Speed Factor of GPU Variants
How fast is a GPU when compared to others? Figure it out here.
Boost for Daz features a curated range of GPU based on cost and speed. The GPU speed is calculated using a metric called Iteration Rate, which is essentially the number of iterations the GPU can do per second.
In the table below, you can find the Speed Factor (the multiple) of a particular variant compared to the base (single) GPU in that family. For example, using the table below you'll see that the speed of a A6000x3 is 2.7 times faster than the speed of a single A6000.
These calculations use the RayDant Benchmark file (available on Daz forums) and are taken at 1800 iterations and convergence set to 95%.
NVIDIA RTX 4000
6.143
base
293
NVIDIA RTX 4000 x2
11.111
1.8
162
NVIDIA RTX 4000 x3
16.667
2.7
108
NVIDIA RTX 4000 x4
19.149
3.1
94
NVIDIA RTX 5000
6.792
base
265
NVIDIA RTX 5000 x2
12.081
1.8
149
NVIDIA RTX 5000 x3
17.308
2.5
104
NVIDIA RTX A4000
10.651
base
169
NVIDIA RTX A4000 x2
20.455
1.9
88
NVIDIA RTX A4000 x3
26.087
2.4
69
NVIDIA RTX A4000 x4
34.615
3.3
52
NVIDIA Tesla V100 x1
9.626
base
187
NVIDIA Tesla V100 x2
15.385
1.6
117
NVIDIA RTX A5000
15.929
base
113
NVIDIA RTX A5000 x2
32.143
2.0
56
NVIDIA RTX A5000 x3
38.298
2.4
47
NVIDIA RTX A5000 x4
48.649
3.1
37
NVIDIA RTX A6000
17.308
base
104
NVIDIA RTX A6000 x2
32.727
1.9
55
NVIDIA RTX A6000 x3
46.154
2.7
39
NVIDIA RTX A40 x1
15.254
base
118
NVIDIA RTX A40 x2
28.125
1.8
64
NVIDIA RTX A40 x3
40.000
2.6
45
NVIDIA RTX A40 x4
54.545
3.6
33
Last updated