Abstract | ||
---|---|---|
In this work we propose a shared floating point unit (FPU) architecture for ultra low power (ULP) system on chips operating at near threshold voltage (NTV). Since high-performance FP units (FPUs) are large and complex, but their utilization is relatively low, adding one FPU per each core in a ULP multicore is costly and power hungry. In our approach, we share a few FPUs among all the cores in the system. This increases the utilization of FPUs leading to an energy-efficient design. As a part of our approach, we propose two different FPU allocation techniques: optimal and random. Experimental results demonstrate that compared to a traditional private-FPU approach, our technique in a multicore system with 8 processors and 2 shared FPUs can increase the performance/(area*power) by 5x for applications with 10% FP operations and by 2.5x for applications with 25% FP operations. |
Year | DOI | Venue |
---|---|---|
2013 | 10.1145/2482767.2482772 | Conf. Computing Frontiers |
Keywords | Field | DocType |
ultra low power,ulp multicore,shared fpus,ultra-low power mpsocs,fp operation,traditional private-fpu approach,shared-fpu architecture,high-performance fp unit,multicore system,energy-efficient design,different fpu allocation technique,multicore | Architecture,Floating-point unit,Computer science,Parallel computing,Real-time computing,Multi-core processor,Threshold voltage,Embedded system | Conference |
Citations | PageRank | References |
0 | 0.34 | 7 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Mohammad Reza Kakoee | 1 | 87 | 8.68 |
Igor Loi | 2 | 445 | 30.66 |
Luca Benini | 3 | 13116 | 1188.49 |