Rik Kisnah - Blog

#Infrastructure

The 2026 GPU Field Guide: H200, B200, B300, GB200, GB300, Vera Rubin, MI3xx and Helios, From the Chip to the Rack

The 2026 GPU Field Guide: H200, B200, B300, GB200, GB300, Vera Rubin, MI3xx and Helios, From the Chip to the Rack

RDMA, RoCEv2 and InfiniBand, Explained to a Ten-Year-Old (and to the New Grad on My Team)

RDMA, RoCEv2 and InfiniBand, Explained to a Ten-Year-Old (and to the New Grad on My Team)

GPU Burn-In at Scale: Commands, Gates, and Troubleshooting for NVIDIA and AMD Clusters

GPU Burn-In at Scale: Commands, Gates, and Troubleshooting for NVIDIA and AMD Clusters

Build the Data Centers

Build the Data Centers

Design Model Weight Distribution to a Thousand Hosts

Design Model Weight Distribution to a Thousand Hosts

The AI Pyramid: Five Layers Between Hardware and AGI

The AI Pyramid: Five Layers Between Hardware and AGI

The Complete NCCL Reference Guide: Commands, Errors, and Troubleshooting for OCI GPU Infrastructure

The Complete NCCL Reference Guide: Commands, Errors, and Troubleshooting for OCI GPU Infrastructure

From First Principles to Zettascale: How OCI's GPU/RDMA Architecture Redefines AI Infrastructure

From First Principles to Zettascale: How OCI's GPU/RDMA Architecture Redefines AI Infrastructure

Oracle AI World 2025: Enterprise AI from Promise to Production

Oracle AI World 2025: Enterprise AI from Promise to Production

Three Weeks in Batam: Bringing NVIDIA GB200 to Life on the Data Plane

Three Weeks in Batam: Bringing NVIDIA GB200 to Life on the Data Plane