Celebrating Diversity: A Mixture of Experts Approach for Runtime Mapping in Dynamic Environments
Matching program parallelism to platform parallelism using thread selection is difficult when the environment and available resources dynamically change. Existing compiler or runtime approaches are typically based on a one-size fits all policy.There is little ability to either evaluate or adapt the policy when encountering new external workloads or hardware resources. This paper focuses on selecting the best number of threads for a parallel application in dynamic environments. It develops a new scheme based on a mixture of experts approach. It learns online which, of a number of existing policies, or experts, is best suited for a particular environment without having to try out each policy. It does so by using a novel environment predictor as a proxy for the quality of an expert thread selection policy. Additional expert policies can easily be added and are selected only when appropriate. We evaluate our scheme in environments with varying external workloads and hardware resources. We then consider the case when workloads use affinity scheduling or are themselves adaptive and show that our approach, in all cases, outperforms existing schemes and surprisingly improves workload performance. On average, we improve 1.69x over OpenMP default, 1.36x over an online scheme, 1.27x over an offline policy and 1.21x over a state-of-art analytic model. Determining the right number and type of experts is an open problem and our initial analysis shows that adding more experts improves accuracy and performance.
Wed 17 Jun Times are displayed in time zone: Tijuana, Baja California change
14:00 - 15:40 | |||
14:00 25mTalk | Celebrating Diversity: A Mixture of Experts Approach for Runtime Mapping in Dynamic Environments Research Papers Media Attached | ||
14:25 25mTalk | Efficient Execution of Recursive Programs on Commodity Vector Hardware Research Papers Bin RenPacific Northwest National Laboratories, Youngjoon JoPurdue University, Sriram KrishnamoorthyPacific Northwest National Laboratories, Kunal AgrawalWashington University in St. Louis, Milind KulkarniPurdue University Media Attached | ||
14:50 25mTalk | Loop and Data Transformations for Sparse Matrix Code Research Papers Anand VenkatUniversity of Utah, Mary HallUniversity of Utah, Michelle StroutColorado State University Media Attached | ||
15:15 25mTalk | Synthesizing Parallel Graph Programs via Automated Planning Research Papers Dimitrios PrountzosThe University of Texas at Austin, Texas, USA, Roman ManevichBen-Gurion University of the Negev, Keshav PingaliUniversity of Texas, Austin Media Attached |