Arm will reportedly start making its own chips this year after signing Meta as a customer, according to the Financial Times (paywalled). TechCrunch reports: The chip is expected to be a CPU for servers in large data centers and can be customized for various customers. Arm will outsource its production. The first in-house Arm chip will be unveiled as early as this summer, the Financial Times also reported.

This is a notable change in strategy for the semiconductor company, which usually licenses its chip blueprints to companies like Apple and Nvidia. Making its own chips will turn some of its existing customers into competitors.

The European Union pledged to mobilize 200 billion euros ($206.15 billion) to invest in AI as the bloc seeks to catch up with the U.S. and China in the race to train the most complex models. From a report: European Commission President Ursula von der Leyen said that the bloc wants to supercharge its ability to compete with the U.S. and China in AI. The plan -- dubbed InvestAI -- includes a new 20 billion-euro fund for so-called AI gigafactories, facilities that rely on powerful chips to train the most complex AI models. "We want Europe to be one of the leading AI continents, and this means embracing a life where AI is everywhere," von der Leyen said at the AI Action Summit in Paris.

The announcement underscores efforts from the EU to position itself as a key player in the AI race. The bloc has been lagging behind the U.S. and China since OpenAI's 2022 release of ChatGPT ushered in a spending bonanza. [...] The EU is aiming to establish gigafactories to train the most complex and large AI models. Those facilities will be equipped with roughly 100,000 last-generation AI chips, around four times more than the number installed in the AI factories being set up right now.

An anonymous reader quotes a report from CBS News: Animal shelters, rescues, and veterinarian clinics around the U.S. are posting on social media telling pet owners to check their four-legged friends' microchips after learning a major microchip company [called Save This Life] is no longer providing services. [...] If you're unsure which company your cats or dogs' chips are registered with, check them. "You can go to your local veterinarian office, a local police station, or even a local animal shelter like HARP, and we can help check that for you and scan your animal. And then you take that number that's on there and there's a tool online where you can go look it up," [said Dan Cody, Executive Director of Humane Animal Rescue of Pittsburgh].

He said you check the number by using the AAHA Universal Microchip Lookup Tool at this link. If you discover your pet's microchip was registered to the company that's ceased operations, you'll need to register with a different company. "So, if you find that you are affected by this, you're going to want to go to one of these other websites that do the registrations. So, things like AKC Reunite, and PetLink. 24PetWatch these are all large companies who've been around for a long time and have good reputations," said Cody.

The American Kennel Club shared a post from its AKC Reunite Facebook page, encouraging people to enroll in microchips with AKC Reunite. The post said in part, "If your dog or cat has a microchip number that starts with 991 or 900164 then it could be a Save This Life microchip. Save This Life suddenly closed, and your pet may not be protected." Cody said if your furry best friend isn't microchipped, take them to a vet or shelter like HARP to get one implanted under their skin so they have a permanent ID. Microchipping can be done at HARP's East Side and North Side Veterinary Medical Center by appointment.

Google DeepMind CEO Demis Hassabis downplayed the technological significance of DeepSeek's latest AI model, despite its market impact. "Despite the hype, there's no actual new scientific advance there. It's using known techniques," Hassabis said on Sunday. "Actually many of the techniques we invented at Google and at DeepMind."

Hassabis acknowledged that Deepseek's AI model "is probably the best work" out of China, but its capabilities, he said, is "exaggerated a little bit."DeepSeek's launch last month triggered a $1 trillion U.S. market sell-off.

Slashdot reader headlessbrick writes: Google security researchers have discovered a way to bypass AMD's security, enabling them to load unofficial microcode into its processors and modify the silicon's behaviour at will. To demonstrate this, they created a microcode patch that forces the chips to always return 4 when asked for a random number.

Beyond simply allowing Google and others to customize AMD chips for both beneficial and potentially malicious purposes, this capability also undermines AMD's secure encrypted virtualization and root-of-trust security mechanisms.
Obligatory XKCD.

