← 研踪 PaperTrailarXiv:2611.25636 · v1
Tenon: A Unified Programming Framework for Retargetable PIM Compilation
首次提交 2026-11-19 15:23
版本更新 2026-11-19 15:23
雷达观测 2026-11-20 19:58
摘要
Processing-in-memory (PIM) is emerging as a building block of heterogeneous systems for memory-bound workloads, and commercial prototypes now span near-bank SIMD engines, bank-level GDDR accelerators, programmable DRAM cores, and bit-level in-SRAM processors. Their programming stacks are fragmented: no abstraction or compiler is shared across vendors, and per-device models range from instruction sequences to C-like languages. Data layout further complicates the problem: tight compute-memory integration makes data layout a first-order optimization, yet existing PIM compilers lack a unified way to represent and optimize it. A key observation is that PIM compilation targets can be modeled as nested grids of compute and storage units connected by legal data-transfer paths. We present Tenon, a unified programming framework for retargetable PIM compilation. It decouples workload, compilation target, and cost specifications, so a workload is expressed once and compiled through a unified programming framework. Tenon generalizes the linear layout algebra from GPU compilers to map logical work and data onto these compute and storage grids. It automatically constructs a minimum-movement data layout given a workload and a target specification. We evaluate Tenon on six PIM backends: five commercial backends and AttAcc, a state-of-the-art academic HBM-based PIM architecture. Across the commercial backends, it matches or outperforms vendor, expert, and compiler baselines, with geometric-mean speedups of up to and individual speedups up to . For end-to-end GPT-13B decoding on AttAcc, Tenon-compiled workloads run -- faster than those compiled by DCC, a state-of-the-art PIM compiler, across all three batch-4 configurations. Together, these results demonstrate Tenon's performance portability across diverse PIM target families.
作者备注:HPCA 2027
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 · 2026-11-19 15:23 · 最新观测版本
Tenon: A Unified Programming Framework for Retargetable PIM Compilation
Processing-in-memory (PIM) is emerging as a building block of heterogeneous systems for memory-bound workloads, and commercial prototypes now span near-bank SIMD engines, bank-level GDDR accelerators, programmable DRAM cores, and bit-level in-SRAM processors. Their programming stacks are fragmented: no abstraction or compiler is shared across vendors, and per-device models range from instruction sequences to C-like languages. Data layout further complicates the problem: tight compute-memory integration makes data layout a first-order optimization, yet existing PIM compilers lack a unified way to represent and optimize it. A key observation is that PIM compilation targets can be modeled as nested grids of compute and storage units connected by legal data-transfer paths. We present Tenon, a unified programming framework for retargetable PIM compilation. It decouples workload, compilation target, and cost specifications, so a workload is expressed once and compiled through a unified programming framework. Tenon generalizes the linear layout algebra from GPU compilers to map logical work and data onto these compute and storage grids. It automatically constructs a minimum-movement data layout given a workload and a target specification. We evaluate Tenon on six PIM backends: five commercial backends and AttAcc, a state-of-the-art academic HBM-based PIM architecture. Across the commercial backends, it matches or outperforms vendor, expert, and compiler baselines, with geometric-mean speedups of up to and individual speedups up to . For end-to-end GPT-13B decoding on AttAcc, Tenon-compiled workloads run -- faster than those compiled by DCC, a state-of-the-art PIM compiler, across all three batch-4 configurations. Together, these results demonstrate Tenon's performance portability across diverse PIM target families.
查看此版本解读 →查看此版本原文 ↗