AI article
I read a CUDA inference engine's symbols with Ghidra, then made it 12% faster on my 3090
Community description: I run one job at home that never sleeps, a coder model quantized to IQ1_M on an RTX 3090, and my...
Dev.to | Oct 16, 2026 | prodbymarcu
Automated excerpt
Every switch threw away whatever that session had already read. Across sessions nothing parked, so every switch pays a full reread. Flat on IQ1_M. --prefill-head-alias, needs short reads off, every prompt read pays 0. 3-0. 6s of head reload.
Selected automatically from source text; not independently written or fact-checked. Read the original for full context.