Nvidia H100 GPU

Evidence (4)

• 6Hugging Face Blog·10/9/2026framework_update
  Ai2 redesigns GPU cluster scheduling for higher-impact research

  Ai2 describes replacing a priority-based GPU scheduler with a system built around GPU time budgets, hierarchical fair-share allocation, and a time-slicing contract. The change is intended to improve how scarce GPU capacity is assigned across research teams while keeping clusters fully occupied and reducing operational issues such as squatting and priority inflation.

• 8Ars Technica AI·10/2/2026security
  US arrests CEO accused of smuggling Nvidia chips to China

  US authorities arrested Earthmade Computer CEO Greg Lui on आरोप of using falsified shipping paperwork to divert more than $300 million in servers containing export-controlled Nvidia chips into China. The indictment says the shipments included Nvidia A100 and H100 GPUs, which are used for large-scale data processing and AI model training, raising national security concerns.

• 7Hacker News·8/21/2026research
  Qwen3-TTS implementation reaches sub-50 ms voice latency

  The authors report a Qwen3-TTS 1.7B CustomVoice implementation that achieves sub-50 ms p95 time-to-first-audio and about 10 requests per second on a single NVIDIA H100 SXM while maintaining real-time playback. They benchmarked multiple serving engines, tuned them for low-latency streaming, and released the implementation and benchmark as open source.

• 6Hacker News·8/12/2026product_launch
  Launch HN: Discovered Materials uses AI agents to find new semiconductor materials

  Discovered Materials says it is building AI agents to discover and evaluate new materials for semiconductor applications, especially to address thermal challenges in high-power GPUs and advanced packaging. The company claims frontier models can identify promising materials computationally, and it is releasing a benchmark plus discovered materials research while exploring IP licensing and tooling sales as its business model.