Low-Latency Asynchronous Logic Design for Inference at the Edge

12/07/2020
by   Adrian Wheeldon, et al.
0

Modern internet of things (IoT) devices leverage machine learning inference using sensed data on-device rather than offloading them to the cloud. Commonly known as inference at-the-edge, this gives many benefits to the users, including personalization and security. However, such applications demand high energy efficiency and robustness. In this paper we propose a method for reduced area and power overhead of self-timed early-propagative asynchronous inference circuits, designed using the principles of learning automata. Due to natural resilience to timing as well as logic underpinning, the circuits are tolerant to variations in environment and supply voltage whilst enabling the lowest possible latency. Our method is exemplified through an inference datapath for a low power machine learning application. The circuit builds on the Tsetlin machine algorithm further enhancing its energy efficiency. Average latency of the proposed circuit is reduced by 10x compared with the synchronous implementation whilst maintaining similar area. Robustness of the proposed circuit is proven through post-synthesis simulation with 0.25 V to 1.2 V supply. Functional correctness is maintained and latency scales with gate delay as voltage is decreased.

READ FULL TEXT
09/02/2021

Self-timed Reinforcement Learning using Tsetlin Machine

We present a hardware design for the learning datapath of the Tsetlin ma...
06/15/2021

Low-Energy and CPA-Resistant Adiabatic CMOS/MTJ Logic for IoT Devices

The tremendous growth in the number of Internet of Things (IoT) devices ...
09/11/2019

An SR Flip-Flop based Physical Unclonable Functions for Hardware Security

Physical Unclonable Functions (PUFs) have emerged as a promising solutio...
04/25/2019

TS Cache: A Fast Cache with Timing-speculation Mechanism Under Low Supply Voltages

To mitigate the ever-worsening Power Wall problem, more and more applica...
10/15/2019

Neural Network Design for Energy-Autonomous AI Applications using Temporal Encoding

Neural Networks (NNs) are steering a new generation of artificial intell...
04/23/2019

IRC: Cross-layer design exploration of Intermittent Robust Computation units for IoTs

Energy-harvesting-powered computing offers intriguing and vast opportuni...
03/05/2021

PCP: Preemptive Circuit Padding against Tor circuit fingerprinting

Online anonymity and privacy has been based on confusing the adversary b...