arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.12378cs.ARcs.AIcs.ET
收藏

StoX-Net:面向高效存内计算DNN加速器的部分和随机处理

StoX-Net: Stochastic Processing of Partial Sums for Efficient In-Memory Computing DNN Accelerators

Iowa State University of Science and Technology(爱荷华州立科技大学)

Ethan G Rogers, Sohan Salahuddin Mugdho, Kshemal Kshemendra Gupte, Cheng Wang

AI总结:

针对存内计算中ADC开销大的瓶颈,提出基于自旋轨道矩磁隧道结概率开关的阵列级部分和随机处理方案,消除ADC并提升能效与面积效率,配合量化感知训练和非均匀采样长度转换,在CIFAR-10上实现高达16倍能效、8倍延迟和10倍面积改进,同时保持接近软件精度。

摘要:

基于交叉阵列的存内计算(IMC)已成为深度神经网络(DNN)硬件加速的有前景平台。然而,IMC系统的能耗和延迟主要受制于外围模数转换器(ADC)的巨大开销。为解决这一ADC瓶颈,本文提出在阵列级部分和(PS)上实现随机处理,以实现高效的IMC。利用自旋轨道矩磁隧道结的概率性开关特性,所提出的PS处理消除了昂贵的ADC,在能量和面积效率上实现了显著提升。为减轻精度损失,我们开发了PS量化感知训练,使得随机PS上的反向传播成为可能。此外,提出了一种采用非均匀采样长度的随机转换新方案。在CIFAR-10数据集上运行ResNet20时,我们的架构到算法协同设计相比采用标准ADC的IMC,在能量、延迟和面积上分别实现了高达16倍、8倍和10倍的改进。使用随机PS的优化设计配置,在能量-延迟积上相比全精度ADC的IMC(稀疏低位ADC的IMC)实现了130倍(24倍)的改进,同时在各基准分类任务中保持了接近软件级的精度。

英文摘要:

Crossbar-based in-memory computing (IMC) has emerged as a promising platform for hardware acceleration of deep neural networks (DNNs). However, the energy and latency of IMC systems are dominated by the large overhead of the peripheral analog-to-digital converters (ADCs). To address such ADC bottleneck, here we propose to implement stochastic processing of array-level partial sums (PS) for efficient IMC. Leveraging the probabilistic switching of spin-orbit torque magnetic tunnel junctions, the proposed PS processing eliminates the costly ADC, achieving significant improvement in energy and area efficiency. To mitigate accuracy loss, we develop PS-quantization-aware training that enables backward propagation across stochastic PS. Furthermore, a novel scheme with an inhomogeneous sampling length of the stochastic conversion is proposed. When running ResNet20 on the CIFAR-10 dataset, our architecture-to-algorithm co-design demonstrates up to 16x, 8x, and 10x improvement in energy, latency, and area, respectively, compared to IMC with standard ADC. Our optimized design configuration using stochastic PS achieved 130x (24x) improvement in Energy-Delay-Product compared to IMC with full precision ADC (sparse low-bit ADC), while maintaining near-software accuracy at various benchmark classification tasks.

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。

↑