Post-silicon CPU adaptation made practical using machine learning
Stephen J. Tarsa, Rangeen Basu Roy Chowdhury, Julien Sébot, Gautham N. Chinya, Jayesh Gaur, Karthik Sankaranarayanan, Chit-Kwan Lin, Robert S. Chappell, Ronak Singhal, Hong Wang · 2019
Processors that adapt architecture to workloads at runtime promise compelling performance per watt (PPW) gains, offering one way to mitigate diminishing returns from pipeline scaling. State-of-the-art adaptive CPUs deploy machine learning (ML) models on-chip to optimize hardware by recognizing workload patterns in event counter data. However, despite breakthrough PPW gains, such designs are not yet widely adopted due to the potential for systematic adaptation errors in the field.