An anonymous reader quotes a report from Ars Technica: In today's issue of Nature, a team at Oxford University describes using quantum teleportation to link two pieces of quantum hardware that were located about 2 meters apart, meaning they could easily have been in different rooms entirely. Once linked, the two pieces of hardware could be treated as a single quantum computer, allowing simple algorithms to be performed that involved operations on both sides of the 2-meter gap. [...] The Oxford team was simply interested in a proof-of-concept, and so used an extremely simplified system. Each end of the 2-meter gap had a single trap holding two ions, one strontium and one calcium. The two atoms could be entangled with each other, getting them to operate as a single unit.

The calcium ion served as a local memory and was used in computations, while the strontium ion served as one of the two ends of the quantum network. An optical cable between the two ion traps allowed photons to entangle the two strontium ions, getting the whole system to operate as a single unit. The key thing about the entanglement processes used here is that a failure to entangle left the system in its original state, meaning that the researchers could simply keep trying until the qubits were entangled. The entanglement event would also lead to a photon that could be measured, allowing the team to know when success had been achieved (this sort of entanglement with a success signal is termed "heralded" by those in the field).

The researchers showed that this setup allowed them to teleport with a specific gate operation (controlled-Z), which can serve as the basis for any other two-qubit gate operation -- any operation you might want to do can be done by using a specific combination of these gates. After performing multiple rounds of these gates, the team found that the typical fidelity was in the area of 70 percent. But they also found that errors typically had nothing to do with the teleportation process and were the product of local operations at one of the two ends of the network. They suspect that using commercial hardware, which has far lower error rates, would improve things dramatically. Finally, they performed a version of Grover's algorithm, which can, with a single query, identify a single item from an arbitrarily large unordered list. The "arbitrary" aspect is set by the number of available qubits; in this case, having only two qubits, the list maxed out at four items. Still, it worked, again with a fidelity of about 70 percent.

While the work was done with trapped ions, almost every type of qubit in development can be controlled with photons, so the general approach is hardware-agnostic. And, given the sophistication of our optical hardware, it should be possible to link multiple chips at various distances, all using hardware that doesn't require the best vacuum or the lowest temperatures we can generate. That said, the error rate of the teleportation steps may still be a problem, even if it was lower than the basic hardware rate in these experiments. The fidelity there was 97 percent, which is lower than the hardware error rates of most qubits and high enough that we couldn't execute too many of these before the probability of errors gets unacceptably high.

Arm has dropped its attempt to terminate Qualcomm's Architecture License Agreement (ALA), allowing Qualcomm to continue developing and producing Arm-compatible chips for PCs, smartphones, and servers. "The Brit biz had sought to end that license in a lawsuit it brought against Qualcomm in 2022," notes The Register. "That suit is rooted in Qualcomm's 2021 acquisition of a startup called Nuvia, which was co-founded by the brains behind Apple's custom processors and had signed an architecture license agreement (ALA) with Arm that allowed it to design its own Arm-compatible CPU cores." From the report: On Wednesday, Qualcomm's latest quarterly financial report [PDF] revealed Arm had indicated on January 8, 2025 it was no longer seeking to kill off Qualcomm's ALA. During Qualcomm's Q1 2025 earnings conference call with Wall Street, CEO Cristiano Amon confirmed Arm "has no current plan to terminate the Qualcomm Architecture License Agreement. We're excited to continue to develop performance leading, world-class products that benefit consumers worldwide that include our incredible Oryon custom CPUs." [...]

On the other side of the fence, Arm noted in a regulatory filing [PDF] that post-trial motions had been filed on both sides to clarify the legal situation following the jury's verdicts, and a new trial may be sought. On its own latest quarterly earnings call, which like Qualcomm's took place on Wednesday, Arm's CFO Jason Child was asked about the impact of the case. He said Arm's revenue forecasts assumed the biz was "not going to prevail in that lawsuit," and that it expected to continue receiving payments from Qualcomm, which licenses various technologies from Arm and doesn't just hold an ALA.

