Abstract
Memristor-based accelerator (MBA) has demonstrated its capability in accelerating matrix-vector multiplication (MVM) with high performance and energy efficiency. However, it is hard to determine whether and how well an application can benefit from MBAs in a heterogeneous computing architecture. In this article, we propose a simulation framework called MHSim to evaluate the energy efficiency and performance of applications running with both MBAs and CPUs. MHSim provides flexible system-level interfaces and circuit-level simulation models for designers to configure heterogeneous computing architectures. We design a general-purpose MBA which enables floating-point computation models for general matrix-matrix multiplication (GEMM). Our simulation framework can quantify the performance and energy efficiency of different MBA architectures for various applications. We validate our simulation framework with SPICE and evaluate the accuracy and performance of MBAs via several case studies. Experimental results demonstrate that the deviations of energy consumption and latency are only 0.47% and 0.49% on average compared with SPICE-based simulation.
Highlights
M ATRIX multiplications are fundamental operations in a wide range of applications such as convolutional neural networks (CNNs) [1] and graph processing [2], [3]
To evaluate the performance benefit of using Memristor-based accelerator (MBA) for more general applications, we propose an effective design in MBAs to support floating-point arithmetic according to the IEEE-754 standard
For VGG16 and ResNet50, we find that the speedups of the MBA are less than that of GPUs because the finite memristor resource limits the number of weight replications due to massive weight parameters
Summary
M ATRIX multiplications are fundamental operations in a wide range of applications such as convolutional neural networks (CNNs) [1] and graph processing [2], [3] They often lead to massive data movement between CPU and main memory, which has become a major performance bottleneck in traditional von Neumann architectures. A general-purpose and flexible simulation framework for various applications requires high-level abstractions of computation operators that can be accelerated by MBAs. Third, existing MBA simulators have not yet supported floating-point operations in MVMs. Since floating-point numbers are commonly used in many real-world applications, it is necessary to support floating-point operands in matrix multiplications. We propose MHSim [24], a general-purpose and open-source simulation framework for benchmarking all real-world applications in the heterogeneous computing architecture. MHSim provides easy-to-use interfaces to design heterogeneous computing systems and evaluate their performance and energy efficiency with high accuracy, without modifying applications’ source codes.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
