– Compact, desktop-amiable AI supercomputer for developers and researchers.
– Powers AI models with up to 200 billion parameters.
– includes NVIDIA GB10 Grace Blackwell Superchip with performance of 1 Petaflop and 128GB unified system memory.
– Enables integration of ASIC chiplets onto NVIDIA chips for semi-custom AI infrastructure.
– Collaboration with partners like MediaTek, Marvell, Synopsys, etc., to create modular systems tailored for industries like automotive and manufacturing.
– Next-gen accelerators optimized for advanced AI tasks including large-scale training.
– Features include B300 GPU offering up to 30 PFLOPS compute power and equipped with high-power requirements (>600W).
– Launch of NVIDIA Accelerated Quantum Research Center (NVAQC) in Boston focused on quantum computing challenges.
Reduce energy usage while supporting expansive GPU clusters via silicon photonics-driven architecture.
to explore more additional read throughout articles from near sidebar link cover here followscreenshots summary stages,” < images through" https://www.nextbigfuture.com