"The primary reason for the lawsuit very much was around defending our IP and that's important," Child said. "But from a financial perspective, we had assumed that we'll continue to be receiving royalties at basically the same rates that they've been paying for in the past and will continue to pay." Qualcomm continues to pursue another case against Arm, alleging the UK outfit didn't honor some of its contractual obligations. Arm reckons that matter will reach the courts in the first half of 2026.

DeepSeek's AI app will highly likely face a US consumer ban after topping download charts on Apple's App Store and Google Play, according to analysts at US investment bank Jefferies. The US federal government, Navy and Texas have already banned the app, and analysts expect broader restrictions using legislation similar to that targeting TikTok.

While consumer access may be blocked, US developers could still be allowed to self-host DeepSeek's model to eliminate security risks, the analysts added. Even if completely banned, DeepSeek's impact on pushing down AI costs will persist as US companies work to replicate its technology, Jefferies said in a report this week reviewed by Slashdot.

The app's pricing advantage remains significant, with OpenAI's latest o3-mini model still costing 100% more than DeepSeek's R1 despite being 63% cheaper than o1-mini. The potential ban comes amid broader US-China tech tensions. While restrictions on H20 chips appear unlikely given their limited training capabilities, analysts expect the Biden administration's AI diffusion policies to remain largely intact under Trump, with some quota increases possible for overseas markets based on their AI activity levels.

Google aims to release commercial quantum computing applications within five years, challenging Nvidia's prediction of a 20-year timeline. "We're optimistic that within five years we'll see real-world applications that are possible only on quantum computers," founder and lead of Google Quantum AI Hartmut Neven said in a statement. Reuters reports: Real-world applications Google has discussed are related to materials science - applications such as building superior batteries for electric cars - creating new drugs and potentially new energy alternatives. [...] Google has been working on its quantum computing program since 2012 and has designed and built several quantum chips. By using quantum processors, Google said it had managed to solve a computing problem in minutes that would take a classical computer more time than the history of the universe.

Google's quantum computing scientists announced another step on the path to real world applications within five years on Wednesday. In a paper published in the scientific journal Nature, the scientists said they had discovered a new approach to quantum simulation, which is a step on the path to achieving Google's objective.

Senator Josh Hawley has introduced a bill that would criminalize the import, export, and collaboration on AI technology with China. What this means is that "someone who knowingly downloads a Chinese developed AI model like the now immensely popular DeepSeek could face up to 20 years in jail, a million dollar fine, or both, should such a law pass," reports 404 Media. From the report: Hawley introduced the legislation, titled the Decoupling America's Artificial Intelligence Capabilities from China Act, on Wednesday of last year. "Every dollar and gig of data that flows into Chinese AI are dollars and data that will ultimately be used against the United States," Senator Hawley said in a statement. "America cannot afford to empower our greatest adversary at the expense of our own strength. Ensuring American economic superiority means cutting China off from American ingenuity and halting the subsidization of CCP innovation."

Hawley's statement explicitly says that he introduced the legislation because of the release of DeepSeek, an advanced AI model that's competitive with its American counterparts, and which its developers claimed was made for a fraction of the cost and without access to as many and as advanced of chips, though these claims are unverified. Hawley's statement called DeepSeek "a data-harvesting, low-cost AI model that sparked international concern and sent American technology stocks plummeting." Hawley's statement says the goal of the bill is to "prohibit the import from or export to China of artificial intelligence technology, "prohibit American companies from conducting AI research in China or in cooperation with Chinese companies," and "Prohibit U.S. companies from investing money in Chinese AI development."

On January 24th Brooklyn blogger Jeffrey Emanuel made the case for shorting NVIDIA, remembers MarketWatch, "due to a number of shifting tides in the AI world, including the emergence of a China-based company called DeepSeek."

