Khawaja Shams shares a distributed systems perspective on scaling LLM inference stacks efficiently. He discusses practical strategies for optimizing GPU utilization, managing KV cache memory cliffs, ...

Genius Group Board approves the incorporation of GENIUSpark to build, own and operate community-first AI data centers through joint ventures with AI companies, data center operators and power partners ...

We asked our Fast Company Executive Board members how they have built a professional community. Here, 35 of them share what ...

Key TakeawaysElon Musk says Terafab, the chip fab Tesla and SpaceX are building in Texas, will rank in the top 20 of country GDP from a single building, as it reportedly courts about 10 Korean ...

Oracle is committing $1 billion over four years to expand partner incentives, AI and cloud resources, migration support and ...

For 35 years, those partnerships have been central to ANDP’s accomplishments. The anniversary celebration offered an opportunity to recognize the people who built that foundation, celebrate those ...

Kimi K3.1 Pro ranked fourth among 215 AI models on an independent intelligence index while Moonshot AI deliberately removed ...

IQuest-Q1 AI model from Beijing lab IQuest Research is the largest known Muon optimizer deployment at 320 billion parameters, ...

Microsoft launched Decision-1 on Oct. 9, 2026, a specialist model built on Alibaba’s Qwen3.5-9B that returns calibrated ...

Nanbeige4.3-33B, a new 33B open-weight model, beats Qwen3.8-27B on key coding benchmarks as H100 rental prices climb nearly 40% in six months.

Cerebras rides Nvidia's H100 price spike and CoreWeave's debt-driven selloff as wafer-scale chips challenge GPU dominance in ...

Recursive Superintelligence released the open-weight model Flint-40B-A2B on October 21, a mixture-of-experts model with roughly 39.8 billion ...