These 10 GPU optimization papers are why your model runs fast today
NandoDF · x · 2026-11-04
A curated thread walks through 10 GPU optimization papers that shaped today's model speed, from FlashAttention's IO-aware exact attention (3x GPT-2 speedup) and FlashAttention-2's work partitioning to Stream-K's balanced GEMM scheduling. The diagrams show SRAM/HBM/DRAM bandwidth gaps and TFLOPs comparisons across implementations.
More from Infra
- Sam Altman? No—Bezos Calls for Nuclear-Powered Offshore Data Center Platforms — beffjezos · 2026-11-04
- NEURA launches Neuron AI, but Europe's Physical AI test is whether the stack matures together — ingliguori · 2026-11-04
- S3-based databases cache in memory and NVMe, making disk vs S3 latency gap smaller than expected — mgill25 · 2026-11-04
- gpu-unlock project claims up to 2x FP16 tensor throughput for RTX 20/30/40 cards — paranoidray · 2026-11-04
- One prompt to audit cross-region traffic: dev cut trans-Atlantic latency by moving a service — DanielLockyer · 2026-11-04
- SurrealDB 4 launches as multi-model ACID database for the agentic era — Scobleizer · 2026-11-04