RDMA

7 posts · 4 sources

NVIDIABlog6d

How DOCA GPUNetIO Unifies GPU-Initiated Networking Across the NVIDIA Software Stack

ModalBlog1w

Modal Clusters are generally available

NVIDIABlog1w

Expanding AI Storage Access with NVIDIA cuObject and the NVIDIA SCADA Server SDK

MetaBlog24 Aug

MetaRoCE: A New RDMA Transport Built for AI-Scale Ethernet

MetaBlog24 Aug

MTIA 300: Meta’s First Training Chip with Built-in NICs and Communication-Offloading Engines

BasetenBlog25 Jun

Live draft model training for speculative decoding

ModalBlog1 Jun

Reinforcement learning is an infrastructure problem