Systolic Tensor Core for Arithmetics calculations

Article type: Research Article

Authors: Ibarra Carrillo, Mario Alfredo^{a; *} | Montiel Pérez, Jesús Yaljá^{b; *} | Molina Lozano, Herón^{c; *}

Affiliations: [a] Centro de Investigación en Computación, Instituto Politécnico Nacional, Laboratorio de Robótica yMecatrónica, México, Ciudad de México, México | [b] Centro de Investigación en Computación, Instituto Politécnico Nacional, Laboratorio de Robótica yMecatrónica, México, Ciudad de México, México | [c] Centro de Investigación en Computación, Instituto Politécnico Nacional, Laboratorio deMicrotecnología y Sisatemas Embebidos, México, Ciudad deMéxico, México

Correspondence: [*] Corresponding authors. Mario Alfredo Ibarra Carrillo. E-mail: [email protected]; Jesús Yaljá Montiel Pérez. E-mail: [email protected] and Herón Molina Lozano. E-mail: [email protected]

Abstract: Today, it is the amount of data that defines the existence of mankind. Scientists respond to the large amount of required calculations by developing hardware in several directions. One of them is to increase the number of arithmetic elements. Another direction is to create new architectures that represent new algorithms for processing numerical data. We have chosen the second direction by developing a new systolic core architecture, which implies an improvement in efficiency, i.e. performing the same task with the same number of arithmetic elements but reducing the latency. Measurements are made in terms of computational capacity and the number of arithmetic elements involved in the operations. The results of the tests are compared with data from a number of selected articles. Today, we have achieved 3.2GFlops with only two modules. In the future, we plan to integrate up to four of our cores in a system with its own memory and management processor and at a higher operating frequency.

Keywords: Systolic array, systolic tensor core, accelerated matrix multiplication, accelerated convolution

DOI: 10.3233/JIFS-219361

Journal: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-14, 2024