Jaewoong Jung

PLINK: A GPU-Initiated I/O Platform Exploiting NVMe Parallelism

As LLM inference scales and KV-cache pressure forces data onto NVMe storage, the frequency and fine-grained nature of GPU-initiated storage accesses have grown substantially. Existing GPU-initiated I/O platforms do not fully exploit NVMe’s multi-queue parallelism because their serialized submit-then-poll path keeps onl...