Ricursive Intelligence, founded by Anna Goldie and Azalia Mirhoseini, is building AI systems that design chips, learn from each iteration and improve the next design. Speaking at TechCrunch Disrupt, the founders said demand for compute could rise 10,000-fold in five years and that the answer is more efficient silicon rather than more data centres. Their approach optimises whole chips instead of partitioning them into dozens of blocks, a method that produced the unexpected donut-shaped layouts of their earlier AlphaChip work. They argue that cutting design time from years to a week would make custom chips viable for space, AR, VR and biological computing.
The memory shortage that started in AI data centres has reached consumer shelves. Samsung, SK Hynix and Micron have shifted production towards high-bandwidth memory for AI accelerators, and 32GB DDR5 kits that sold for under $100 now trade above $400 in many markets. Gartner forecast DRAM prices would rise 47% in 2026, and SK Hynix says its HBM, DRAM and NAND capacity is essentially sold out for the year. Graphics card prices have also climbed as AI buyers take cards in bulk.
TSMC reported third-quarter revenue of US$46.72 billion, up 41.2 percent from a year earlier and above its own guidance, with net income of NT$763.06 billion, a sixth straight record quarter. High-performance computing, the bucket for AI accelerators, grew 22 percent from the previous quarter to 68 percent of revenue, while consumer electronics fell to 1 percent. Chief executive C.C. Wei said AI demand remains far above TSMC's ability to supply. The 2nm node is already 8 percent of wafer revenue, and the company guided fourth-quarter revenue to US$52.3-54.3 billion.
Microsoft introduced Azure Maia 300 at the OCP Global Summit in San Jose, its third-generation custom AI accelerator. The company says the chip delivers up to 40% more tokens per megawatt than leading GPU-based systems, and that a 96-accelerator rack offers seven times the rack-scale throughput of Maia 200. Production deployments begin in 2027, alongside multi-gigawatt Azure Maia buildouts. Microsoft now calls Maia the primary AI inference system for its own products and services, a shift from treating custom silicon as a supplement to Nvidia and AMD hardware.
Google launched a prototype satellite on a SpaceX Falcon 9 from Vandenberg on 1 October, carrying four of its own Tensor Processing Units into orbit. The mission is part of Project Suncatcher, a first step towards running data centres in space. The chips run a version of Google's Gemma model for about 15 minutes at a time because of heat limits, and the test will measure how they handle radiation and the stress of spaceflight. Google plans two more satellites in 2027 to test laser links. Google calls the mission a very minimal test. SpaceX and the startup Starcloud are pursuing similar orbital computing projects.
Character Technologies, the company behind the Character.ai chatbot platform, filed for Chapter 11 bankruptcy in Delaware on Oct. 12, listing estimated assets and liabilities of between $50 million and $100 million each. Disney agreed before the filing to act as stalking horse bidder, offering $20 million for a non-exclusive perpetual license to certain intellectual property plus transition services from former Character.ai employees it has hired. The platform is meant to keep running during the sale process. The company has faced lawsuits claiming its app contributed to mental health problems and suicides among teenagers.
Google has confirmed that its first Project Suncatcher satellite is operating as expected, weeks after launching on a SpaceX Falcon 9 from California on 1 October. The spacecraft carries four Trillium tensor processing units, the same family of AI chips that power Google's Gemini models, and was built with satellite company Planet. Google says an orbiting solar panel can generate several times the energy of the same panel on Earth. The chips survived launch forces of 50 to 100 times Earth's gravity and proton-beam radiation testing at UC Davis with no hard failures. Two satellites are to test laser links in 2027.
Companies racing to fund AI chips have sold about $489 billion of related bonds so far in 2026, already more than in all of 2025, according to credit strategist Amanda Lynam. Broadcom is arranging more than $50 billion to finance OpenAI's custom accelerator, while SpaceX is in talks over a $40 billion package to buy Nvidia processors. Credit markets are responding: Broadcom's five-year credit default swap spread has more than tripled from about 41 basis points to roughly 136, and Oracle's has reached a record near 261. Oracle's shares are down more than 40% since June 2 and Broadcom's about 25%, while Nvidia's stock has held up, trading at $237.47.
TSMC is reported to be considering a second US campus, this time in Dallas, according to UDN and Reuters reports citing unnamed sources. The company has already pledged $265 billion for American manufacturing, and Taiwan's Ministry of Economic Affairs says it has approved $44 billion of TSMC investment in the United States so far. The first Arizona fab began high-volume production in 2024, with three more under construction and plans for more than nine facilities in the state. Announced US private-sector manufacturing commitments have now passed $2.09 trillion, led by semiconductor projects. AI data centres are driving demand, and advanced packaging capacity remains tight.
Semiconductor shares climbed on Friday as investors looked past a report that OpenAI revenue may fall short of earlier estimates. Broadcom, Micron, Marvell, AMD and Lam Research all rose, and futures for the Nasdaq 100 gained about 1 percent. The S&P 500 closed above 7,800 for the first time earlier in the week. Analysts expect S&P 500 earnings to grow about 30 percent for the July-to-September quarter, led by energy and technology. Chipmakers sit at the centre of the AI buildout, and demand for advanced processors and high-bandwidth memory remains the main driver. Traders expect the Federal Reserve to hold rates steady at its October meeting.