Browse > Article
http://dx.doi.org/10.7236/JIIBC.2019.19.3.165

Analysis of Tensor Processing Unit and Simulation Using Python  

Lee, Jongbok (Dept. of EI Engineering, Hansung University)
Publication Information
The Journal of the Institute of Internet, Broadcasting and Communication / v.19, no.3, 2019 , pp. 165-171 More about this Journal
Abstract
The study of the computer architecture has shown that major improvements in price-to-energy performance stems from domain-specific hardware development. This paper analyzes the tensor processing unit (TPU) ASIC which can accelerate the reasoning of the artificial neural network (NN). The core device of the TPU is a MAC matrix multiplier capable of high-speed operation and software-managed on-chip memory. The execution model of the TPU can meet the reaction time requirements of the artificial neural network better than the existing CPU and the GPU execution models, with the small area and the low power consumption even though it has many MAC and large memory. Utilizing the TPU for the tensor flow benchmark framework, it can achieve higher performance and better power efficiency than the CPU or CPU. In this paper, we analyze TPU, simulate the Python modeled OpenTPU, and synthesize the matrix multiplication unit, which is the key hardware.
Keywords
neural network; machine learning; matrix multiply unit;
Citations & Related Records
연도 인용수 순위
  • Reference
1 N. P. Jouppi et al., "In-Datacenter Performance Analysis of a Tensor Processing Unit," 44th International Symposium on Computer Architecture (ISCA), Jun. 2017.
2 P. Ienne, T. Cornu, and G. Kuhn, "Special Purpose digital hardware for neural networks: An architectural survery," Journal of VLSI signal processing systems for signal, image and video technology, Vol. 1, No. 13, 1996.
3 D. Hammerstrom, "A VLSI Architecture for high-performnace, low-cost, on-chip learning," International Joint Conference on Neural Networks, Jun. 1990.
4 U. Ramacher et. al., "Design of a 1st Generation Nerocomputer," VLSI design of Neural Networks. 1991.
5 K. Asanovik et. al, "Training Neural Networks with Spert-II," Parallel Architectures for Artificial Networks : Paradigm and Implementations, Nov. 1998.
6 https://github.com/UCSBarchlab/OpenTPU