|
| This Article | ||
| ||
| Share | ||
| Bibliographic References | ||
| Add to: | ||
| | ||
| Search | ||
| ||
| ASCII Text | x | ||
| Jaume Abella, Antonio Gonz?lez, "Low-Complexity Distributed Issue Queue," 2013 IEEE 19th International Symposium on High Performance Computer Architecture (HPCA), pp. 73, 10th International Symposium on High Performance Computer Architecture (HPCA'04), 2004. | |||
| BibTex | x | ||
| @article{ 10.1109/HPCA.2004.10013, author = {Jaume Abella and Antonio Gonz?lez}, title = {Low-Complexity Distributed Issue Queue}, journal ={2013 IEEE 19th International Symposium on High Performance Computer Architecture (HPCA)}, volume = {0}, year = {2004}, issn = {1530-0897}, pages = {73}, doi = {http://doi.ieeecomputersociety.org/10.1109/HPCA.2004.10013}, publisher = {IEEE Computer Society}, address = {Los Alamitos, CA, USA}, } | |||
| RefWorks Procite/RefMan/Endnote | x | ||
| TY - CONF JO - 2013 IEEE 19th International Symposium on High Performance Computer Architecture (HPCA) TI - Low-Complexity Distributed Issue Queue SN - 1530-0897 SP EP A1 - Jaume Abella, A1 - Antonio Gonz?lez, PY - 2004 KW - null VL - 0 JA - 2013 IEEE 19th International Symposium on High Performance Computer Architecture (HPCA) ER - | |||
As technology evolves, power density significantly increases and cooling systems become more complex and expensive. The issue logic is one of the processor hotspots and, at the same time, its latency is crucial for the processor performance.
This paper presents a low-complexity FP issue logic (MB_distr) that achieves high performance with small energy requirements. The MB_distr scheme is based on classifying instructions and dispatching them into a set of queues depending on their data dependences. These instructions are selected for issuing based on an estimation of when their operands will be available, so the conventional wakeup activity is not required. Additionally, the functional units are distributed across the different queues.
The energy required by the proposed scheme is substantially lower than that required by a conventional issue design, even if the latter has the ability of waking-up only unready operands. MB_distr scheme reduces the energy-delay2 product by 35% and the energy-delay product by 18% with respect to a state-of-the-art approach.
