NVIDIA is making AI-predicted protein complex structures for more than 2,800 viruses openly available with GoogleDeepMind, emblebi, and research partners. NVIDIA says this gives scientists a head start in preparing for potential outbreaks.
Jiantao Jiao says an NVIDIA team built SWE-Serve to test whether AI agents can develop an inference engine and serve real models. The benchmark has 53 tasks drawn from real sgl_project engineering work.
A Codex planner and creator turn each skill into tested terminal workflows, using a rubric taken from that skill's own quality criteria. Outcome-only RL moved Qwen3.8-27B from 49.4% to 54.1% on Terminal-Bench 2.1, while adding the rubric reached 50.1% but won more trajectory comparisons.
NVIDIA AI reports a 14.72% error rate for Nemotron 3 Diarization on voicearena_ai’s initial Diarization-Bench, about 24% lower than the runner-up.
The Runway AI Summit is one week after September 23, 2026. The announced session is Grounding Intelligence in the Physical World, with Ming-Yu Liu, Jon Barron, and Alex Toshev.
In ComfyUI, encoding is up to about 2.2 times faster and decoding is about 1.4–2.7 times faster. ComfyUI reports a 1344×768, 129-frame round trip falling from 24.3 seconds to 12.7 seconds.
Sluicebox’s AI supplier agent Lucy uses NVIDIA Nemotron 3 Ultra to collect and check supplier data while engineers design a data center’s carbon footprint. In Sluicebox’s tests, Nemotron improved accuracy, cut costs by 51–80%, and returned responses up to 2x faster.
NVIDIA said Grok 4.7 has landed and congratulated @SpaceXAI on what it called the team’s most capable model yet for coding and knowledge work. In the quoted post, Elon Musk described Grok 4.7 as a strong combination of intelligence, speed, and low cost.
NVIDIA quoted Einride’s announcement of a strategic collaboration with NVIDIA DRIVE to build the next generation of Einride Driver on the NVIDIA Hyperion platform, targeting autonomous trucking on highways and suburban roads.
