←Search results

Search “NVIDIA”

By relevance · Total 2010 results Clear search

Wednesday

  4. Data Center Knowledge

    Drastic AI Power Fluctuations Force a Co-Design Revolution from Chips to the Power Grid

    AI data center power can surge and plunge within seconds or even milliseconds, and a sudden drop of hundreds of megawatts could cause wide-area grid disturbances; the industry is calling for chipmakers, operators, utilities and infrastructure parties to respond collaboratively. Nvidia electrical design engineer Sai Somayajula said rack power has risen from about 40 kW in the H100 era to about 150 kW with Blackwell, with Vera Rubin targeting about 240 kW, and under future 800V DC power distribution it could reach 1 MW per rack; Nvidia has added rack-level energy storage and control and introduced the DSX architecture. Oracle's Stargate data center campus is already 60% to 70% operational, with plans for gigawatt-scale capacity.

    hundreds of megawatts could cause wide-area grid disturbances; the industry is calling for chipmakers, operators, utilities and infrastructure parties to respond collaboratively. Nvidia electrical design engineer Sai Somayajula said rack power has risen from about 40 kW in the H100 era to about 150 kW with Blackwell, with Vera Rubin targeting about 240 kW, and under future 800V DC power distribution it could reach 1 MW per rack; Nvidia has added rack-level energy storage and control and introdu

    Matched fieldsSummary、Body

    #hot-signal#Industry news#English text is being prepared.#English text is being prepared.

Tuesday

  13. UCSD Hao AI LabSelected

    UCSD Hao AI Lab open-sources Minecraft agent Jev, prompt evolution without training

    UCSD Hao AI Lab has open-sourced the Minecraft agent Jev, publishing how it works after 1,795 matches. It does not train a model; instead, a SOTA LLM agent evolves prompts from the bot's own logs under human supervision; in Block UHC it won 70% of rounds against human players. Each decision takes 24 ms on NVIDIA Blackwell, which the team says is 2.5 times faster than a mainstream inference engine, and it has open-sourced the inference engine DJev-serve, completing each decision in a single CUDA graph on GB200 NVL72 and B200.

    ed the Minecraft agent Jev, publishing how it works after 1,795 matches. It does not train a model; instead, a SOTA LLM agent evolves prompts from the bot's own logs under human supervision; in Block UHC it won 70% of rounds against human players. Each decision takes 24 ms on NVIDIA Blackwell, which the team says is 2.5 times faster than a mainstream inference engine, and it has open-sourced the inference engine DJev-serve, completing each decision in a single CUDA graph on GB200 NVL72 and B200.

    Matched fieldsSummary

    #lab#English text is being prepared.#Agents#English text is being prepared.