He published his 12,000-word post "on his personal blog and then shared it with the Value Investors Club website and across Reddit, X and other platforms." The next day he saw 35 people read his post. "But then the post started to go viral..." Well-known venture capitalist Chamath Palihapitiya shared Emanuel's post on Nvidia's short case with his 1.8 million X followers. Successful early stage investor Naval Ravikant shared the post with his 2.6 million followers... Morgan Brown, a vice president of product and growth at Dropbox, pointed to it in a thread that was viewed over 13 million times. Emanuel's own X post got nearly half a million views. He also quickly gained about 13,000 followers on the platform, going from about 2,000 to more than 15,000 followers...

[Emanuel] pointed to the fact that so many people in San Jose were reading his blog post. He theorized that many of them were Nvidia employees with thousands — or even millions — of dollars worth of Nvidia stock tied up in employee stock options. With that much money in a single asset, Emanuel speculated that many were already debating whether to hold the stock or sell it to lock in profits. He believes his blog post helped convince some of them to sell. "A lot of the sell pressure you saw on Monday morning wasn't necessarily what you might think. I believe a fair amount of that was from shares that had never been active because they had been sitting in workplace.schwab.com accounts..."
Emanuel stresses he's "the most bullish on AI," with MarketWatch emphasizing that "while the points Emanuel laid out in his blog post might be bearish for Nvidia, he still thinks they paint a positive future for AI." Nevertheless, Monday NVIDIA's market capitalization dropped $600 billion, which MarketWatch calls "the largest single-day market-cap drop to date for any company." What countless Wall Street firms and investment analysts had seemingly missed was being pointed out by some guy in his apartment.... Matt Levine, the prominent Bloomberg News financial columnist, noted the online chatter that claimed Emanuel's post "was an important catalyst" for the stock-market selloff and said it was a "candidate for the most impactful short research report ever." Emanuel spent the rest of the week booked solid as hedge funds paid him $1,000 per hour to speak on the phone and give his take on Nvidia and AI...

Emanuel wrote that the industry may be running low on quality data to train that AI — that is, a potential "data wall" is looming that could slow down AI scaling and reduce some of that need for training resources... Some of these companies, like Alphabet, have also been investing in building out their own semiconductor chips. For a while, Nvidia's hardware has been the best for training AI, but that might not be the case forever as more companies, such as Cerebras, build better hardware. And other GPU makers like AMD are updating their drivers software to be more competitive with Nvidia... Add all these things together — unsustainable spending and data-center building, less training data to work with, better competing hardware and more efficient AI — and you get a future where it's harder to imagine Nvidia's customers spending as much as they currently are on Nvidia hardware... "If you know that a company will only earn supersized returns for a couple years, you don't apply a multiple. You certainly don't put a 30-times multiple," Emanuel told MarketWatch.
The article notes that DeepSeek "is open-source and has been publishing technical papers out in the open for the past few months... The $5.6 million training-cost statistic that many investors cited for sparking the DeepSeek market panic was actually revealed in the V3 technical paper published on Dec. 26."

Nearly three years ago a team of Chinese AI engineers working for DeepSeek's parent company unveiled an earlier AI supercomputer that the Washington Post says was constructed from 10,000 A100 GPUs purchased from Nvidia. Roughly six months later "Washington had banned Nvidia from selling any more A100s to China," the article notes.

Remember that number as you read this. 10,000 A100 GPUs... DeepSeek's new chatbot caused a panic in Silicon Valley and on Wall Street this week, erasing $1 trillion from the stock market. That impact stemmed in large part from the company's claim that it had trained one of its recent models on a minuscule $5.6 million in computing costs and with only 2,000 or so of Nvidia's less-advanced H800 chips.

