The Turbo-Charged Mapper: Fast and Optimal Mapping for Energy-efficient and Low-latency Accelerator Design

TL;DR

The Turbo-Charged Mapper (TCM) is a novel, fast algorithm that finds optimal mappings for DNN accelerators, significantly reducing search time and improving energy-delay performance by pruning the search space.

Contribution

We introduce dataplacement, a new mapping concept, enabling TCM to perform full search space exploration and find optimal mappings efficiently.

Findings

Reduces search space by up to 32 orders of magnitude.

Improves energy-delay-product by 1.2-6.5 times.

Reduces mapping search time from 5 hours to 17 seconds.

Abstract

The energy and latency of an accelerator running a deep neural network (DNN) depend on how the computation and data movement are scheduled in the accelerator (i.e., mapping), and picking an optimal mapping is essential to achieve high-performance accelerators. However, it is challenging to find mappings that maximize accelerator performance. The space of mappings is large, and prior works cannot guarantee finding optimal mappings because they use heuristics or metaheuristics to narrow the search space. To address this challenge, we propose the Turbo-Charged Mapper (TCM), a fast mapper that finds optimal mappings. The key to our approach is that we define a new mapping concept called dataplacement, which, like the prior concept of dataflow, allows for clear analysis and comparison of mappings. Through it, we identify opportunities to prune redundant and suboptimal mappings, reducing search space by up to 32 orders of magnitude ($10^{37}\rightarrow10^5$). TCM leverages these insights to perform full mapspace searches, making it the first mapper that can find optimal mappings in feasible runtime. Compared to prior mappers, TCM improves accelerator energy-delay-product by $1.2-6.5\times$ while simultaneously reducing mapping search time by $1000\times$ (5 hours $\rightarrow$ 17 seconds).