Single-node Partitioned-Memory for Huge Graph Analytics: Cost and Performance Trade-offs

Conference · Sun Nov 14 04:00:00 EST 2021

DOI:https://doi.org/10.1145/3458817.3476156· OSTI ID:1836010

Ghosh, Sayan ^[1]; Tallent, Nathan R. ^[1]; Minutoli, Marco ^[1]; Halappanavar, Mahantesh ^[1]; Peri, Ramesh ^[2]; Kalyanaraman, Ananth ^[3]

---

1. BATTELLE (PACIFIC NW LAB)
2. Intel
3. Washington State University

Nonvolatile memory NVDIMMs, available as Intel Optane, are less expensive than DRAM and bring large byte-addressable storage within reach to many applications. Evaluations on graph analytics have shown promising performance only when DRAM is used as a hardware cache (Memory mode). An open question is whether graph applications can exploit Optane and DRAM directly (AppDirect mode) and achieving better-than-DRAM average bandwidth and run times. We evaluate Optane as a volatile pool on two large-scale graph applications with very different computational patterns, Grappolo and Ripples. We show that AppDirect mode can deliver better-than-DRAM performance, by allocating data structures to Optane and DRAM according to their access characteristics, resulting in higher average memory bandwidth and lower average latency. Memory mode provides DRAM-competitive performance with capacity equal to persistent memory. We demonstrate occasional 4x improvement using the latest AppDirect option and frequently observe competitive performance between Optane AppDirect Memory modes and DRAM.

• Research Organization: Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)
• Sponsoring Organization: USDOE
• DOE Contract Number: AC05-76RL01830
• OSTI ID: 1836010
• Report Number(s): PNNL-SA-161359
• Country of Publication: United States
• Language: English