Name
Title | ||
|---|---|---|
Runtime data management on non-volatile memory-based heterogeneous memory for task-parallel programs. |
Abstract | ||
|---|---|---|
Non-volatile memory (NVM) provides a scalable solution to replace DRAM as main memory. Because of relatively high latency and low bandwidth of NVM (comparing with DRAM), NVM often pairs with DRAM to build a heterogeneous main memory system (HMS). Deciding data placement on NVM-based HMS is critical to enable future NVM-based HPC. In this paper, we study task-parallel programs, and introduce a runtime system to address the data placement problem on NVM-based HMS. Leveraging semantics and execution mode of task-parallel programs, we efficiently characterize memory access patterns of tasks and reduce data movement overhead. We also introduce a performance model to predict performance for tasks with various data placements on HMS. Evaluating with a set of HPC benchmarks, we show that our runtime system achieves higher performance than a conventional HMS-oblivious runtime (24% improvement on average) and two state-of-the-art HMS-aware solutions (16% and 11% improvement on average, respectively). |
Year | DOI | Venue |
|---|---|---|
2018 | 10.1109/SC.2018.00034 | SC |
Keywords | Field | DocType |
Task analysis,Nonvolatile memory,Random access memory,Runtime,Programming,Data models,Analytical models | Dram,Data modeling,Computer science,Latency (engineering),Parallel computing,Non-volatile memory,Bandwidth (signal processing),Data management,Runtime system,Scalability,Embedded system | Conference |
ISBN | Citations | PageRank |
978-1-5386-8384-2 | 3 | 0.39 |
References | Authors | |
31 | 3 | |