Energy-efficient coarse-grain out-of-order execution
Abstract/Contents
- Abstract
- Throughout the past decade, energy efficient computing has been a major problem in the computer system space from device technologies, to chip design, to computer architecture and software systems. Today, we have more smartphone devices in the hands of consumers than ever before, and we generate more data on the internet than ever before. This trend implies that moving forward, building energy efficient systems remains an increasingly significant challenge both in the mobile industry and in the server industry. Building energy efficient processors is at the forefront of solving these technological challenges. This doctoral dissertation describes the Coarse-Grain Out-of-Order (CG-OoO) processor architecture. Block-level code processing is at the heart of the CG-OoO architecture; CG-OoO speculates, fetches, schedules, and commits code at block-level granularity. CG-OoO eliminates unnecessary accesses to energy consuming tables, and turns large tables into smaller and distributed tables that are cheaper to access. CG-OoO leverages compiler-level code optimizations to deliver more efficient static code. It exploits instruction-level parallelism and block-level parallelism. CG-OoO introduces the Skipahead model which is a complexity effective, limited out-of-order instruction issue model. It is an energy-performance proportional design that can scale according to the program load. Through the energy efficiency techniques applied to the compiler and processor pipeline stages, CG-OoO delivers over 50% energy reduction at the performance of the baseline out-of-order processor.
Description
Type of resource | text |
---|---|
Form | electronic; electronic resource; remote |
Extent | 1 online resource. |
Publication date | 2015 |
Issuance | monographic |
Language | English |
Creators/Contributors
Associated with | Mohammadi, Milad | |
---|---|---|
Associated with | Stanford University, Department of Electrical Engineering. | |
Primary advisor | Dally, William J | |
Thesis advisor | Dally, William J | |
Thesis advisor | Aamodt, Tor M | |
Thesis advisor | Aiken, Alexander | |
Thesis advisor | Kozyrakis, Christoforos, 1974- | |
Advisor | Aamodt, Tor M | |
Advisor | Aiken, Alexander | |
Advisor | Kozyrakis, Christoforos, 1974- |
Subjects
Genre | Theses |
---|
Bibliographic information
Statement of responsibility | Milad Mohammadi. |
---|---|
Note | Submitted to the Department of Electrical Engineering. |
Thesis | Thesis (Ph.D.)--Stanford University, 2015. |
Location | electronic resource |
Access conditions
- Copyright
- © 2015 by Milad Mohammadi
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...