Advancements in semiconductor technology, GPU development, and data center infrastructure that power AI computing capabilities.
Microsoft has launched the Maia 200, an AI inference accelerator built on TSMC’s 3-nanometer process, designed to improve performance and efficiency for large-scale AI workloads. The chip delivers up to 10 petaflops in 4-bit precision and powers models including GPT-5.2 and Microsoft Copilot.
Samsung Electronics is set to start mass production of its next-generation HBM4 memory chips next month, with shipments expected to go to Nvidia after completing final qualification stages.
Intel expects first-quarter revenue and profit to fall below analyst estimates as it struggles to meet demand for server chips used in AI data centers, leading to a 13% drop in its share price.
Nvidia CEO Jensen Huang said the global expansion of artificial intelligence is fueling what he described as the largest infrastructure buildout in human history, projecting $85 trillion in investments over the next 15 years.
Austin-based Neurophos has raised $110 million in a Series A round led by Gates Frontier to develop its photonic AI chips, which use micron-scale metamaterial optical modulators to boost performance and energy efficiency.
Portland-based AheadComputing has raised $30 million in a Seed2 round led by Eclipse, Toyota Ventures, and Cambium to accelerate development of its RISC-V-based CPU architecture designed for AI applications.
BOS Semiconductors has licensed Ceva’s SensPro AI DSP architecture for its Eagle-A automotive SoC, designed to enhance real-time perception and sensor fusion in advanced driver assistance systems.
Jon Peddie Research has released an extensive report analyzing the rapidly expanding AI processor market, highlighting over $28.5 billion in investments and activity from 120 startups.
OpenAI has entered a multi-year agreement worth over $10 billion with AI chipmaker Cerebras Systems to secure 750 megawatts of computing capacity through 2028. The partnership aims to enhance the speed and efficiency of OpenAI’s inference workloads, including ChatGPT.
OpenAI and SoftBank Group are investing $1 billion in SB Energy to develop large-scale AI data centers and energy infrastructure in the U.S., including a 1.2 GW facility in Texas set to begin operations in 2026.
Meta has signed agreements with TerraPower, Oklo, and Vistra to supply more than 6 gigawatts of nuclear power for its Prometheus AI data center in Ohio, securing long-term energy for its growing AI operations.
Supermicro has announced expanded production and liquid-cooling capabilities to support NVIDIA's upcoming Vera Rubin and Rubin platforms, aiming to accelerate delivery of next-generation AI data center solutions.
Tenstorrent unveiled a compact AI accelerator device developed with Razer at CES 2026, designed to bring generative AI capabilities to Thunderbolt-equipped laptops for mobile and scalable edge AI development.
Nvidia has launched its new Rubin platform, a six-chip architecture designed to power next-generation AI supercomputers. The system introduces major advances in compute efficiency, networking, and storage, and will be deployed by major cloud providers starting in the second half of 2026.
Nvidia has reached a non-exclusive licensing deal with AI chip startup Groq valued at around $20 billion. The agreement includes licensing Groq’s inference technology and hiring key executives, including founder Jonathan Ross and president Sunny Madra, while Groq continues operating independently.
Alphabet has agreed to acquire Intersect, a data center and energy infrastructure developer, for $4.75 billion in cash plus assumed debt. The deal will help Google expand data center power generation capacity and is expected to close in the first half of 2026.
Cerebras Systems is preparing to file for a U.S. IPO next week, targeting a second-quarter 2026 listing. The move follows the company's clearance from U.S. regulators and the removal of G42 from its investor list after previous delays.
Hut 8 has announced a partnership with Anthropic and Fluidstack to develop up to 2.3GW of AI data center capacity in the United States, beginning with a 245MW facility in Louisiana.
OpenAI is in early discussions with Amazon about a potential investment exceeding $10 billion and an agreement to use Amazon’s AI chips, following the ChatGPT maker’s recent restructuring and expanded partnership flexibility.
Intel is negotiating to acquire AI chipmaker SambaNova Systems for about $1.6 billion including debt, with a potential deal expected as soon as next month. SambaNova, founded by Stanford professors, designs custom AI chips and was valued at $5 billion in 2021.
Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.
