mirror/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2024-12-15 18:09:11 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity
34,013 Commits 6 Branches 127 Tags 3 GiB
9f123f8d74
Commit Graph

9 Commits

Author SHA1 Message Date
Yuantao Feng
3afe8ddaf8
core: Rename cv::float16_t to cv::hfloat (#25217) 
* rename cv::float16_t to cv::fp16_t

* add typedef fp16_t float16_t

* remove zero(), bits() from fp16_t class

* fp16_t -> hfloat

* remove cv::float16_t::fromBits; add hfloatFromBits

* undo changes in conv_winograd_f63.simd.hpp and conv_block.simd.hpp

* undo some changes in dnn
 2024-03-21 23:44:19 +03:00
Alexander Alekhin
40533dbf69
Merge pull request #24918 from opencv-pushbot:gitee/alalek/core_convertfp16_replacement 
core(OpenCL): optimize convertTo() with CV_16F (convertFp16() replacement) #24918

relates #24909
relates #24917
relates #24892

Performance changes:

- [x] 12700K (1 thread) + Intel iGPU

|Name of Test|noOCL|convertFp16|convertTo BASE|convertTo PATCH|
|---|:-:|:-:|:-:|:-:|
|ConvertFP16FP32MatMat::OCL_Core|3.130|3.152|3.127|3.136|
|ConvertFP16FP32MatUMat::OCL_Core|3.030|3.996|3.007|2.671|
|ConvertFP16FP32UMatMat::OCL_Core|3.010|3.101|3.056|2.854|
|ConvertFP16FP32UMatUMat::OCL_Core|3.016|3.298|2.072|2.061|
|ConvertFP32FP16MatMat::OCL_Core|2.697|2.652|2.723|2.721|
|ConvertFP32FP16MatUMat::OCL_Core|2.752|4.268|2.662|2.947|
|ConvertFP32FP16UMatMat::OCL_Core|2.706|2.601|2.603|2.528|
|ConvertFP32FP16UMatUMat::OCL_Core|2.704|3.215|1.999|1.988|

Patched version is not worse than convertFp16 and convertTo baseline (except MatUMat 32->16, baseline uses CPU code+dst buffer map).
There are still gaps against noOpenCL(CPU only) mode due to T-API implementation issues (unnecessary synchronization).


- [x] 12700K + AMD dGPU

|Name of Test|noOCL|convertFp16 dGPU|convertTo BASE dGPU|convertTo PATCH dGPU|
|---|:-:|:-:|:-:|:-:|
|ConvertFP16FP32MatMat::OCL_Core|3.130|3.133|3.172|3.087|
|ConvertFP16FP32MatUMat::OCL_Core|3.030|1.713|9.559|1.729|
|ConvertFP16FP32UMatMat::OCL_Core|3.010|6.515|6.309|4.452|
|ConvertFP16FP32UMatUMat::OCL_Core|3.016|0.242|23.597|0.170|
|ConvertFP32FP16MatMat::OCL_Core|2.697|2.641|2.713|2.689|
|ConvertFP32FP16MatUMat::OCL_Core|2.752|4.076|6.483|4.191|
|ConvertFP32FP16UMatMat::OCL_Core|2.706|9.042|16.481|1.834|
|ConvertFP32FP16UMatUMat::OCL_Core|2.704|0.229|15.730|0.176|

convertTo-baseline can't compile OpenCL kernel for FP16 properly - FIXED.
dGPU has much more power, so results are x16-17 better than single cpu core. 
Patched version is not worse than convertFp16 and convertTo baseline.
There are still gaps against noOpenCL(CPU only) mode due to T-API implementation issues (unnecessary synchronization) and required memory transfers.

Co-authored-by: Alexander Alekhin <alexander.a.alekhin@gmail.com>
 2024-01-26 12:56:52 +03:00
Alexander Alekhin
624d532000 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-12-17 21:05:34 +00:00
Alexander Alekhin
c240355cc6 dnn(ocl): avoid mess FP16/FP32 in convolution layer 2020-12-15 08:51:24 +00:00
Alexander Alekhin
f8791f072d core: avoid function type cast, make happy UBSAN 
backporting of commit: d3d13c41c4
 2019-06-11 19:36:47 +00:00
Alexander Alekhin
d3d13c41c4 core: avoid function type cast, make happy UBSAN 
oss-fuzz: https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=14115
 2019-06-11 07:06:29 +00:00
Alexander Alekhin
dfef04b325 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2019-02-12 17:54:40 +03:00
Alexander Alekhin
39b90ae9fb core: dispatch convert 2019-02-08 18:32:10 +03:00
Alexander Alekhin
5527c41468 core: clone convert.dispatch.cpp 2019-02-08 16:29:16 +03:00