GPU Performance Details: GeForce GTX 960
System Configuration
MATLAB Release: R2016b
|
Host
| Name | Intel(R) Core(TM) i5-6400T CPU @ 2.20GHz |
|---|
| Clock | 2201 MHz |
|---|
| Cache | 1024 KB |
|---|
| NumProcessors | 4 |
|---|
| OSType | Windows |
|---|
| OSVersion | Microsoft Windows 10 Home |
|---|
|
GPU
| Name | GeForce GTX 960 |
|---|
| Clock | 1.200500e+03 MHz |
|---|
| NumProcessors | 8 |
|---|
| ComputeCapability | 5.2 |
|---|
| TotalMemory | 4.00 GB |
|---|
| CUDAVersion | 8 |
|---|
| DriverVersion | 6.14.13.6930 (369.30) |
|---|
|
Results for MTimes (double)
These results show the performance of the GPU or host PC when calculating a
matrix multiplication
of two NxN real matrices. The number of operations is assumed to be
2*N^3 - N^2.
This calculation is usually compute-bound, i.e. the performance depends mainly
on how fast the GPU or host PC can perform floating-point operations.
|
Raw data for GeForce GTX 960 - MTimes (double)
|
Array size
(elements) | Num
Operations | Time
(ms) | GigaFLOPS |
|---|
| 1,024 | 64,512 | 0.07 | 0.91 |
| 4,096 | 520,192 | 0.06 | 8.38 |
| 16,384 | 4,177,920 | 0.14 | 30.32 |
| 65,536 | 33,488,896 | 0.62 | 53.68 |
| 262,144 | 268,173,312 | 4.60 | 58.32 |
| 1,048,576 | 2,146,435,072 | 32.42 | 66.20 |
| 4,194,304 | 17,175,674,880 | 252.31 | 68.07 |
(N gigaflops = Nx109 operations per second)
|
|
|
Results for Backslash (double)
These results show the performance of the GPU or host PC when calculating the
matrix left division
of an NxN matrix with an Nx1 vector. The number of operations
is assumed to be 2/3*N^3 + 3/2*N^2.
This calculation is usually compute-bound, i.e. the performance depends
mainly on how fast the GPU or host PC can perform floating-point operations.
|
Raw data for GeForce GTX 960 - Backslash (double)
|
Array size
(elements) | Num
Operations | Time
(ms) | GigaFLOPS |
|---|
| 1,024 | 23,381 | 0.94 | 0.02 |
| 4,096 | 180,907 | 0.96 | 0.19 |
| 16,384 | 1,422,677 | 1.62 | 0.88 |
| 65,536 | 11,283,115 | 3.51 | 3.21 |
| 262,144 | 89,871,701 | 7.57 | 11.88 |
| 1,048,576 | 717,400,747 | 22.80 | 31.46 |
| 4,194,304 | 5,732,914,517 | 117.32 | 48.87 |
(N gigaflops = Nx109 operations per second)
|
|
|
Results for FFT (double)
These results show the performance of the GPU or host PC when calculating the
Fast-Fourier-Transform
of a vector of complex numbers. The number of operations for a vector
of length N is assumed to be 5*N*log2(N).
This calculation is usually memory-bound, i.e. the performance depends mainly
on how fast the GPU or host PC can read and write data.
|
Raw data for GeForce GTX 960 - FFT (double)
|
Array size
(elements) | Num
Operations | Time
(ms) | GigaFLOPS |
|---|
| 1,024 | 51,200 | 0.09 | 0.56 |
| 4,096 | 245,760 | 0.12 | 2.03 |
| 16,384 | 1,146,880 | 0.14 | 8.36 |
| 65,536 | 5,242,880 | 0.33 | 15.79 |
| 262,144 | 23,592,960 | 0.84 | 28.17 |
| 1,048,576 | 104,857,600 | 3.04 | 34.49 |
| 4,194,304 | 461,373,440 | 12.31 | 37.49 |
| 16,777,216 | 2,013,265,920 | 54.61 | 36.87 |
(N gigaflops = Nx109 operations per second)
|
|
|
Results for MTimes (single)
These results show the performance of the GPU or host PC when calculating a
matrix multiplication
of two NxN real matrices. The number of operations is assumed to be
2*N^3 - N^2.
This calculation is usually compute-bound, i.e. the performance depends mainly
on how fast the GPU or host PC can perform floating-point operations.
|
Raw data for GeForce GTX 960 - MTimes (single)
|
Array size
(elements) | Num
Operations | Time
(ms) | GigaFLOPS |
|---|
| 1,024 | 64,512 | 0.08 | 0.79 |
| 4,096 | 520,192 | 0.07 | 7.19 |
| 16,384 | 4,177,920 | 0.07 | 60.49 |
| 65,536 | 33,488,896 | 0.09 | 362.21 |
| 262,144 | 268,173,312 | 0.21 | 1288.35 |
| 1,048,576 | 2,146,435,072 | 1.12 | 1924.36 |
| 4,194,304 | 17,175,674,880 | 7.78 | 2206.56 |
| 16,777,216 | 137,422,176,256 | 60.41 | 2274.96 |
| 67,108,864 | 1,099,444,518,912 | 484.59 | 2268.84 |
(N gigaflops = Nx109 operations per second)
|
|
|
Results for Backslash (single)
These results show the performance of the GPU or host PC when calculating the
matrix left division
of an NxN matrix with an Nx1 vector. The number of operations
is assumed to be 2/3*N^3 + 3/2*N^2.
This calculation is usually compute-bound, i.e. the performance depends
mainly on how fast the GPU or host PC can perform floating-point operations.
|
Raw data for GeForce GTX 960 - Backslash (single)
|
Array size
(elements) | Num
Operations | Time
(ms) | GigaFLOPS |
|---|
| 1,024 | 23,381 | 0.79 | 0.03 |
| 4,096 | 180,907 | 0.93 | 0.19 |
| 16,384 | 1,422,677 | 1.25 | 1.14 |
| 65,536 | 11,283,115 | 1.34 | 8.43 |
| 262,144 | 89,871,701 | 4.95 | 18.14 |
| 1,048,576 | 717,400,747 | 9.66 | 74.25 |
| 4,194,304 | 5,732,914,517 | 23.84 | 240.44 |
| 16,777,216 | 45,838,150,315 | 98.34 | 466.14 |
| 67,108,864 | 366,604,539,221 | 557.27 | 657.86 |
(N gigaflops = Nx109 operations per second)
|
|
|
Results for FFT (single)
These results show the performance of the GPU or host PC when calculating the
Fast-Fourier-Transform
of a vector of complex numbers. The number of operations for a vector
of length N is assumed to be 5*N*log2(N).
This calculation is usually memory-bound, i.e. the performance depends mainly
on how fast the GPU or host PC can read and write data.
|
Raw data for GeForce GTX 960 - FFT (single)
|
Array size
(elements) | Num
Operations | Time
(ms) | GigaFLOPS |
|---|
| 1,024 | 51,200 | 0.08 | 0.61 |
| 4,096 | 245,760 | 0.09 | 2.74 |
| 16,384 | 1,146,880 | 0.12 | 9.53 |
| 65,536 | 5,242,880 | 0.14 | 36.84 |
| 262,144 | 23,592,960 | 0.33 | 70.49 |
| 1,048,576 | 104,857,600 | 1.04 | 100.66 |
| 4,194,304 | 461,373,440 | 3.37 | 136.95 |
| 16,777,216 | 2,013,265,920 | 12.88 | 156.26 |
(N gigaflops = Nx109 operations per second)
|
|
|
|