Performance characterization of a DRAM-NVM hybrid memory architecture for HPC applications using intel optane DC persistent memory modules

Onkar Patil, Latchesar Ionkov, Jason Lee, Frank Mueller, Michael Lang

Proceedings of the International Symposium on Memory Systems · 2019 · 64 citations · 29 references

DOIFull text

Open access

Concepts

TL;DR

Intel’s Optane DC PMM introduces byte‑addressable, non‑volatile memory that persists data without power, expands the memory hierarchy with up to eight times the capacity of DDR4 DRAM, and enables HPC applications to scale larger problem sizes. The study evaluates how a DRAM‑NVM hybrid memory system affects performance in high‑performance computing workloads. Using a STREAM‑like benchmark, the authors compare Optane DC to DDR4 DRAM and measure execution times for mini‑apps VPIC, SNAP, LULESH, and AMG under various Optane DC configurations. Optane‑only runs are 2–16 % slower than DRAM‑only and Memory‑mode for VPIC and up to six times slower for LULESH.

Abstract

Non-volatile, byte-addressable memory (NVM) has been introduced by Intel in the form of NVDIMMs named Intel® Optane™ DC PMM. This memory module has the ability to persist the data stored in it without the need for power. This expands the memory hierarchy into a hybrid memory system due the differences in access latency and memory bandwidth from DRAM, which has been the predominant byte-addressable main memory technology. The Optane DC memory modules have up to 8x the capacity of DDR4 DRAM modules which can expand the byte-address space up to 6 TB per node. Many applications can now scale up the their problem size given such a memory system. We evaluate the capabilities of this DRAM-NVM hybrid memory system and its impact on High Performance Computing (HPC) applications. We characterize the Optane DC in comparison to DDR4 DRAM with a STREAM-like custom benchmark and measure the performance for HPC mini-apps like VPIC, SNAP, LULESH and AMG under different configurations of Optane DC PMMs. We find that Optane-only executions are slower in terms of execution time than DRAM-only and Memory-mode executions by a minimum of 2 to 16% for VPIC and maximum of 6x for LULESH.

References

29