled from American components'. Advertisement Should China achieve full-scale automation 'without the US following suit', it would pose 'an existential threat' to the world's largest economy, according to the report. That scenario would see China benefit from a 'massive expansion in production capacities [across various industries] supported by intelligent robotics systems'.
USA Today
15-02-2025
- Business
- USA Today
You may have sold Nvidia for the wrong reason. 3 reasons to consider buying this AI stock
You may have sold Nvidia for the wrong reason. 3 reasons to consider buying this AI stock Show Caption Hide Caption Trump, China, and the AI Summit: A global power struggle The AI Summit in Paris has drawn world leaders, tech executives, and researchers to shape the future of artificial intelligence. While many seek global cooperation, the event highlights deep geopolitical divides, particularly between the US and China. unbranded - Newsworthy Nvidia's stock price has declined in early 2025 due to concerns about competition and US government restrictions. The emergence of a Chinese AI model claiming to be cheaper to train raised concerns about demand for Nvidia's chips. However, analysis suggests the Chinese model's costs may be higher than initially claimed, and Nvidia's stock remains a good long-term investment opportunity. Increased demand for AI applications and continued investment in AI infrastructure by major companies suggest a positive outlook for Nvidia. Nvidia (NASDAQ: NVDA) has had a forgettable start to 2025 as shares of the semiconductor giant are down more than 3% as of this writing, and multiple factors out of the company's control have played a part in its decline. For instance, the previous Biden administration proposed wide-ranging restrictions on sales of Nvidia's chips to foreign customers, but the impact of those restrictions was mitigated to some extent by the recent announcement of the Stargate project that could see $500 billion being poured into artificial intelligence (AI) infrastructure in the U.S. This development gave Nvidia stock a shot in the arm, but the chip designer would witness another sell-off very soon. Nvidia stock fell thanks to DeepSeek, but investors may have jumped the gun Chinese AI start-up DeepSeek released its R1 reasoning model and claimed that it was trained for a paltry $5.6 million. DeepSeek's model was good enough to compete with the o1 reasoning model from OpenAI, a company that has been spending billions to build its AI infrastructure using chips from Nvidia. So, the low-cost nature and efficiency of the Chinese company's model sent Nvidia stock packing. Investors were worried about a potential drop in demand for its graphics cards that are being used for AI training and inference by major cloud computing companies and governments. The semiconductor giant shed almost $600 billion of its market cap in a single day on Jan. 27 following DeepSeek's purported breakthrough. However, a report from semiconductor industry analysis company SemiAnalysis (via Tom's Hardware) suggests that DeepSeek may not have revealed the actual cost of training its AI model. What is DeepSeek? How a small Chinese startup shook up the AI sector SemiAnalysis points out that DeepSeek reportedly incurred $1.6 billion in hardware expenses. It also adds that the Chinese start-up has access to 50,000 of Nvidia's previous-generation Hopper graphics processing units (GPUs), including 10,000 units of the flagship H100 processor. SemiAnalysis further points out that the $6 million figure highlighted by DeepSeek only refers to the potential money spent on training the model. It doesn't consider other costs associated with research, data processing, fine-tuning the model, and infrastructure expenses. Given that DeepSeek has reportedly spent over $500 million on AI funding since its inception in 2023 and has its own data centers, there is a good chance that the cost of training the R1 model that sent Nvidia stock plunging was actually higher than what's being touted by the Chinese company. If that's indeed the case, then investors may have hit the panic button for the wrong reason. However, the good part is that Nvidia's poor start to the year means that investors have a window to buy this fast-growing company on the dip. Here are three reasons doing that could turn out to be a smart move. Three reasons to buy Nvidia right now The first reason to buy Nvidia is its valuation. The stock's expensive valuation following its tremendous rally over the past couple of years has been a cause for concern for investors and analysts. However, it is now trading at quite attractive levels. Though Nvidia's trailing price-to-earnings (P/E) ratio of 51 is higher than the tech-laden Nasdaq-100 index's 33.4, the forward earnings multiple of 30 is lower than that. The second reason why investors should consider buying Nvidia is related to DeepSeek. Assuming that DeepSeek's claims are indeed true and it has managed to develop a low-cost model, it could lead to an increase in the demand for AI applications. British economist William Stanley Jevons observed in 1865 that the increased efficiency in coal consumption wouldn't reduce the demand for coal. It would instead spur the usage of coal in more industries. This concept, known as the Jevons Paradox, can be seen in many other applications. For instance, the increasing fuel efficiency of vehicles has reportedly led to an increase in distance traveled while the arrival of low-cost LED bulbs hasn't necessarily brought down electricity bills as people have tended to install more lights because of reduced costs. So, efficient AI models could lead to an increase in their demand, and that's why the need for Nvidia's chips is likely to remain solid. As such, DeepSeek's purported breakthrough isn't necessarily a bad thing for Nvidia. The final reason to buy Nvidia stock following its recent pullback is that cloud computing giants are set to keep spending more money on AI infrastructure. President Donald Trump's $500 billion AI infrastructure push that's being driven by the likes of Oracle, OpenAI, SoftBank, and Abu Dhabi-based AI investment vehicle MGX to build AI data centers will require more chips. Meanwhile, recent announcements from the CEOs of Meta Platforms and Microsoft supporting higher spending on AI also point toward healthy AI chip demand. Finally, a spike in bookings for the advanced machines sold by Dutch semiconductor equipment giant ASML provides further evidence that the appetite for advanced chips to support AI workloads isn't going away. So, it won't be surprising to see Nvidia's fortunes turn around, which is why investors should consider using the recent pullback in this AI stock given the healthy growth that it is capable of delivering in the long run. Randi Zuckerberg, a former director of market development and spokeswoman for Facebook and sister to Meta Platforms CEO Mark Zuckerberg, is a member of The Motley Fool's board of directors. Harsh Chauhan has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends ASML, Meta Platforms, Microsoft, Nvidia, and Oracle. 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. The Motley Fool is a USA TODAY content partner offering financial news, analysis and commentary designed to help people take control of their financial lives. Its content is produced independently of USA TODAY. Should you invest $1,000 in Nvidia right now? Offer from the Motley Fool: Before you buy stock in Nvidia, consider this: The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Nvidia wasn't one of them. The 10 stocks that made the cut could produce monster returns in the coming years. Consider when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you'd have $829,128!* Stock Advisor provides investors with an easy-to-follow blueprint for success, including guidance on building a portfolio, regular updates from analysts, and two new stock picks each month. TheStock Advisorservice has more than quadrupled the return of S&P 500 since 2002*. Learn more » *Stock Advisor returns as of February 7, 2025
