#ai-chips

#ai-chips

Four generations, MTIA 300, 400, 450, and 500, have been produced within less than two years, with several already in production and others scheduled for mass deployment in 2026 and 2027. The quick pace is deliberate. Rather than betting on a single chip generation and waiting years for results, Meta has adopted a roughly six-month cadence per generation, using modular chiplet architecture to enable incremental upgrades without replacing entire rack systems.

An international research effort is mobilising to develop chips with the potential to reduce the energy consumption of artificial intelligence (AI) datacentres by using light instead of electricity to process information. The work is being pioneered by a team of scientists from Europe and South Korea, working together as the Haetae consortium, who have already received €1.49m in funding from the European Union (EU) to financially support their work.

Driving the news: Meta will buy millions of chips from Nvidia, ranging from standalone Grace CPUs to next-gen Blackwell GPUs and upcoming Vera Rubin systems, to use across its U.S. data center buildout. This makes Meta the first Big Tech firm to commit to buying standalone central processing units from Nvidia, which are used to run AI rather than train AI. This signals a shift toward inference over training, with the latter typically requiring more intense and expensive general processing units, or GPUs

UK semiconductor start-up Fractile has announced a £100 million expansion of its British operations, scaling up in London and Bristol as ministers intensify calls for greater domestic ownership of critical artificial intelligence technology. The investment, to be deployed over the next three years, will fund a new industrial hardware engineering facility in Bristol, alongside the expansion of Fractile's existing UK sites and a significant increase in its domestic workforce. The company is focused on developing AI chips optimised for inference, the stage at which large language models generate outputs, an area of growing strategic importance as demand for real-time AI applications accelerates.

And now, Nvidia is the largest suppler by a long shot of chips for artificial intelligence. Its GPUs have proven to be the best option for training and running AI models. Developers from OpenAI to Google, Anthropic, Meta and Amazon, along with a host of specialized startups, are fighting to get their hands on Nvidia chips. That high demand has made the manufacturer the most valuable company in the world.

The family solved the problem last year, The Wall Street Journal has discovered, by using the simplest solution that businesses can now employ for a policy obstacle: They seem to have made a deal with the Trump family. Sheikh Tahnoon bin Zayed Al Nahyan, a member of the Abu Dhabi royal family sometimes called the "Spy Sheikh," purchased a 49 percent share in World Liberty Financial, the Trump family's crypto firm, thus sending $187 million to Trump-family-controlled entities.

MediaTek's share price has risen sharply in a short period of time, partly due to increasing attention to the company's role in Google's AI strategy. In two trading days, the share price rose by approximately 19 percent. This brought the Taiwanese chip designer to a new record high on the Taipei stock exchange. The jump in share price follows a period of sustained optimism among investors.

CEO Jensen Huang showed off Cosmos, an AI foundation model trained on massive datasets, capable of simulating environments governed by actual physics. He also announced Alpamayo, an AI model specifically designed for autonomous driving. Huang revealed that Nvidia's next generation AI superchip platform, dubbed Vera Rubin, is in full production, and that Nvidia has a new partnership with Siemens. All of this shows Nvidia is going to fight increased competition to retain its reputation as the backbone of the AI industry.