SCALE-Sim V3: a Modular Cycle-Accurate Systolic Accelerator Simulator for End-To-End System Analysis
Ritik Raj, Sarbartha Banerjee, Nikhil Chandra, Zishen Wan, Jianming Tong, Ananda Samajdhar, Tushar Krishna
Abstract
The rapid advancements in AI, scientific computing, and high-performance computing (HPC) have driven the need for versatile and efficient hardware accelerators. Existing tools like SCALE-Sim v2 provide valuable cycle-accurate simulations for systolic-array-based architectures but fall short in supporting key modern features such as sparsity, multi-core scalability, and comprehensive memory analysis. To address these limitations, we present SCALE-Sim v3 (GitHub Repository), a modular, cycle-accurate simulator that extends the capabilities of its predecessor. SCALE-Sim v3 introduces five significant enhancements: multi-core simulation with spatio-temporal partitioning and hierarchical memory structures, support for sparse matrix multiplications (SpMM) with layer-wise and row-wise sparsity, integration with Ramulator for detailed DRAM analysis, precise data layout modeling to minimize memory stalls, and energy and power estimation via Accelergy. These improvements enable deeper end-to-end system analysis for modern AI accelerators, accommodating a wide variety of systems and workloads and providing detailed full-system insights into latency, bandwidth, and power efficiency. A <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$128 \times 128$</tex> array is <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$6.53 \times$</tex> faster than a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$32 \times 32$</tex> array for ViTbase, using only latency as a metric. However, SCALE-Sim v3 finds that <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$32 \times 32$</tex> is <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$2.86 \times$</tex> more energy-efficient due to better utilization and lower leakage energy. For EdP, <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$64 \times 64$</tex> outperforms both <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$128 \times 128$</tex> and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$32 \times 32$</tex> for ViT-base. SCALE-Sim v2 shows a 21 % reduction in compute cycles for six ResNet18 layers using weight-stationary (WS) dataflow compared to outputstationary (OS). However, when factoring in DRAM stalls, OS dataflow exhibits 30.1% lower execution cycles compared to WS, highlighting the critical role of detailed DRAM analysis.