arXiv 2502.021422025-02-04cs.AR

迈向基于LUT的高效PIM：一种面向现代工作负载的可扩展低功耗方法

Towards Efficient LUT-based PIM: A Scalable and Low-Power Approach for Modern Workloads

Bahareh Khabbazan, Marc Riera, Antonio González

备注：: 18 pages, 13 figures

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出基于LUT的PuM架构Lama，利用DRAM mat级并行性减少激活命令开销，支持8位运算且面积开销仅2.47%，并进一步提出LamaAccel加速器，在性能和能效上显著超越现有PuM架构、CPU、TPU和GPU。

摘要：

在深度学习等内存密集型工作负载中，数据移动所产生的能耗比计算本身的能耗高出三个数量级以上。由于这些工作负载涉及内存与处理单元之间频繁的数据传输，解决数据移动开销对于提升性能至关重要。利用内存进行计算（Processing-using-memory，PuM）通过支持内存内计算来最大限度地减少数据传输，提供了一种有效的解决方案。本文提出Lama，一种基于查找表（LUT）的PuM架构，旨在通过支持DRAM子阵列内每个mat的独立列访问来高效执行SIMD操作。Lama利用DRAM的mat级并行性和开放页策略，显著减少了高能耗内存激活（ACT）命令的数量，而ACT命令是大多数PuM架构中的主要开销来源。与以往的PuM方案不同，Lama支持高达8位操作数精度而无需分解计算，同时仅产生2.47%的面积开销。我们的评估表明，对于批量8位乘法，Lama相较于最先进的PuM架构实现了平均8.5倍的性能提升，相较于CPU实现了3.8倍的性能提升，同时分别获得了6.9倍和8倍的能效收益。我们还提出了LamaAccel，一种基于HBM的PuM加速器，利用Lama加速基于注意力机制的模型推理。LamaAccel采用指数量化来优化点积运算中的乘积/累加操作，将其转化为加法和计数等更简单的任务。与TPU/GPU相比，LamaAccel实现了高达9.3倍/19.2倍的能耗降低和4.8倍/9.8倍的加速，同时相较于最先进的PuM基线实现了高达5.8倍的能耗降低和2.1倍的加速。

英文摘要：

Data movement in memory-intensive workloads, such as deep learning, incurs energy costs that are over three orders of magnitude higher than the cost of computation. Since these workloads involve frequent data transfers between memory and processing units, addressing data movement overheads is crucial for improving performance. Processing-using-memory (PuM) offers an effective solution by enabling in-memory computation, thereby minimizing data transfers. In this paper we propose Lama, a LUT-based PuM architecture designed to efficiently execute SIMD operations by supporting independent column accesses within each mat of a DRAM subarray. Lama exploits DRAM's mat-level parallelism and open-page policy to significantly reduce the number of energy-intensive memory activation (ACT) commands, which are the primary source of overhead in most PuM architectures. Unlike prior PuM solutions, Lama supports up to 8-bit operand precision without decomposing computations, while incurring only a 2.47% area overhead. Our evaluation shows Lama achieves an average performance improvement of 8.5x over state-of-the-art PuM architectures and a 3.8x improvement over CPU, along with energy efficiency gains of 6.9x/8x, respectively, for bulk 8-bit multiplication. We also introduce LamaAccel, an HBM-based PuM accelerator that utilizes Lama to accelerate the inference of attention-based models. LamaAccel employs exponential quantization to optimize product/accumulation in dot-product operations, transforming them into simpler tasks like addition and counting. LamaAccel delivers up to 9.3x/19.2x reduction in energy and 4.8x/9.8x speedup over TPU/GPU, along with up to 5.8x energy reduction and 2.1x speedup over a state-of-the-art PuM baseline.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

AI总结：

摘要：

英文摘要：

相关论文