Details
| Original language | English |
|---|---|
| Title of host publication | DIMES '23 |
| Subtitle of host publication | Proceedings of the 1st Workshop on Disruptive Memory Systems |
| Pages | 52-59 |
| Number of pages | 8 |
| ISBN (electronic) | 9798400703003 |
| Publication status | Published - 23 Oct 2023 |
| Event | 1st Workshop on Disruptive Memory Systems, DIMES 2023, co-located with the 29th ACM Symposium on Operating Systems Principles, SOSP 2023 - Koblenz, Germany Duration: 23 Oct 2023 → 23 Oct 2023 |
Abstract
The tremendous growth of RAM capacity - now exceeding multiple terabytes - necessitates a reevaluation of traditional memory-management methods, which were developed when resources were scarce. Current virtual-memory subsystems handle address-space regions as sets of individual 4-KiB pages with demand paging and copy-on-write, resulting in significant management overhead. Although huge pages reduce the number of managed entities, they induce internal fragmentation and have a coarse copy granularity.To address these problems, we introduce Morsels, a novel virtual-memory-management paradigm that is purely based on hardware data structures and enables the efficient sharing of virtual-memory objects between processes and devices while being well suited for non-volatile memory. Our benchmarks show that Morsels reduce the mapping time for a 6.82-GiB machine-learning model by up to 99.8 percent compared to conventional memory mapping in Linux.
ASJC Scopus subject areas
- Computer Science(all)
- Computational Theory and Mathematics
- Computer Science(all)
- Computer Science Applications
- Computer Science(all)
- Software
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
DIMES '23: Proceedings of the 1st Workshop on Disruptive Memory Systems. 2023. p. 52-59.
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Morsels
T2 - 1st Workshop on Disruptive Memory Systems, DIMES 2023, co-located with the 29th ACM Symposium on Operating Systems Principles, SOSP 2023
AU - Halbuer, Alexander
AU - Dietrich, Christian
AU - Rommel, Florian
AU - Lohmann, Daniel
N1 - Funding Information: We thank our reviewers for their valuable feedback. This work was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – 468988364, 501887536.
PY - 2023/10/23
Y1 - 2023/10/23
N2 - The tremendous growth of RAM capacity - now exceeding multiple terabytes - necessitates a reevaluation of traditional memory-management methods, which were developed when resources were scarce. Current virtual-memory subsystems handle address-space regions as sets of individual 4-KiB pages with demand paging and copy-on-write, resulting in significant management overhead. Although huge pages reduce the number of managed entities, they induce internal fragmentation and have a coarse copy granularity.To address these problems, we introduce Morsels, a novel virtual-memory-management paradigm that is purely based on hardware data structures and enables the efficient sharing of virtual-memory objects between processes and devices while being well suited for non-volatile memory. Our benchmarks show that Morsels reduce the mapping time for a 6.82-GiB machine-learning model by up to 99.8 percent compared to conventional memory mapping in Linux.
AB - The tremendous growth of RAM capacity - now exceeding multiple terabytes - necessitates a reevaluation of traditional memory-management methods, which were developed when resources were scarce. Current virtual-memory subsystems handle address-space regions as sets of individual 4-KiB pages with demand paging and copy-on-write, resulting in significant management overhead. Although huge pages reduce the number of managed entities, they induce internal fragmentation and have a coarse copy granularity.To address these problems, we introduce Morsels, a novel virtual-memory-management paradigm that is purely based on hardware data structures and enables the efficient sharing of virtual-memory objects between processes and devices while being well suited for non-volatile memory. Our benchmarks show that Morsels reduce the mapping time for a 6.82-GiB machine-learning model by up to 99.8 percent compared to conventional memory mapping in Linux.
UR - http://www.scopus.com/inward/record.url?scp=85176912779&partnerID=8YFLogxK
U2 - 10.1145/3609308.3625267
DO - 10.1145/3609308.3625267
M3 - Conference contribution
AN - SCOPUS:85176912779
SP - 52
EP - 59
BT - DIMES '23
Y2 - 23 October 2023 through 23 October 2023
ER -