BioMed Research International / 2014 / Article / Tab 1

Research Article

A Performance/Cost Evaluation for a GPU-Based Drug Discovery Application on Volunteer Computing

Table 1

Hardware features for our local test-bed infrastructure.

Intel system

ProcessorIntel Xeon E5620 @ 2.4 GHz
GPU 0Nvidia 7300GT
Memory16 GB DDR3 @ 1333 MHz
Maximum power draw80 W
Experimental idle power 38 W

GPU 1: Nvidia GTX 465
GPU familyGF100
Manufacturing process40 nm
Core clock607 MHz
Memory size1024 MB
Memory clock 2 × 1603 MHz
Memory bus width256 bits
Memory bandwidth102.6 GB/sec
Stream processors352
Maximum power draw200 W
Experimental idle power24 W

GPU 2: Nvidia GTX 480
GPU familyGF100
Manufacturing process40 nm.
Core clock700 MHz
Memory size1536 MB
Memory clock2 × 1848 MHz
Memory bus width384 bits
Memory bandwidth177.4 GB/sec
Stream processors480
Maximum power draw250 W
Experimental idle power37 W

GPU 3: Nvidia Tesla C2070
GPU familyGF100
Process40 nm.
Core clock573.5 MHz
Memory size6143 MB
Memory clock2 × 1494 MHz
Memory bus width384 bits
Memory bandwidth143.4 GB/sec
Stream processors448
Maximum power draw247 W
Experimental idle power107 W

GPU 4: Nvidia GTX 590
GPU familyGF100
Manufacturing process40 nm.
Core clock1215 MHz
Memory size2 × 1536 MB
Memory clock2 × 1707 MHz
Memory bus width 2 × 384 bits
Memory bandwidth2 × 327.7 GB/sec
Stream processors1024
Maximum power draw365 W
Experimental idle power140 W

GPU 5: Nvidia Tesla K20c
GPU familyGK110
Manufacturing process28 nm.
Core clock705 MHz
Memory size5120 MB
Memory clock2 × 2600 MHz
Memory bus width320 bits
Memory bandwidth:208 GB/sec
Stream processors2496
Maximum power draw225 W
Experimental idle power27 W

Article of the Year Award: Outstanding research contributions of 2020, as selected by our Chief Editors. Read the winning articles.