Nvidia saw its soaring value crater by $589 billion Monday as DeepSeek rocketed to the top of download charts, prompting President Donald Trump to call for U.S. industry to be "laser focused" on competing... But a closer look at DeepSeek reveals that its parent company deployed a large and sophisticated chip set in its supercomputer, leading experts to assess the total cost of the project as much higher than the relatively paltry sum that U.S. markets reacted to this week... Lennart Heim, an AI expert at Rand, said DeepSeek's evident access to [the earlier] supercomputer would have made it easier for the company to develop a more efficient model, requiring fewer chips.
That earlier project "suggests that DeepSeek had a major boost..." according to the article, "with technology comparable to that of the leading U.S. AI companies." And while DeepSeek claims it only spent $5.6 million to train one of its advanced models, "its parent company has said that building the earlier supercomputer had cost 1 billion yuan, or $139 million.") Yet the article also cites the latest insights Friday from chip investment company SemiAnalysis, summarizing their finding that DeepSeek "has spent more than half a billion dollars on GPUs, with total capital expenditures of almost $1.3 billion."

The article notes Thursday remarks by OpenAI CEO Sam Altman that DeepSeek's energy-efficiency claims were "wildly overstated... This is a model at a capability level that we had quite some time ago." And Palmer Luckey called DeepSeek "legitimately impressive" on X but called the $5.6 million training cost figure "bogus" and said the Silicon Valley meltdown was "hysteria." Even with these higher total costs in mind, experts say, U.S. companies are right to be concerned about DeepSeek upending the market. "We know two things for sure: DeepSeek is pricing their services very competitively, and second, the performance of their models is comparable to leading competitors," said Kai-Shen Huang, an AI expert at the Research Institute for Democracy, Society and Emerging Technology, a Taipei-based think tank. "I think DeepSeek's pricing strategy has the potential to disrupt the market globally...."

China's broader AI policy push has helped create an environment conducive for a company like DeepSeek to rise. Beijing announced an ambitious AI blueprint in 2017, with a goal to become a global AI leader by 2030 and promises of funding for universities and private enterprise. Local governments across the nation followed with their own programs to support AI.

Microsoft is charging business customers a $400 premium for Surface devices equipped with Intel's latest Core Ultra processors compared to models using Qualcomm's Arm-based chips, the company has disclosed. The Intel-powered Surface Pro tablet and Surface Laptop, starting at $1,499, come with a second-generation Core Ultra 5 processor featuring eight cores, 16GB of memory and 256GB storage.

Comparable Qualcomm-based models begin at $1,099. The new Intel devices will be available to business customers from February 18, though versions with cellular connectivity will launch later. Consumer Surface devices will only be offered with Qualcomm processors. Microsoft also unveiled a USB 4 Dock supporting dual 4K displays and the Surface Hub 3, a conference room computer available in 50-inch or 85-inch touchscreen versions.

An anonymous reader quotes a report from MIT Technology Review: A Canadian startup called Xanadu has built a new quantum computer it says can be easily scaled up to achieve the computational power needed to tackle scientific challenges ranging from drug discovery to more energy-efficient machine learning. Aurora is a "photonic" quantum computer, which means it crunches numbers using photonic qubits -- information encoded in light. In practice, this means combining and recombining laser beams on multiple chips using lenses, fibers, and other optics according to an algorithm. Xanadu's computer is designed in such a way that the answer to an algorithm it executes corresponds to the final number of photons in each laser beam. This approach differs from one used by Google and IBM, which involves encoding information in properties of superconducting circuits.

Aurora has a modular design that consists of four similar units, each installed in a standard server rack that is slightly taller and wider than the average human. To make a useful quantum computer, "you copy and paste a thousand of these things and network them together," says Christian Weedbrook, the CEO and founder of the company. Ultimately, Xanadu envisions a quantum computer as a specialized data center, consisting of rows upon rows of these servers. This contrasts with the industry's earlier conception of a specialized chip within a supercomputer, much like a GPU. [...]

