Anthropic just confirmed it is building a custom silicon team to design AI chips, aiming to co-design hardware and models to boost Claude's speed and efficiency. Despite existing compute deals with Nvidia, AWS, and Google, soaring demand is pushing the company toward in-house silicon—following OpenAI's Broadcom-built Jalapeño chip.
SpaceX just unveiled STARMIND, an orbital AI compute satellite design featuring a 75-meter wingspan and up to 250 kW peak compute power. By tapping continuous solar energy in sun-synchronous orbit and radiating heat into deep space, the project aims to bypass terrestrial grid and cooling limits, targeting mass deployment via Starship as early as late 2027.
Microsoft just unveiled its in-house MAI model lineup, rolling out over a dozen task-specific models that slash GPU costs by 50% to 89% across tools like GitHub Copilot and PowerPoint. Its new MAI-Cyber-1-Flash topped the CyberGym benchmark at half the cost of prior options. Why it matters: Big tech is shifting away from running mega-models on everyday tasks, proving specialized smaller models can match frontier quality at a fraction of the budget.
Ai2 just launched the OlmoEarth Platform, an infrastructure stack built to run Earth observation foundation models at planetary scale. It can process continent-sized satellite datasets in roughly a day for fractions of a penny per square kilometer. In a test mapping North America wildfire risk, it coordinated 19,600 CPUs and 994 GPUs to deliver a 155x speedup.
Anthropic CEO Dario Amodei just clarified that the company has never advocated for banning open-weights AI models, rejecting accusations of corporate protectionism. Instead of blunt model bans, Amodei wants Washington focused on strict chip export controls, stopping industrial-scale model distillation, and requiring mandatory pre-release safety testing across both open and closed frontier models.
Nvidia just forged a massive multibillion-dollar partnership with Safe Superintelligence, the stealth lab founded by former OpenAI chief scientist Ilya Sutskever. The deal provides SSI with Nvidia's Vera Rubin GPU platform, boosting its compute power by an order of magnitude and phasing out Google TPUs. Why it matters: Nvidia is securing long-term loyalty from the most elite AI safety researchers, while Ilya bets that brute-force GPU scaling is still the true path forward.
NVIDIA CEO Jensen Huang just dropped his first post on X, sharing a signed letter advocating for open-weight AI models. Huang argued that open models strengthen cybersecurity, accelerate innovation, and support national sovereignty, emphasizing that the tech landscape requires both closed and open frontier models to thrive together.
Lunar Outpost just announced its next moon rover will use Nvidia Jetson chips to run its lidar system, potentially putting the first GPU on the lunar surface. Slated to launch on a Falcon 9 rocket before the end of the year, the mission will test if commercial physical AI chips can survive cosmic radiation and freezing lunar nights on minimal power.
AMD just unveiled its 6th Gen EPYC 9006 server CPUs built on TSMC's 2nm Zen 6 architecture, packing up to 256 cores and 512 threads per socket. AMD claims the chips deliver a 174% geomean performance leap over Intel Xeon 6980P in agentic AI pipelines like orchestration and database retrieval.
AMD just launched Helios, a massive rack-scale AI platform packing 72 Instinct MI455X GPUs to rival Nvidia's Vera Rubin NVL72. The system offers 2.9 exaflops of FP4 compute and 31 TB of HBM4, with AMD claiming a 15% edge in compute and 50% more memory capacity. Heavyweights like Microsoft, OpenAI, and Anthropic have already signed on.
AMD and Cerebras just teamed up on a disaggregated AI inference solution, pairing AMD Helios rackscale systems with the Cerebras Wafer-Scale Engine. AMD tackles heavy prompt processing while Cerebras handles ultra-fast token decoding, promising up to 5x higher tokens per second per watt. It is built for real-time agentic workflows and launches on Cerebras Cloud in H2 2026.
The Trump administration just accused China's Moonshot AI of secretly distilling Anthropic's top-tier Fable model to build its new 2.8-trillion parameter K3 system, while allegedly tapping Nvidia GB300 chips in Thailand. The Treasury is now weighing trade blacklists and sanctions, though Chinese officials firmly denied the allegations.
Microsoft just pledged $60 million to the Department of Energy's Genesis Mission, aiming to double U.S. research productivity over the next decade. The commitment provides $40 million in Azure compute, $20 million in engineering services, and launches a hub called SPARK to connect 17 National Laboratories.
Nearly 200 US utility companies and data center operators just signed Donald Trump's voluntary rate payer protection pledge, covering about 80% of power delivered across the country. The agreement aims to prevent consumers from footing the bill for surging AI data center energy use, but it includes zero enforcement mechanisms or penalties, making it little more than a pinky promise.
OpenAI just committed $17 million in resources to the U.S. Department of Energy’s Genesis Mission, aiming to double national research productivity within a decade. The pledge includes $4 million in Codex access for 2,000 scientists, targeted support for high-temperature superconductors, and deploying reasoning models on the Venado supercomputer. Why it matters: OpenAI is embedding frontier models directly into federal scientific infrastructure to compress decades of research into years.
Google is reportedly developing a custom server chip called "Frozen v2" that physically etches Gemini model architecture into silicon to cut data movement. Slated for 2028, it could deliver 6 to 10 times more tokens per watt than current TPUs, targeting a cloud compute crunch that has forced Google to turn away customers.
New York just became the first US state to enact a statewide data center moratorium, freezing new environmental permits for facilities over 50 megawatts for up to a year. The pause gives regulators time to study surging power bills and resource impacts fueled by AI demand. Meanwhile, an even stricter 20-megawatt legislative bill still awaits the governor's signature.
Sam Altman just called out Elon Musk for pitching "short-term space datacenters" to investors, and experts agree with him. Space startups and Google's orbital compute team note that orbital AI processing won't scale without cheap, truly reusable rockets and low-cost satellites. While SpaceX plans early flights next year, true commercial scale remains a story for the 2030s.
Sunrun just launched a pilot to build a distributed AI network by paying solar and battery customers to host mini compute nodes at home. The aggregated power will then be sold to enterprise AI buyers. With over 70% of Americans pushing back against massive new data centers, turning private homes into decentralized server rooms offers a clever workaround.
Anthropic just partnered with engineering giant UST to deploy Claude across 20,000 engineers and consultants worldwide. UST is embedding Claude Code directly into its chip validation pipeline to read schematics and auto-generate tests, slashing cycle times by up to 70% while rolling out similar workflows across healthcare, telecom, and banking.
