NVIDIA's SANA family runs 20× smaller and 100× faster than Flux-12B, generating 1024px images in 0.1s on H100 with full Diffusers, ComfyUI, and SGLang integration on day one.
A 32,000-GPU-hour benchmark confirms the harness layer outweighs model choice — identical backbones swing 3× in accuracy depending on agent framework.
Anthropic's '2028' essay on US-China AI competition fractures the industry — Nvidia, HuggingFace, and OpenAI each offer incompatible counter-doctrines.
Jensen Huang's foundation bought $108.3M in CoreWeave compute for donation to universities and nonprofits. Largest single AI compute philanthropic donation on record.
US clears ~10 Chinese companies for Nvidia H200 chips including Alibaba, Tencent, ByteDance, JD.com. Approvals granted but no deliveries made yet.
NVIDIA's Jim Fan debuts World Action Models (WAM) and DreamDojo as the core architecture for physical AGI at Sequoia AI Ascent 2026.
Jensen Huang: 'AI became useful in the last several months.' Circulating as implicit concession that trillions in AI investment preceded demonstrated utility by years. GPU prices still rising.
NVIDIA B200 spot rate doubled from $2.31 to $4.95/hr in six weeks. Blackwell supply squeeze widens the price gap vs prior-gen H200 chips.
Six concurrent signals in 24 hours mark a structural AI capital inflection: Cerebras $40B IPO, Anthropic's PE joint venture, EU compute criticism, and lenders quietly exiting AI debt.
NVIDIA open-releases Nemotron 3 Nano Omni (30B MoE/3B active): unified video/audio/image/text model with 9× video-reasoning capacity improvement vs. predecessors.
Huawei projects $12B AI chip revenue in 2026 (+60%), driven by Ascend 950PR orders as Nvidia loses ground in China to export controls.
Matthew Berman argues DeepSeek confirms US open-source AI has no business model without government subsidy; proposes compute quotas and sovereign procurement as policy fixes.
DeepSeek v4 reignites debate on US open-source AI: Berman argues the business model is broken, leaving Nvidia as the only credible US champion.
Nvidia's Nemotron 3 Nano Omni is a 30B-A3B MoE open multimodal model built for agentic AI, with vision and speech capabilities.
NVIDIA's multi-agent LLMs self-evolve the ABC chip design toolchain, beating decades of human-tuned baselines — the first production self-improvement system on real semiconductor infrastructure.
NVIDIA Ising uses AI to shrink quantum computer calibration from days to hours and provides 3D neural network error correction faster than existing open-source methods.
NVIDIA's ~100-model unified API catalog at build.nvidia.com lets developers prototype across text, image, audio, video, and agentic workflows without owning infrastructure.
DeepSeek V4's 10× KV-cache compression restructures AI cost economics globally, exposing a structural threat to US lab pricing and strategic positioning.
OpenAI and NVIDIA piloted a whole-company Codex rollout, and OpenAI is now inviting other enterprises to adopt the same all-employee model.
Commerce Secretary Lutnick confirms Nvidia's H200 chips have not been sold to Chinese companies, with Chinese government approval also absent — a direct export-control reaffirmation.
Curated AI insights — sent when there's something worth your inbox.