Stanford CS336 Lecture: GPU Kernel Optimization Centers on Memory
Stanford CS336's latest lecture by Percy Liang explains GPU kernel optimization, arguing that memory traffic rather than tensor expression length is the real performance bottleneck.
2026-10-21 ~ 2026-10-21 · 2 related posts
- Stanford CS336 explains GPU kernels: data movement, not math, is what costs you — stanfordnlp · 2026-10-21
- Stanford CS336 Lecture 6: fused Triton kernels cut softmax memory traffic ~4x — stanfordnlp · 2026-10-21