Xanadu's 12 qubits may seem like a paltry number next to IBM's 1,121, but Tiwari says this doesn't mean that quantum computers based on photonics are running behind. In his opinion, the number of qubits reflects the amount of investment more than it does the technology's promise. [...] Xanadu's next goal is to improve the quality of the photons in the computer, which will ease the error correction requirements. "When you send lasers through a medium, whether it's free space, chips, or fiber optics, not all the information makes it from the start to the finish," he says. "So you're actually losing light and therefore losing information." The company is working to reduce this loss, which means fewer errors in the first place. Xanadu aims to build a quantum data center, with thousands of servers containing a million qubits, in 2029. The company published its work on chip design optimization and fabrication in the journal Nature.

Alibaba has unveiled a new version of its AI model, called Qwen2.5-Max, claiming benchmark scores that surpass both DeepSeek's recently released R1 model and industry standards like GPT-4o and Claude-3.5-Sonnet. The model achieves these results using a mixture-of-experts architecture that requires significantly less computational power than traditional approaches.

The release comes amid growing concerns about China's AI capabilities, following DeepSeek's R1 model launch last week that sent Nvidia's stock tumbling 17%. Qwen2.5-Max scored 89.4% on the Arena-Hard benchmark and demonstrated strong performance in code generation and mathematical reasoning tasks. Unlike U.S. companies that rely heavily on massive GPU clusters -- OpenAI reportedly uses over 32,000 high-end GPUs for its latest models -- Alibaba's approach focuses on architectural efficiency. The company claims this allows comparable AI performance while reducing infrastructure costs by 40-60% compared to traditional deployments.

An anonymous reader quotes a report from Ars Technica: Apple-designed chips powering Macs, iPhones, and iPads contain two newly discovered vulnerabilities that leak credit card information, locations, and other sensitive data from the Chrome and Safari browsers as they visit sites such as iCloud Calendar, Google Maps, and Proton Mail. The vulnerabilities, affecting the CPUs in later generations of Apple A- and M-series chip sets, open them to side channel attacks, a class of exploit that infers secrets by measuring manifestations such as timing, sound, and power consumption. Both side channels are the result of the chips' use of speculative execution, a performance optimization that improves speed by predicting the control flow the CPUs should take and following that path, rather than the instruction order in the program. [...]

The researchers published a list of mitigations they believe will address the vulnerabilities allowing both the FLOP and SLAP attacks. They said that Apple officials have indicated privately to them that they plan to release patches. In an email, an Apple representative declined to say if any such plans exist. "We want to thank the researchers for their collaboration as this proof of concept advances our understanding of these types of threats," the spokesperson wrote. "Based on our analysis, we do not believe this issue poses an immediate risk to our users." FLOP, short for Faulty Load Operation Predictor, exploits a vulnerability in the Load Value Predictor (LVP) found in Apple's A- and M-series chipsets. By inducing the LVP to predict incorrect memory values during speculative execution, attackers can access sensitive information such as location history, email content, calendar events, and credit card details. This attack works on both Safari and Chrome browsers and affects devices including Macs (2022 onward), iPads, and iPhones (September 2021 onward). FLOP requires the victim to interact with an attacker's page while logged into sensitive websites, making it highly dangerous due to its broad data access capabilities.

SLAP, on the other hand, stands for Speculative Load Address Predictor and targets the Load Address Predictor (LAP) in Apple silicon, exploiting its ability to predict memory locations. By forcing LAP to mispredict, attackers can access sensitive data from other browser tabs, such as Gmail content, Amazon purchase details, and Reddit comments. Unlike FLOP, SLAP is limited to Safari and can only read memory strings adjacent to the attacker's own data. It affects the same range of devices as FLOP but is less severe due to its narrower scope and browser-specific nature. SLAP demonstrates how speculative execution can compromise browser process isolation.

Nvidia shares plunged 17% on Monday, wiping nearly $600 billion from its market value, after Chinese AI firm DeepSeek's breakthrough, but analysts are questioning the cost narrative. DeepSeek said to have trained its December V3 model for $5.6 million, but chip consultancy SemiAnalysis suggested this figure doesn't reflect total investments. "DeepSeek has spent well over $500 million on GPUs over the history of the company," Dylan Patel of SemiAnalysis said. "While their training run was very efficient, it required significant experimentation and testing to work."

