Loading...
Please wait, while we are loading the content...
Similar Documents
Supporting transparent thread assignment in heterogeneous multicore processors using reinforcement learning.
| Content Provider | CiteSeerX |
|---|---|
| Author | Yan, Xiaolei Sawalha, Lina Mcgovern, Amy Barnes, Ronald D. |
| Abstract | Abstract—Heterogeneity in multicore processor systems creates challenges in effectively mapping processes to diverse cores. While most approaches require programmer partitioning between core types or permutation of thread schedules to find the optimal mapping, we introduce a new machine learning approach to automated thread assignment. We train a reinforcement learning agent to assign threads to the best performing core given the state of the program and the processor cores. We present preliminary results demonstrating the promise of this approach for two and four heterogeneous cores using multiprogram workloads from SPEC CPU2006 benchmarks. We further discuss the limitations of this initial approach and propose future directions for improving our technique. I. |
| File Format | |
| Access Restriction | Open |
| Subject Keyword | Transparent Thread Assignment Thread Schedule Spec Cpu2006 Benchmark Multiprogram Workload Programmer Partitioning Initial Approach Abstract Heterogeneity Processor Core Propose Future Direction Thread Assignment Core Type Heterogeneous Core Performing Core Multicore Processor System New Machine Present Preliminary Result Optimal Mapping |
| Content Type | Text |
