Search AI Products and News

Nvidia has unveiled its next-generation AI chip, Rubin, promising a performance leap of up to 900 times that of its current flagship Hopper architecture. This significant advancement points towards a new era in AI processing capabilities.

NVIDIA showcased its new supercomputer Hopper in the MLPerf benchmark tests. Hopper is equipped with over 10,000 powerful H100 GPUs. Compared to Intel's Gaudi2, Hopper delivers 4 times the performance per dollar. Few vendors besides NVIDIA submitted results for MLPerf. While the MLPerf training benchmarks are increasingly losing comparative significance, Hopper still demonstrates superior computing power.

{Point 1: NVIDIA launched the GH200 Grace Hopper Superchip platform to meet the growing demand for generative AI. Point 2: NVIDIA introduced NVIDIA AI Enterprise 4.0, providing the tools needed for generative AI. Point 3: Huang Renxun announced NVIDIA's new products and updates, including AI Workbench, OpenUSD, and more.}