Abstract
The article proposes a selective compressed memory system (SCMS) focusing on a compressed cache architecture, in which only data blocks with good compression efficiency are compressed selectively and all compressed blocks are stored in a fixed memory space. The selective compression technique can reduce decompression overhead caused by online data decompression and also the fixed memory space allocation allows efficient management of the compressed blocks. The results from a trace driven simulation show that the SCMS approach can provide around a 35% decrease in the on-chip cache miss ratio as well as a 53% decrease in the data traffic over conventional memory systems. Furthermore, a large amount of the decompression overhead can be reduced, and thus the average memory access time can also be reduced by a maximum 20% against conventional memory systems.
| Original language | English |
|---|---|
| Title of host publication | Proceedings - 25th EUROMICRO Conference on Informatics |
| Subtitle of host publication | Theory and Practice for the New Millennium, EUROMICRO 1999 |
| Pages | 224-227 |
| Number of pages | 4 |
| Volume | 1 |
| DOIs | |
| Publication status | Published - 1999 Dec 1 |
| Event | 25th EUROMICRO Conference on Informatics: Theory and Practice for the New Millennium, EUROMICRO 1999 - Milan, Italy Duration: 1999 Sep 8 → 1999 Sep 10 |
Other
| Other | 25th EUROMICRO Conference on Informatics: Theory and Practice for the New Millennium, EUROMICRO 1999 |
|---|---|
| Country | Italy |
| City | Milan |
| Period | 99/9/8 → 99/9/10 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Control and Systems Engineering
- Hardware and Architecture
Cite this
}
A selective compressed memory system by on-line data decompressing. / Lee, Jang Soo; Hong, Won Kee; Kim, Shin-Dug.
Proceedings - 25th EUROMICRO Conference on Informatics: Theory and Practice for the New Millennium, EUROMICRO 1999. Vol. 1 1999. p. 224-227 794470.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - A selective compressed memory system by on-line data decompressing
AU - Lee, Jang Soo
AU - Hong, Won Kee
AU - Kim, Shin-Dug
PY - 1999/12/1
Y1 - 1999/12/1
N2 - The article proposes a selective compressed memory system (SCMS) focusing on a compressed cache architecture, in which only data blocks with good compression efficiency are compressed selectively and all compressed blocks are stored in a fixed memory space. The selective compression technique can reduce decompression overhead caused by online data decompression and also the fixed memory space allocation allows efficient management of the compressed blocks. The results from a trace driven simulation show that the SCMS approach can provide around a 35% decrease in the on-chip cache miss ratio as well as a 53% decrease in the data traffic over conventional memory systems. Furthermore, a large amount of the decompression overhead can be reduced, and thus the average memory access time can also be reduced by a maximum 20% against conventional memory systems.
AB - The article proposes a selective compressed memory system (SCMS) focusing on a compressed cache architecture, in which only data blocks with good compression efficiency are compressed selectively and all compressed blocks are stored in a fixed memory space. The selective compression technique can reduce decompression overhead caused by online data decompression and also the fixed memory space allocation allows efficient management of the compressed blocks. The results from a trace driven simulation show that the SCMS approach can provide around a 35% decrease in the on-chip cache miss ratio as well as a 53% decrease in the data traffic over conventional memory systems. Furthermore, a large amount of the decompression overhead can be reduced, and thus the average memory access time can also be reduced by a maximum 20% against conventional memory systems.
UR - http://www.scopus.com/inward/record.url?scp=84888994889&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84888994889&partnerID=8YFLogxK
U2 - 10.1109/EURMIC.1999.794470
DO - 10.1109/EURMIC.1999.794470
M3 - Conference contribution
AN - SCOPUS:84888994889
SN - 0769503217
SN - 9780769503219
VL - 1
SP - 224
EP - 227
BT - Proceedings - 25th EUROMICRO Conference on Informatics
ER -