TY - JOUR
T1 - A Case Study of a DRAM-NVM Hybrid Memory Allocator for Key-Value Stores
AU - Kim, Minjae
AU - Kim, Bryan S.
AU - Lee, Eunji
AU - Lee, Sungjin
N1 - Funding Information:
This research was supported by the MOTIE (Ministry of Trade, Industry & Energy under Grant 1415181081 and KSRC (Korea Semiconductor Research Consortium) under Grant 20019402 support program for the development of the future semiconductor device, the Samsung Research Funding Incubation Center of Samsung Electronics under Grant SRFCIT1902-03, and the NRF grant funded by the Korea government (Ministry of Science and ICT) under Grants NRF-2018R1A5A1060031 and NRF-2019R1A2C1090337.
Publisher Copyright:
IEEE
PY - 2022
Y1 - 2022
N2 - As non-volatile memory (NVM) technologies advance, commercial NVDIMM devices have been made readily available for various computing systems. To efficiently utilize the high-density and high-capacity of NVM, the latest Xeon CPUs support a special Memory Mode that turns the DRAM into a last-level (L4) cache and uses NVM as the user-addressable system memory. Unfortunately, Memory Mode often provides low performance, even slower than when only NVM is used without any DRAM cache. According to our analysis, this is due to the inefficient management of a DRAM cache by the integrated memory controller, which results in high miss rates. This paper proposes a new hybrid memory allocator, called TARMAC. By employing intelligent yet lightweight memory management policies at the memory allocator level, TARMAC manages two different types of memory devices more efficiently, achieving 37% higher cache hit rate, 67% higher throughput, and 40% shorter memory latency than the hardware-based Memory Mode, on average. TARMAC exposes memory interfaces compatible with traditional memory allocators, enabling existing software to use TARMAC without any manual modification.
AB - As non-volatile memory (NVM) technologies advance, commercial NVDIMM devices have been made readily available for various computing systems. To efficiently utilize the high-density and high-capacity of NVM, the latest Xeon CPUs support a special Memory Mode that turns the DRAM into a last-level (L4) cache and uses NVM as the user-addressable system memory. Unfortunately, Memory Mode often provides low performance, even slower than when only NVM is used without any DRAM cache. According to our analysis, this is due to the inefficient management of a DRAM cache by the integrated memory controller, which results in high miss rates. This paper proposes a new hybrid memory allocator, called TARMAC. By employing intelligent yet lightweight memory management policies at the memory allocator level, TARMAC manages two different types of memory devices more efficiently, achieving 37% higher cache hit rate, 67% higher throughput, and 40% shorter memory latency than the hardware-based Memory Mode, on average. TARMAC exposes memory interfaces compatible with traditional memory allocators, enabling existing software to use TARMAC without any manual modification.
KW - Analytical models
KW - Data models
KW - Memory Allocator
KW - Memory Performance Analysis
KW - Memory management
KW - Non-volatile Memory
KW - Nonvolatile memory
KW - Performance evaluation
KW - Random access memory
KW - Throughput
UR - http://www.scopus.com/inward/record.url?scp=85136038674&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85136038674&partnerID=8YFLogxK
U2 - 10.1109/LCA.2022.3197654
DO - 10.1109/LCA.2022.3197654
M3 - Article
AN - SCOPUS:85136038674
SN - 1556-6056
SP - 1
EP - 4
JO - IEEE Computer Architecture Letters
JF - IEEE Computer Architecture Letters
ER -