← 研踪 PaperTrailarXiv:2611.18145 · v1
Compiling Specialized GPU-Initiated Networking
首次提交 2026-11-13 21:19
版本更新 2026-11-13 21:19
雷达观测 2026-11-16 11:41
摘要
Machine learning systems specialize computation to the workload and hardware, but communication remains in generic libraries whose device code and transport decisions are not visible to high-level compilers. We present Obaleth, a compiler that makes GPU-initiated communication part of a kernel's intermediate representation, co-optimizes it alongside computation, and lowers it to specialized transport code. Obaleth plans completions around buffer lifetimes and optimizes memory ordering, synchronization, and batching across NVIDIA and AMD GPUs. On H100 GPUs, Obaleth sends small messages up to 3.3 faster, and on 256 GPUs runs all-to-all 1.25 and 2.3 faster than NVSHMEM and NCCL GIN, with much smaller memory and register use. In vLLM, it serves Qwen3-235B on 16 B200 GPUs and answers requests up to 1.29 faster than with DeepEP.
作者备注:20 pages, 13 figures, 6 tables. Equal contribution from first two authors. Code: https://github.com/ParCoreLab/aboleth
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-13 21:19 · 最新观测版本
Compiling Specialized GPU-Initiated Networking
Machine learning systems specialize computation to the workload and hardware, but communication remains in generic libraries whose device code and transport decisions are not visible to high-level compilers. We present Obaleth, a compiler that makes GPU-initiated communication part of a kernel's intermediate representation, co-optimizes it alongside computation, and lowers it to specialized transport code. Obaleth plans completions around buffer lifetimes and optimizes memory ordering, synchronization, and batching across NVIDIA and AMD GPUs. On H100 GPUs, Obaleth sends small messages up to 3.3 faster, and on 256 GPUs runs all-to-all 1.25 and 2.3 faster than NVSHMEM and NCCL GIN, with much smaller memory and register use. In vLLM, it serves Qwen3-235B on 16 B200 GPUs and answers requests up to 1.29 faster than with DeepEP.
查看此版本解读 →查看此版本原文 ↗