CDNA kernel devs say assembly with syntactic sugar may beat the compiler
salykova_ · x · 2026-09-19
@nirw4nna argues the hardest part of writing kernels on CDNA GPUs is instruction scheduling — it's extremely difficult to delimit a region and schedule a balanced mix of VALU + MATRIX ops against a compiler that insists on doing its own thing, so the right programming model may be "an assembler with some syntactic sugar." @salykova picks up the thread, suggesting it may be time to write high-performance AMD GPU kernels directly in assembly, or at least make clean asm easier to write.
More from Infra
- Joseph Suarez hits SOTA in 5 hours on consumer cards in a warehouse — yacineMTB · 2026-09-19
- MiMo RL livestream: infra restarts waste 33% of compute, over $400K lost — dustinvtran · 2026-09-19
- Why Jev-class models could become a near-free judgment primitive running on-device — signulll · 2026-09-19
- Pedro Domingos: Opposing Data Centers Means Keeping Your Country Stupid — pmddomingos · 2026-09-19
- Pedro Domingos: OpenAI and Anthropic's real moat is their massive secured compute — pmddomingos · 2026-09-19
- Fully Local, Private Computer-Use AI Arrives — Runs on a 3060 Ti 12GB — WolframRvnwlf · 2026-09-19