Algorithm-Directed Data Placement in Explicitly Managed Non-Volatile Memory
- 31 May 2016
- conference paper
- conference paper
- Published by Association for Computing Machinery (ACM) in Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing
- p. 141-152
- https://doi.org/10.1145/2907294.2907321
Abstract
The emergence of many non-volatile memory (NVM) techniques is poised to revolutionize main memory systems because of the relatively high capacity and low lifetime power consumption of NVM. However, to avoid the typical limitation of NVM as the main memory, NVM is usually combined with DRAM to form a hybrid NVM/DRAM system to gain the benefits of each. However, this integrated memory system raises a question on how to manage data placement and movement across NVM and DRAM, which is critical for maximizing the benefits of this integration. The existing solutions have several limitations, which obstruct adoption of these solutions in the high performance computing (HPC) domain. In particular, they cannot take advantage of application semantics, thus losing critical optimization opportunities and demanding extensive hardware extensions; they implement persistent semantics for resilience purpose while suffering large performance and energy overhead. In this paper, we re-examine the current hybrid memory designs from the HPC perspective, and aim to leverage the knowledge of numerical algorithms to direct data placement. With explicit algorithm management and limited hardware support, we optimize data movement between NVM and DRAM, improve data locality, and implement a relaxed memory persistency scheme in NVM. Our work demonstrates significant benefits of integrating algorithm knowledge into the hybrid memory design to achieve multi-dimensional optimization (performance, energy, and resilience) in HPC.Keywords
Funding Information
- National Science Foundation (CCF-1553645, CCF-1305622, ACI-1305624, CCF-1513201)
- Department of Energy, Office of Science (ASCR)
This publication has 36 references indexed in Scilit:
- Optimizing the LU Factorization for Energy Efficiency on a Many-Core ArchitectureLecture Notes in Computer Science, 2014
- Designing LU-QR Hybrid Solvers for Performance and StabilityPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2014
- Correcting soft errors online in LU factorizationPublished by Association for Computing Machinery (ACM) ,2013
- Algorithm-based fault tolerance for dense matrix factorizationsACM SIGPLAN Notices, 2012
- NV-HeapsACM SIGARCH Computer Architecture News, 2011
- Scalable high performance main memory system using phase-change memory technologyACM SIGARCH Computer Architecture News, 2009
- A higher order estimate of the optimum checkpoint interval for restart dumpsFuture Generation Computer Systems, 2006
- PinACM SIGPLAN Notices, 2005
- Direct Cache Access for High Bandwidth Network I/OACM SIGARCH Computer Architecture News, 2005
- The cache performance and optimizations of blocked algorithmsACM SIGPLAN Notices, 1991