phenomenon · first seen 4 Nov 2026
LLM prefill on CPU
Prompt-processing phase of LLM inference on CPUs, often bound by DRAM bandwidth rather than compute.
Mentions per week
PapersCodeDiscussion
Earliest sources
- NUMA-Torch: Compiler-Driven NUMA-Aware LLM Prefill on Multi-Socket CPUsPapers · 4 Nov 2026 · Thisara Rathnayaka Weerakoon, Ishan Dewpura, Niman Ransindu, Shavin Anjitha, Tha…
Latest sources
- NUMA-Torch: Compiler-Driven NUMA-Aware LLM Prefill on Multi-Socket CPUsPapers · 4 Nov 2026 · Thisara Rathnayaka Weerakoon, Ishan Dewpura, Niman Ransindu, Shavin Anjitha, Tha…