The steep sell-off led to the Philadelphia Semiconductor index's worst daily drop since March 2020 at 9.2%, generating $6.75 billion in profits for short sellers, according to data group S3 Partners. DeepSeek's engineers also demonstrated they could write code without relying on Nvidia's Cuda software platform, which is widely seen as crucial to the Silicon Valley chipmaker's dominance of AI development.

Nvidia has responded to the market panic over Chinese AI group DeepSeek, arguing that the startup's breakthrough still requires "significant numbers of NVIDIA GPUs" for its operation. The US chipmaker, which saw more than $600 billion wiped from its market value on Monday, characterized DeepSeek's advancement as "excellent" but asserted that the technology remains dependent on its hardware.

"DeepSeek's work illustrates how new models can be created using [test time scaling], leveraging widely-available models and compute that is fully export control compliant," Nvidia said in a statement Monday. However, it stressed that "inference requires significant numbers of NVIDIA GPUs and high-performance networking." The statement came after DeepSeek's release of an AI model that reportedly achieves performance comparable to those from US tech giants while using fewer chips, sparking the biggest one-day drop in Nvidia's history and sending shockwaves through global tech stocks.

Nvidia sought to frame DeepSeek's breakthrough within existing technical frameworks, citing it as "a perfect example of Test Time Scaling" and noting that traditional scaling approaches in AI development - pre-training and post-training - "continue" alongside this new method. The company's attempt to calm market fears follows warnings from analysts about potential threats to US dominance in AI technology. Goldman Sachs earlier warned of possible "spillover effects" from any setbacks in the tech sector to the broader market. The shares stabilized somewhat in afternoon trading but remained on track for their worst session since March 2020, when pandemic fears roiled markets.

Meta plans to spend between $60 billion and $65 billion this year to build out AI infrastructure, CEO Mark Zuckerberg said on Friday, joining a wave of Big Tech firms unveiling hefty investments to capitalize on the technology. From a report: As part of the investment, Meta will build a more than 2-gigawatt data center that would be large enough to cover a significant part of Manhattan. The company -- one of the largest customers of Nvidia's coveted artificial intelligence chips -- plans to end the year with more than 1.3 million graphics processors.

"This will be a defining year for AI," Zuckerberg said in a Facebook post. "This is a massive effort, and over the coming years it will drive our core products and business." Zuckerberg expects Meta's AI assistant -- available across its services, including Facebook and Instagram -- to serve more than 1 billion people in 2025, while its open-source Llama 4 would become the "leading state-of-the-art model."

The U.S. may have led China in the AI race for the past decade, according to Alexandr Wang, CEO of Scale AI, but on Christmas Day, everything changed. From a report: Wang, whose company provides training data to key AI players including OpenAI, Google and Meta , said Thursday at the World Economic Forum in Davos, Switzerland, that DeepSeek, the leading Chinese AI lab, released an "earth-shattering model" on Christmas Day, then followed it up with a powerful reasoning-focused AI model, DeepSeek-R1, which competes with OpenAI's recently released o1 model.

"What we've found is that DeepSeek ... is the top performing, or roughly on par with the best American models," Wang said. In an interview with CNBC, Wang described the artificial intelligence race between the U.S. and China as an "AI war," adding that he believes China has significantly more Nvidia H100 GPUs -- AI chips that are widely used to build leading powerful AI models -- than people may think, especially considering U.S. export controls. [...] "The United States is going to need a huge amount of computational capacity, a huge amount of infrastructure," Wang said, later adding, "We need to unleash U.S. energy to enable this AI boom." DeepSeek's holding company is a quant firm, which happened to have a lot of GPUs for trading and mining. DeepSeek is their "side project."