Abstract | ||
---|---|---|
Spatially-correlated intra-die process variations result in significant core-to-core frequency variations in chip-multiprocessors. An analytical model for frequency island chip-multiprocessor throughput is introduced. The improved variability-tolerance of FI-CMPs over their globally-clocked counterparts is quantified across a range of core counts and sizes under constant die area. The benefits are highest for designs consisting of many small cores, with the throughput of a globally-clocked design with 70 small cores increasing by 8.8% when per-core frequency islands are used. The small- core FI-CMP also loses only 7.2% of its nominal performance to process variations, the least among any of the designs. |
Year | DOI | Venue |
---|---|---|
2008 | 10.1145/1391469.1391550 | Anaheim, CA |
Keywords | Field | DocType |
analytical model,chip-multiprocessor variability-tolerance,core count,globally-clocked counterpart,significant core-to-core frequency variation,per-core frequency island,constant die area,globally-clocked design,spatially-correlated intra-die process variation,frequency island chip-multiprocessor throughput,small core,throughput,monte carlo methods,chip,correlation,logic design,process design,frequency control,fault tolerance,mathematical model,process variation,spatial correlation,displays,monte carlo analysis | Logic synthesis,Monte Carlo method,Computer science,Parallel computing,Automatic frequency control,Electronic engineering,Multiprocessing,Real-time computing,Chip,Fault tolerance,Process design,Throughput | Conference |
ISSN | ISBN | Citations |
0738-100X | 978-1-60558-115-6 | 39 |
PageRank | References | Authors |
1.99 | 10 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Sebastian Herbert | 1 | 256 | 9.83 |
Diana Marculescu | 2 | 2725 | 223.87 |