arXiv cs.AI· Orhun G\"orkem, Athanasios Vasilopoulos, Georgios Pagonis, Victoria Clerico, Hadjer Benmeziane, Irem Boybat, Hsinyu Tsai, Abu Sebastian, William Andrew Simon·· 12 小时前AI 评分22
CAMP：面向 3D-IMC 阵列上 LLM 的上下文感知权重映射与放置算法
CAMP: Context-Aware Mapping and Placement for LLMs on 3D-IMC Arrays
AI 导读
CAMP 是一种面向 3D-IMC 的 LLM 加速器的推理上下文感知权重映射算法，通过冲突矩阵刻画架构、目标网络与运行时上下文引发的资源争用，并用性能分析驱动的反馈循环剪枝搜索空间。该离线软件方案无需改动硬件，将 3D-IMC 计算利用率从 42.4% 提升至 60.3%，相比 SOTA 映射算法实现最高 2.0 倍的 3D-IMC 吞吐和 1.74 倍的端到端吞吐。
正文 · 原文
Abstract:Large Language Models (LLMs) incur severe memory-access overheads due to the cost of loading large weight matrices and key-value (KV) caches. Analog In-Memory Computing (AIMC) is an emerging solution that enables computation directly on static weights stored in memory, substantially reducing energy and latency. 3D-IMC further stacks memory arrays into vertical tiers, increasing capacity but revealing a new challenge: how weights are placed in memory determines which matrix-vector multiplications (MVMs) can execute concurrently, affecting latency and throughput. The concurrency aspect of 3D-IMC weight-mapping remains underexplored in the literature, leading to sub-optimal utilization of compute resources on 3D-IMC chips. In this context we propose CAMP (Context-Aware Mapping and Placement), an inference-context-aware weight mapping algorithm for 3D-IMC-based LLM accelerators. CAMP captures the potential resource contention resulting from the architecture, targeted network, and runtime context in a conflict matrix, enabling throughput-optimized weight mapping. This mapping is refined through a profiling-driven feedback loop that efficiently prunes the search space and returns optimized mappings within practical runtimes. CAMP is an offline, software-based solution requiring no hardware changes and improving 3D-IMC compute utilization from 42.4% to 60.3%, delivering up to 2.0x higher 3D-IMC throughput and 1.74x higher end-to-end throughput over state-of-the-art mapping algorithms.
| Subjects: | Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.07707 [cs.AR] |
| (or arXiv:2611.07707v1 [cs.AR] for this version) | |
| https://doi.org/10.48550/arXiv.2611.07707 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Orhun Görkem [view email]
[v1]
Thu, 5 Nov 2026 09:27:31 UTC (2,435 KB)
来源：arXiv cs.AI · arxiv.org