Meta’s all-new AI Research SuperCluster (RSC) is touted as one of the world’s fastest AI supercomputers. Even in its unfinished form, it’s currently performing tasks like translating text between languages and identifying potentially harmful content. Hardware wise, the supercomputer is comprised of 760 NVIDIA DGX A100 systems as its compute nodes, for a combined total of 6,080 GPUs — with each A100 GPU being more powerful than the V100 used in our previous system.
Each of the NVIDIA DGX A100 systems communicate through an NVIDIA Quantum 1600 Gb/s InfiniBand two-level Clos fabric that has no oversubscription. The AI Research SuperCluster’s storage tier boasts 175 petabytes of Pure Storage FlashArray, 46 petabytes of cache storage in Penguin Computing Altus systems, and 10 petabytes of Pure Storage FlashBlade. Compared to Meta’s legacy production and research infrastructure, the RSC is capable of running computer vision workflows up to 20 times faster, the NVIDIA Collective Communication Library (NCCL) more than nine times faster, and trains large-scale NLP models three times faster.
- Aspect Ratio:16:9.Connectivity Technology: HDMI,USB,Ethernet,WiFi,Bluetooth
- Dominate the Game: With the 10th Gen Intel Core i5-10300H processor, your Nitro 5 is packed with incredible power for all your games
- RTX, It's On: The latest NVIDIA GeForce RTX 3050 (4GB dedicated GDDR6 VRAM) is powered by award-winning architecture with new Ray Tracing Cores,...
We expect such a step function change in compute capability to enable us not only to create more accurate AI models for our existing services, but also to enable completely new user experiences, especially in the metaverse. Our long-term investments in self-supervised learning and in building next-generation AI infrastructure with RSC are helping us create the foundational technologies that will power the metaverse and advance the broader AI community as well,” said Kevin Lee, Technical Program Manager.