Kernel & Perf
NVIDIA Minerva: finding what slows down an AI cluster
An AI job can slow down without a GPU failing outright. Minerva connects infrastructure and workload signals to help operators investigate the dependencies behind that slowdown.
Kernel, Perf & Observability
Kernel & Perf: observing and speeding up an AI stack
eBPF, perf, traces, NUMA, scheduling: diagnosing and optimizing an inference stack at the system level, where the abstractions stop.
4 articles
Articles: Kernel & Perf
Kernel & Perf
Flapping Airplanes: ThunderKittens, NVIDIA and the Megakernel Bet
A GPU can be busy without using all its resources. Follow an inference request to understand ThunderKittens, megakernels and the research behind Flapping Airplanes.
Kernel & Perf
Scion: why Anthropic replaced Git for its agents
An agent finishes a patch, then waits for test machines to fetch the code. Scion provides a starting point for understanding that hidden cost and the infrastructure choices behind it.
Kernel & Perf
FlashAttention-4 explains why faster matrix multiplication is not enough
FlashAttention-4 keeps evolving on Blackwell. Between matrix products, softmax and data movement can determine how fast the kernel progresses.