Self-timed Reinforcement Learning using Tsetlin Machine
This work addresses energy efficiency for pervasive AI applications, but it is incremental as it builds on previous inference hardware design.
The paper tackles the problem of designing low-energy hardware for the Tsetlin machine algorithm by using asynchronous techniques like Petri nets and dual-rail, resulting in a design suitable for energy-limited applications such as personalized healthcare and IoT devices.
We present a hardware design for the learning datapath of the Tsetlin machine algorithm, along with a latency analysis of the inference datapath. In order to generate a low energy hardware which is suitable for pervasive artificial intelligence applications, we use a mixture of asynchronous design techniques - including Petri nets, signal transition graphs, dual-rail and bundled-data. The work builds on previous design of the inference hardware, and includes an in-depth breakdown of the automaton feedback, probability generation and Tsetlin automata. Results illustrate the advantages of asynchronous design in applications such as personalized healthcare and battery-powered internet of things devices, where energy is limited and latency is an important figure of merit. Challenges of static timing analysis in asynchronous circuits are also addressed.