Abstract
The presence of FPGAS in data centers has been growing due to their superior performance as accelerators. Thermal management, particularly battling the cooling cost in these high performance systems, is a primary concern. Introduction of new heterogeneous components only adds further complexities to thermal modeling and management. The thermal behavior of multi-FPGA systems deployed within large compute clusters is little explored. In this paper, we first show that the thermal behaviors of different FPGAS of the same generation can vary due to their physical locations in a rack and process variation, even though they are running the same tasks. We present a machine learning based model to capture the thermal behavior of a multi-node FPGA cluster. We then propose to mitigate thermal variation and hotspots across the cluster by proactive task placement guided by our thermal model. Our experiments show that through proper placement of tasks on the multi-FPGA system, we can reduce the peak temperature by up to 11.50°C with no impact on performance.
| Original language | English (US) |
|---|---|
| Title of host publication | Proceedings - 2018 IEEE 36th International Conference on Computer Design, ICCD 2018 |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 537-544 |
| Number of pages | 8 |
| ISBN (Electronic) | 9781538684771 |
| DOIs | |
| State | Published - Jan 16 2019 |
| Event | 36th International Conference on Computer Design, ICCD 2018 - Orlando, United States Duration: Oct 7 2018 → Oct 10 2018 |
Publication series
| Name | Proceedings - 2018 IEEE 36th International Conference on Computer Design, ICCD 2018 |
|---|
Conference
| Conference | 36th International Conference on Computer Design, ICCD 2018 |
|---|---|
| Country | United States |
| City | Orlando |
| Period | 10/7/18 → 10/10/18 |
Fingerprint
Keywords
- HPC
- Task Placement
- Thermal Modeling
ASJC Scopus subject areas
- Computer Networks and Communications
- Hardware and Architecture
- Safety, Risk, Reliability and Quality
Cite this
}
Minimizing Thermal Variation in Heterogeneous HPC Systems with FPGA Nodes. / Luo, Yingyi; Wang, Xiaoyang; Memik, Seda Ogrenci; Memik, Gokhan; Yoshii, Kazutomo; Beckman, Pete.
Proceedings - 2018 IEEE 36th International Conference on Computer Design, ICCD 2018. Institute of Electrical and Electronics Engineers Inc., 2019. p. 537-544 8615736 (Proceedings - 2018 IEEE 36th International Conference on Computer Design, ICCD 2018).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - Minimizing Thermal Variation in Heterogeneous HPC Systems with FPGA Nodes
AU - Luo, Yingyi
AU - Wang, Xiaoyang
AU - Memik, Seda Ogrenci
AU - Memik, Gokhan
AU - Yoshii, Kazutomo
AU - Beckman, Pete
PY - 2019/1/16
Y1 - 2019/1/16
N2 - The presence of FPGAS in data centers has been growing due to their superior performance as accelerators. Thermal management, particularly battling the cooling cost in these high performance systems, is a primary concern. Introduction of new heterogeneous components only adds further complexities to thermal modeling and management. The thermal behavior of multi-FPGA systems deployed within large compute clusters is little explored. In this paper, we first show that the thermal behaviors of different FPGAS of the same generation can vary due to their physical locations in a rack and process variation, even though they are running the same tasks. We present a machine learning based model to capture the thermal behavior of a multi-node FPGA cluster. We then propose to mitigate thermal variation and hotspots across the cluster by proactive task placement guided by our thermal model. Our experiments show that through proper placement of tasks on the multi-FPGA system, we can reduce the peak temperature by up to 11.50°C with no impact on performance.
AB - The presence of FPGAS in data centers has been growing due to their superior performance as accelerators. Thermal management, particularly battling the cooling cost in these high performance systems, is a primary concern. Introduction of new heterogeneous components only adds further complexities to thermal modeling and management. The thermal behavior of multi-FPGA systems deployed within large compute clusters is little explored. In this paper, we first show that the thermal behaviors of different FPGAS of the same generation can vary due to their physical locations in a rack and process variation, even though they are running the same tasks. We present a machine learning based model to capture the thermal behavior of a multi-node FPGA cluster. We then propose to mitigate thermal variation and hotspots across the cluster by proactive task placement guided by our thermal model. Our experiments show that through proper placement of tasks on the multi-FPGA system, we can reduce the peak temperature by up to 11.50°C with no impact on performance.
KW - HPC
KW - Task Placement
KW - Thermal Modeling
UR - http://www.scopus.com/inward/record.url?scp=85062238722&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85062238722&partnerID=8YFLogxK
U2 - 10.1109/ICCD.2018.00086
DO - 10.1109/ICCD.2018.00086
M3 - Conference contribution
T3 - Proceedings - 2018 IEEE 36th International Conference on Computer Design, ICCD 2018
SP - 537
EP - 544
BT - Proceedings - 2018 IEEE 36th International Conference on Computer Design, ICCD 2018
PB - Institute of Electrical and Electronics Engineers Inc.
ER -