The AI race between the US and China enters a new phase: Washington loosens Nvidia export rules, but Beijing reportedly halts purchases. China wants to shield its chip industry and may require buyers to also purchase domestic chips.
UK company Basecamp Research has developed AI models together with researchers from Nvidia and Microsoft that generate potential new therapies against cancer and multidrug-resistant bacteria from a database of over one million species.
Epoch AI has released a comprehensive database of AI chip sales showing that global computing capacity now exceeds 15 million H100 equivalents. This metric compares the performance of various chips to Nvidia's H100 processor. The data, published on January 8, 2026, reveals that Nvidia's new B300 chip now generates the majority of the company's AI revenue, while the older H100 has dropped below ten percent. The analysis covers chips from Nvidia, Google, Amazon, AMD, and Huawei.
Epoch AI estimates this hardware collectively requires over 10 gigawatts of power - roughly twice what New York City consumes. The figures are based on financial reports and analyst estimates, since exact sales numbers are often not disclosed directly. The dataset is freely available and aims to bring transparency to computing capacity and energy consumption.
Nvidia is reportedly in advanced negotiations to acquire Israeli AI start-up AI21 Labs, according to a report by Calcalist. The deal could be worth between two and three billion dollars. AI21 was valued at $1.4 billion during its last funding round in 2023, with both Google and Nvidia among its investors.
Similar to its recent Groq deal, Nvidia appears primarily interested in AI21's workforce. The company employs around 200 people, most of whom hold advanced degrees and have rare expertise in AI development. At the reported deal value, Nvidia would be paying roughly $10 to $15 million per employee.
AI21, based in Tel Aviv, was founded in 2017 by Professor Amnon Shashua, Professor Yoav Shoham, and Ori Goshen. The company builds large language models and offers software tools, including an agent builder. Its products target enterprise customers, with estimated annual revenue of around $50 million.
Nvidia's China business is ramping up again. According to Reuters, the chip company is in talks with contract manufacturer TSMC to expand production of its H200 AI chips to meet surging demand from China. Chinese tech companies have ordered more than two million H200 chips for 2026, but Nvidia currently only has 700,000 units available.
Production at TSMC is set to begin in the second quarter of 2026, with Nvidia planning to sell the chips for around $27,000 each. ByteDance alone, TikTok's parent company, plans to spend around 100 billion yuan (about $14 billion) on Nvidia chips in 2026, according to the South China Morning Post.
The Trump administration recently cleared H200 chip exports to China with a 25 percent fee attached. However, China hasn't approved the imports yet. Officials there are still debating whether access to advanced foreign chips might slow down the country's domestic chip industry. One proposal under consideration would require every H200 order to include a certain percentage of chips made in China.
China is also working on a rule that would require domestic chip manufacturers to use at least 50 percent locally made equipment.
In case there was any doubt that Nvidia's Groq deal is anything but a takeover in disguise: according to Axios, roughly 90 percent of the workforce—including CEO Jonathan Ross and President Sunny Madra—is moving to Nvidia. Groq will continue as an independent company under new CEO Simon Edwards.
Though officially a non-exclusive license agreement worth around $20 billion, employees and shareholders are walking away with significant payouts. Staff moving to Nvidia get cash for vested shares and Nvidia stock for unvested ones; even those at Groq for less than a year will have their vesting cliff waived for immediate liquidity. Shareholders receive about 85 percent upfront, another 10 percent in mid-2026, and the rest by year's end.
Since 2016, Groq has raised around $3.3 billion from investors including Blackrock, Samsung, and Social Capital. They're now seeing substantial returns, as the deal pushed the startup's valuation from $7 billion to roughly $20 billion. For a more in-depth look at why Nvidia made this move, see my analysis.
Nvidia is spending $20 billion on a holiday gift to itself: a struggling chip startup and its founders. The deal offers both tax advantages and a defense against Google’s TPUs in one move.
Nvidia has released a new base model for gaming agents. NitroGen is an open vision action model trained on 40,000 hours of gameplay videos from more than 1,000 games. The researchers tapped into a previously overlooked resource: YouTube and Twitch videos with visible controller overlays. Using template matching and a fine-tuned SegFormer model, they extracted player inputs directly from these recordings.
NitroGen builds on Nvidia's GR00T N1.5 robotics model. According to the researchers, it's the first model to demonstrate that robotics foundation models can work as universal agents across virtual environments with different physics engines and visual styles. The model handles various genres—action RPGs, platformers, roguelikes, and more. When dropped into unfamiliar games, it achieves up to 52 percent better success rates than models trained from scratch.
The team, which includes researchers from Nvidia, Stanford, Caltech, and other universities, has made the dataset, model weights, paper, and code publicly available.
