Difference between revisions of "Performance & Benchmark"

From CUVI Wiki
Line 11: Line 11:




<!--
 
<tabs>
<tabs>


<tab name="NX">
<tab name="RTX 2060">
{|class="wikitable"
{|class="wikitable"
|+ style="caption-side:bottom; color:#e76700;"|''NX with CUVI v1.8.0
|+ style="caption-side:bottom; color:#e76700;"|''Kernel Time in milliseconds (ms) with CUVI v1.8.0 on RTX 2060
|-
|-
! Algorithm / Image Size
! Algorithm / Image Size
Line 24: Line 24:
! 8k (7680x4320)
! 8k (7680x4320)
|-
|-
|-
|[[Function:Add|add]] - 2 Images
|0.06
|0.14
|0.51
|2.01
|-
|[[Function:ChannelMix|channelMix]]
|0.07
|0.14
|0.55
|2.25
|-
|[[Function:Demosaic|demosaic]]
|0.24
|0.53
|2.10
|8.10
|-
|[[Function:DemosaicDFPD|demosaicDFPD]]
|0.52
|1.22
|4.53
|18.1
|-
|[[Function:GammaCorrect|gammaCorrect]]
|0.12
|0.28
|1.02
|4.30
|-
|[[Function:HistEq|histEq]] - Single Channel
|0.21
|0.24
|0.84
|3.10
|-
|[[Function:LUT|LUT]]
|0.03
|0.08
|0.29
|1.20
|-
|[[Function:blackGammaLUT|blackGammaLUT]]
|0.069
|0.16
|0.61
|2.50
|-
|[[Function:RGB2Gray|rgb2gray]]
|0.04
|0.09
|0.34
|1.43
<!--
|-
|[[Function:GoodFeaturesToTrack|goodFeaturesToTrack]]
|26,720.11
|20,384.87
|11,828.72
|8,466.54
|-
|[[Function:opticalFlowPyrLKDense|opticalFlowPyrLKDense]] - 2 Images
|26,720.11
|20,384.87
|11,828.72
|8,466.54
|-
|[[Function:FocusStack|focusStack]] - 2 Images
|26,720.11
|20,384.87
|11,828.72
|8,466.54
-->
|-
|[[Function:BitConversion|bitConversion]] - From 8 to 16 bits
|0.01
|0.24
|0.95
|3.81
|-
|[[Function:Crop|crop]]
|0.04
|0.12
|0.41
|1.70
|-
|[[Function:Resize|resize]] - Scale=2.0
|0.25
|0.55
|2.21
|8.70
|-
|[[Function:Rotate|rotate]] - Non Cropping, Angle = -3.76f
|0.04
|0.09
|0.36
|1.11
|-
|[[Function:WarpPerspective|warpPerspective]]
|0.08
|0.20
|0.77
|3.10
|-
|[[Function:ImageFilter|imageFilter]] - 5x5 floating point window
|0.65
|1.56
|5.81
|13.7
|-
|[[Function:UnderwaterFilter|underwaterFilter]]
|0.53
|1.10
|4.00
|15.2
|-
|[[Function:haarFwd|haarFwd]]
|0.14
|0.30
|1.21
|4.90
|}
|}
</tab>
</tab>
-->
 
<tab name="RTX 2060">
<tab name="RTX 2060">
{|class="wikitable"
{|class="wikitable"
Line 176: Line 298:


</tabs>
</tabs>
-->

Revision as of 20:32, 23 October 2022

If one thing CUVI gives you, it's performance boost over competitive libraries and solutions. Using GPGPU as the underlying hardware, Imaging and Vision modules get maximum benefit due to their inherent parallel algorithms. In addition to cost cutting on CPU-based clusters, CUVI gives up to 15x speedup over Intel IPP.

File:CUVI Speedup.jpg

Applications using CUVI are generally ten times faster than their CPU counterpart. CUVI framework also gives the ease to scale the application on more than one GPU making it as fast as you want.

File:CUVI Bench.jpg

Benchmark

The following benchmark via NVIDIA's Performance tools for Windows and Linux. Timing figure represents frames per second (rounded) based on only the processing time on the single GPU. The benchmarks are performed on color 8-bit per channel images except where mentioned otherwise.


Kernel Time in milliseconds (ms) with CUVI v1.8.0 on RTX 2060
Algorithm / Image Size 720p 1080p 4k (3840x2160) 8k (7680x4320)
add - 2 Images 0.06 0.14 0.51 2.01
channelMix 0.07 0.14 0.55 2.25
demosaic 0.24 0.53 2.10 8.10
demosaicDFPD 0.52 1.22 4.53 18.1
gammaCorrect 0.12 0.28 1.02 4.30
histEq - Single Channel 0.21 0.24 0.84 3.10
LUT 0.03 0.08 0.29 1.20
blackGammaLUT 0.069 0.16 0.61 2.50
rgb2gray 0.04 0.09 0.34 1.43
bitConversion - From 8 to 16 bits 0.01 0.24 0.95 3.81
crop 0.04 0.12 0.41 1.70
resize - Scale=2.0 0.25 0.55 2.21 8.70
rotate - Non Cropping, Angle = -3.76f 0.04 0.09 0.36 1.11
warpPerspective 0.08 0.20 0.77 3.10
imageFilter - 5x5 floating point window 0.65 1.56 5.81 13.7
underwaterFilter 0.53 1.10 4.00 15.2
haarFwd 0.14 0.30 1.21 4.90
Kernel Time in milliseconds (ms) with CUVI v1.8.0 on RTX 2060
Algorithm / Image Size 720p 1080p 4k (3840x2160) 8k (7680x4320)
add - 2 Images 0.06 0.14 0.51 2.01
channelMix 0.07 0.14 0.55 2.25
demosaic 0.24 0.53 2.10 8.10
demosaicDFPD 0.52 1.22 4.53 18.1
gammaCorrect 0.12 0.28 1.02 4.30
histEq - Single Channel 0.21 0.24 0.84 3.10
LUT 0.03 0.08 0.29 1.20
blackGammaLUT 0.069 0.16 0.61 2.50
rgb2gray 0.04 0.09 0.34 1.43
bitConversion - From 8 to 16 bits 0.01 0.24 0.95 3.81
crop 0.04 0.12 0.41 1.70
resize - Scale=2.0 0.25 0.55 2.21 8.70
rotate - Non Cropping, Angle = -3.76f 0.04 0.09 0.36 1.11
warpPerspective 0.08 0.20 0.77 3.10
imageFilter - 5x5 floating point window 0.65 1.56 5.81 13.7
underwaterFilter 0.53 1.10 4.00 15.2
haarFwd 0.14 0.30 1.21 4.90
Retrieved from "https://wiki.cuvilib.com/index.php?title=Performance_%26_Benchmark&oldid=1671"