The proliferation of hardware accelerators has enabled the pervasive use of machine-learning algorithms in a range of diverse real-world applications, from computer vision to natural language processing. In addition to building the systems and accelerators that have enabled this current momentum in artificial intelligence, the computer architecture community has also explored these new models to improve and optimize the computing systems that we build.
This is a less-explored but promising research direction, with important implications across the full computing stack: from software performance and profiling to operating systems, compilers, architecture, microarchitecture, and circuit design. Potential improvements involve increasing hardware performance and efficiency, performing design space explorations, improving design automation, and reducing the efforts of architecting and designing hardware.
This special issue of IEEE Micro will explore broadly the use of machine learning including supervised, unsupervised, and reinforcement learning-based techniques to improve computer architecture and computer systems. Papers on the following topics are solicited:
Use of machine learning to improve:
- Computer Architecture, Microarchitecture, and Accelerators
- Circuit Design and Layout
- Interconnects and Networking
- Memory and Storage Systems
- Runtime Systems
- Datacenter Management
- Computing at the Edge
- Algorithm Optimization of Hardware and Software Systems
- Hardware/Software Co-Design
- Source Code Optimization
- Modeling and Simulation Techniques
- Workload Characterization
- Profiling and Performance Optimization
- Submissions due: CLOSED
- Reviews due: 6 April 2020
- Revisions due: 8 June 2020
- Final reviews due: 29 June 2020
- Final notifications: 13 July 2020
- Publication: Sept/Oct 2020
Please see the Author Information page and the Magazine Peer Review page for more information. Please submit electronically through ScholarOne Manuscripts (https://mc.manuscriptcentral.