[GPU/OpenCL] Check fp16(half) support #2608
Conversation
Added check for fp16 support on OpenCL. Will show proper message if not found. Signed-off-by: Debadri Samaddar <[email protected]>
s-debadri
requested review from
myungjoo,
jijoongmoon,
again4you,
jaeyun-jung,
leemgs,
wooksong,
helloahn,
kparichay,
gichan-jang,
anyj0527,
zhoonit,
lhs8928,
songgot,
jihochu,
DonghakPark,
SeoHyungjun,
baek2sm,
skykongkong8,
djeong20,
EunjuYang and
a team
as code owners
|
📝 TAOS-CI Version: 1.5.20200925. Thank you for submitting PR #2608. Please a submit 1commit/1PR (one commit per one PR) policy to get comments quickly from reviewers. Your PR must pass all verificiation processes of cibot before starting a review process from reviewers. If you are new member to join this project, please read manuals in documentation folder and wiki page. In order to monitor a progress status of your PR in more detail, visit http://ci.nnstreamer.ai/. |
|
|
cibot: 
There was a problem hiding this comment.
@s-debadri, 💯 All CI checkers are successfully verified. Thanks.
| @@ -155,6 +155,31 @@ bool ContextManager::CreateDefaultGPUDevice() { | |||
| device_id_ = devices[0]; | |||
| platform_id_ = platform_id_; | |||
|
|
|||
| #ifdef ENABLE_FP16 | |||
| // check for fp16 (half) support available on device | |||
There was a problem hiding this comment.
which half-precision does it support? _Float16, __fp16, or half
There was a problem hiding this comment.
Ideally all of _Float16, __fp16 and half can be used interchangeably inside kernels when cl_khr_fp16 extension is enabled. I have tested with __fp16 and half myself, they worked exactly the same. However I did not test with __Float16 but it is supported by OpenCL 2.1 or later.
There was a problem hiding this comment.
half is specifically designed for use in OpenCL kernels. Some are architecture specific (e.g: __fp16 for ARM). So half will be more portable and benefit from further optimizations..
There was a problem hiding this comment.
thanks for the clarification! if using the half type has advantages, should it be added as a tensor data type?
nntrainer/api/ccapi/include/tensor_dim.h
Lines 25 to 31 in 1a6537f
There was a problem hiding this comment.
half will only work inside OpenCL kernels. So I think adding it to tensor data type won't be useful in this scenario.
Added check for OpenCL fp16
halfsupport on device.Will show proper message if not found.
To use fp16 enable
cl_khr_fp16as following with the kernels.Signed-off-by: Debadri Samaddar [email protected]