High-performance throughput computing

被引：39

作者：

Chaudhry, S ^{[1
]}

Caprioli, P ^{[1
]}

Yip, S ^{[1
]}

Tremblay, M ^{[1
]}

机构：

[1] Sun Microsyst Inc, Sunnyvale, CA 94085 USA

来源：

IEEE MICRO | 2005年 / 25卷 / 03期

关键词：

D O I：

10.1109/MM.2005.49

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

Throughput computing, achieved through multithreading and multicore technology, can lead to performance improvements that are 10 to 30x those of conventional processors and systems. However, such systems should also offer good single-thread performance. Here, the authors show that hardware scouting increases the performance of an already robust core by up to 40 percent for commercial benchmarks.

引用

页码：32 / 45

页数：14

共 21 条

[1]

Balasubramonian R, 2001, CONF PROC INT SYMP C, P26, DOI 10.1109/ISCA.2001.937428

[2]

CHAUDHRY S, Patent No. 6415356

[3] Microarchitecture optimizations for exploiting memory-level parallelism [J].

Chou, Y ;

Fahs, B ;

Abraham, S .

31ST ANNUAL INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE, PROCEEDINGS, 2004, :76-87

[4]

Clabes J, 2004, DES AUT CON, P670

[5] Dynamic speculative precomputation [J].

Collins, JD ;

Tullsen, DM ;

Wang, H ;

Shen, JP .

34TH ACM/IEEE INTERNATIONAL SYMPOSIUM ON MICROARCHITECTURE, MICRO-34, PROCEEDINGS, 2001, :306-317

[6] Out-of-order commit processors [J].

Cristal, A ;

Ortega, D ;

Llosa, J ;

Valero, M .

10TH INTERNATIONAL SYMPOSIUM ON HIGH PERFORMANCE COMPUTER ARCHITECTURE, PROCEEDINGS, 2004, :48-59

[7]

DUNDAS J, 1997, P 11 INT C SUP, P68

[8]

Flautner K, 2000, ACM SIGPLAN NOTICES, V35, P129, DOI 10.1145/384264.379233

[9]

GREENLEY D, 2004, MICR FOR IN STAT

[10]

JACOBSON Q, 2003, ULTRASPARC 4 PROCESS

← 1 2 3 →