Abstract
Parallel multithreaded architectures take advantage of the ability to execute more than one thread simultaneously on a single chip at low synchronisation and communication costs and high hardware resource utilisation. However, a high bandwidth cache, such as a multibank cache, is especially critical to serve memory accesses issued at the same time from different threads. To prevent bank conflicts of multibank cache from seriously degrading system performance, a loop partition method is proposed to reduce or even eliminate bank conflicts. The partition allows each thread access to certain bank modules and prevents any two from accessing the same bank module. The method neither slows down the clock rate nor increases the array subscript expression complexity. The performance gains of the bank-conflict-free loop partition approach are shown in simulation results.
| Original language | English |
|---|---|
| Pages (from-to) | 30-36 |
| Number of pages | 7 |
| Journal | IEE Proceedings: Computers and Digital Techniques |
| Volume | 143 |
| Issue number | 1 |
| DOIs | |
| Publication status | Published - 1996 Jan 1 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Theoretical Computer Science
- Hardware and Architecture
- Computational Theory and Mathematics
Cite this
}
A loop partition technique for reducing cache bank conflict in multithreaded architecture. / Wu, Chao-Chin; Chen, C.
In: IEE Proceedings: Computers and Digital Techniques, Vol. 143, No. 1, 01.01.1996, p. 30-36.Research output: Contribution to journal › Article
TY - JOUR
T1 - A loop partition technique for reducing cache bank conflict in multithreaded architecture
AU - Wu, Chao-Chin
AU - Chen, C.
PY - 1996/1/1
Y1 - 1996/1/1
N2 - Parallel multithreaded architectures take advantage of the ability to execute more than one thread simultaneously on a single chip at low synchronisation and communication costs and high hardware resource utilisation. However, a high bandwidth cache, such as a multibank cache, is especially critical to serve memory accesses issued at the same time from different threads. To prevent bank conflicts of multibank cache from seriously degrading system performance, a loop partition method is proposed to reduce or even eliminate bank conflicts. The partition allows each thread access to certain bank modules and prevents any two from accessing the same bank module. The method neither slows down the clock rate nor increases the array subscript expression complexity. The performance gains of the bank-conflict-free loop partition approach are shown in simulation results.
AB - Parallel multithreaded architectures take advantage of the ability to execute more than one thread simultaneously on a single chip at low synchronisation and communication costs and high hardware resource utilisation. However, a high bandwidth cache, such as a multibank cache, is especially critical to serve memory accesses issued at the same time from different threads. To prevent bank conflicts of multibank cache from seriously degrading system performance, a loop partition method is proposed to reduce or even eliminate bank conflicts. The partition allows each thread access to certain bank modules and prevents any two from accessing the same bank module. The method neither slows down the clock rate nor increases the array subscript expression complexity. The performance gains of the bank-conflict-free loop partition approach are shown in simulation results.
UR - http://www.scopus.com/inward/record.url?scp=0029779810&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0029779810&partnerID=8YFLogxK
U2 - 10.1049/ip-cdt:19960007
DO - 10.1049/ip-cdt:19960007
M3 - Article
AN - SCOPUS:0029779810
VL - 143
SP - 30
EP - 36
JO - IEE Proceedings: Computers and Digital Techniques
JF - IEE Proceedings: Computers and Digital Techniques
SN - 1350-2387
IS - 1
ER -