Approximate memory compression for energy-efficiency
Authors:
- 1 July 2017
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Abstract
Memory subsystems are a major energy bottleneck in computing platforms due to frequent transfers between processors and off-chip memory. We propose approximate memory compression, a technique that leverages the intrinsic resilience of emerging workloads such as machine learning and data analytics to reduce off-chip memory traffic and energy. To realize approximate memory compression, we enhance the memory controller to be aware of memory regions that contain approximation-resilient data, and to transparently compress/decompress the data written to/read from these regions. To provide control over approximations, the quality-aware memory controller conforms to a specified error constraint for each approximate memory region. We design a software interface that programmers can use to identify data structures that are resilient to approximations. We also propose a runtime quality control framework that automatically determines the error constraints for the identified data structures such that a given target application-level quality is maintained. We evaluate our proposal by implementing a hardware prototype using the Intel UniPHY-DDR3 memory controller and NIOS-II processor, a Hynix DDR3 DRAM module, and a Stratix-IV FPGA development board. Across a suite of 8 machine learning benchmarks, approximate memory compression obtains a 1.28× benefit in DRAM energy and a simultaneous 11.5% improvement in execution time for a small (<; 1.5%) loss in output quality.Keywords
This publication has 10 references indexed in Scilit:
- STAxCache: An approximate, energy efficient STT-MRAM cache2017
- Concise loads and stores: The case for an asymmetric compute-memory architecture for approximation2016
- Efficient reliability management in SoCs - an approximate DRAM perspective2016
- Approximate Computing: A SurveyIEEE Design & Test, 2015
- Approximate storage for energy efficient spintronic memories2015
- Approximate computing and the quest for computing efficiency2015
- Load Value Approximation2014
- Approximate storage in solid-state memories2013
- Flikker2011
- Hardware-assisted data compression for energy minimization in systems with embedded processors2002