Abstract
Non-uniform memory architecture (NUMA) system has numerous nodes with shared last level cache (LLC). Their shared LLC has brought many benefits in the cache utilization. However, LLC can be seriously polluted by tasks that cause huge I/O traffic for a long time since inclusive cache architecture of LLC replaces valid cache line by back-invalidate. Many research on the page coloring, partitioning, and pollute buffer mechanism handled this cache pollution. But, there are no scheduling approaches considering I/O-intensive tasks in NUMA systems. To address the above problem, OS scheduling that reduces cache pollution is highly needed in NUMA systems. In this paper, we propose a software-based mechanism that reduces shared LLC miss in NUMA systems. Our mechanism includes I/O traffic measurement and devil conscious scheduling. The experimental results show that LLC miss rate can be reduced up to 37.6%, and our approach improves execution time to 1.48%.
| Original language | English |
|---|---|
| Title of host publication | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
| Pages | 272-282 |
| Number of pages | 11 |
| Volume | 7335 LNCS |
| Edition | PART 3 |
| DOIs | |
| Publication status | Published - 2012 Jul 23 |
| Event | 12th International Conference on Computational Science and Its Applications, ICCSA 2012 - Salvador de Bahia, Brazil Duration: 2012 Jun 18 → 2012 Jun 21 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Number | PART 3 |
| Volume | 7335 LNCS |
| ISSN (Print) | 03029743 |
| ISSN (Electronic) | 16113349 |
Other
| Other | 12th International Conference on Computational Science and Its Applications, ICCSA 2012 |
|---|---|
| Country | Brazil |
| City | Salvador de Bahia |
| Period | 12/6/18 → 12/6/21 |
Fingerprint
Keywords
- Cache Performance
- Cache Pollution
- I/O Intensive Task
- Last Level Cache
- NUMA Scheduling
- Task Characteristics
ASJC Scopus subject areas
- Computer Science(all)
- Theoretical Computer Science
Cite this
Reducing last level cache pollution in NUMA multicore systems for improving cache performance. / An, Deukhyeon; Kim, Jeehong; Han, Junghyun; Eom, Young Ik.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7335 LNCS PART 3. ed. 2012. p. 272-282 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7335 LNCS, No. PART 3).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Reducing last level cache pollution in NUMA multicore systems for improving cache performance
AU - An, Deukhyeon
AU - Kim, Jeehong
AU - Han, Junghyun
AU - Eom, Young Ik
PY - 2012/7/23
Y1 - 2012/7/23
N2 - Non-uniform memory architecture (NUMA) system has numerous nodes with shared last level cache (LLC). Their shared LLC has brought many benefits in the cache utilization. However, LLC can be seriously polluted by tasks that cause huge I/O traffic for a long time since inclusive cache architecture of LLC replaces valid cache line by back-invalidate. Many research on the page coloring, partitioning, and pollute buffer mechanism handled this cache pollution. But, there are no scheduling approaches considering I/O-intensive tasks in NUMA systems. To address the above problem, OS scheduling that reduces cache pollution is highly needed in NUMA systems. In this paper, we propose a software-based mechanism that reduces shared LLC miss in NUMA systems. Our mechanism includes I/O traffic measurement and devil conscious scheduling. The experimental results show that LLC miss rate can be reduced up to 37.6%, and our approach improves execution time to 1.48%.
AB - Non-uniform memory architecture (NUMA) system has numerous nodes with shared last level cache (LLC). Their shared LLC has brought many benefits in the cache utilization. However, LLC can be seriously polluted by tasks that cause huge I/O traffic for a long time since inclusive cache architecture of LLC replaces valid cache line by back-invalidate. Many research on the page coloring, partitioning, and pollute buffer mechanism handled this cache pollution. But, there are no scheduling approaches considering I/O-intensive tasks in NUMA systems. To address the above problem, OS scheduling that reduces cache pollution is highly needed in NUMA systems. In this paper, we propose a software-based mechanism that reduces shared LLC miss in NUMA systems. Our mechanism includes I/O traffic measurement and devil conscious scheduling. The experimental results show that LLC miss rate can be reduced up to 37.6%, and our approach improves execution time to 1.48%.
KW - Cache Performance
KW - Cache Pollution
KW - I/O Intensive Task
KW - Last Level Cache
KW - NUMA Scheduling
KW - Task Characteristics
UR - http://www.scopus.com/inward/record.url?scp=84863952156&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84863952156&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-31137-6_21
DO - 10.1007/978-3-642-31137-6_21
M3 - Conference contribution
AN - SCOPUS:84863952156
SN - 9783642311369
VL - 7335 LNCS
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 272
EP - 282
BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ER -