|
E-prints UPC >
Altres >
Enviament des de DRAC >
Empreu aquest identificador per citar o enllaçar aquest ítem:
http://hdl.handle.net/2117/13283
|
| Citació: | Alastruey, J. [et al.]. Selection of the register file size and the resource policy on SMT processors. A: International Symposium on Computer Architecture and High Performance Computing. "20th International Symposium on Computer Architecture and High Performance Computing". Campo Grande: IEEE Computer Society, 2008, p. 63-70. |
| Títol: | Selection of the register file size and the resource policy on SMT processors |
| Autor: | Alastruey, Jesús; Monreal Arnal, Teresa ; Cazorla Almeida, Francisco Javier ; Viñals Yufera, Víctor ; Valero Cortés, Mateo  |
| Editorial: | IEEE Computer Society |
| Data: | 2008 |
| Tipus de document: | Conference lecture |
| Resum: | The performance impact of the Physical Register File (PRF) size on Simultaneous Multithreading processors has not been extensively studied in spite of being a critical shared resource. In this paper we analyze the effect on performance of the PRF size for a broad set of resource allocation policies (Icount, Stall, Flush, Flush++, Static, Dcra and Hill-climbing) and evaluate them under two metrics: instructions per second (IPS) for throughput and harmonic mean of weighted IPCs (Hmean-wIPC) for fairness. We have found that resource allocation policy and PRF size should be considered together in order to obtain
the best score in the proposed metrics. For instance, for the analyzed 2 and 4-threaded SPEC CPU2000 workloads, small PRFs are best managed by Flush, whereas for larger PRFs, Hill-climbing and Static lead to the best values for the throughput and fairness metrics, respectively. The second contribution of this work is a simple procedure that, for a given resource allocation policy, selects the PRF size that maximizes IPS and obtains for HmeanwIPC
a value close to its maximum. According to our results, Hill-climbing with a 320-entry PRF achieves the best figures for 2-threaded workloads. When executing 4-threaded workloads, Hill-Climbing with a 384-entry PRF
achieves the best throughput whereas Static obtains the best throughput-fairness balance. |
| ISBN: | 9780769534237 |
| URI: | http://hdl.handle.net/2117/13283 |
| Versió de l'editor: | 10.1109/SBAC-PAD.2008.17 |
| Apareix a les col·leccions: | Altres. Enviament des de DRAC Departament d'Arquitectura de Computadors. Ponències/Comunicacions de congressos CAP - Grup de Computació d´Altes Prestacions. Ponències/Comunicacions de congressos
|
| Comparteix: |
|
Queda prohibida la reproducció, transformació, distribució i comunicació pública d'aquesta obra. Es permet, en tot cas, la reproducció per a ús privat sempre i quan la còpia que se'n faci no sigui objecte d'utilització col·lectiva ni lucrativa (art. 31.2 del Reial Decret Legislatiu 1/1996, de 12 d'abril, pel qual s'aprova el Text Refós de la Llei de Propietat Intel·lectual, http://bibliotecnica.upc.es/sepi/legislacio.asp).
Per a qualsevol ús que es vulgui fer diferent al permès, dirigiu-vos a: sepi@upc.edu
|