Document Type

Article

Author ORCID Identifier

Noel Daniel Gundi https://orcid.org/0000-0001-6331-6844

Journal/Book Title/Conference

Journal of Low Power Electronics and Applications

Volume

14

Issue

1

Publisher

MDPI AG

Publication Date

1-9-2024

Journal Article Version

Version of Record

First Page

1

Last Page

18

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Abstract

Artificial Intelligence (AI) hardware accelerators have seen tremendous developments in recent years due to the rapid growth of AI in multiple fields. Many such accelerators comprise a Systolic Multiply–Accumulate Array (SMA) as its computational brain. In this paper, we investigate the faulty output characterization of an SMA in a real silicon FPGA board. Experiments were run on a single Zybo Z7-20 board to control for process variation at nominal voltage and in small batches to control for temperature. The FPGA is rated up to 800 MHz in the data sheet due to the max frequency of the PLL, but the design is written using Verilog for the FPGA and C++ for the processor and synthesized with a chosen constraint of a 125 MHz clock. We then operate the system at a frequency range of 125 MHz to 450 MHz for the FPGA and the nominal 667 MHz for the processor core to produce timing errors in the FPGA without affecting the processor. Our extensive experimental platform with a hardware–software ecosystem provides a methodological pathway that reveals fascinating characteristics of SMA behavior under an overclocked environment. While one may intuitively expect that timing errors resulting from overclocked hardware may produce a wide variation in output values, our post-silicon evaluation reveals a lack of variation in erroneous output values. We found an intriguing pattern where error output values are stable for a given input across a range of operating frequencies far exceeding the rated frequency of the FPGA.

Download

Share

COinS