Thanks for downloading and using CodeAnalyst.
Are you trying to run CodeAnalyst on "Intel" platform?
The scenario mentioned above is a known issue with 3.6.1197 version.
If you are using Intel platform, i would suggest you to wait for couple of days. We will be releasing CodeAnalyst 3.7 version that has this fix.
Yes, I am running on an Intel platform. Sorry about that, we did buy an AMD video card though.
Can you point me a previous version 3.6.xxx that does work?
And when is 3.7 going to be available?
I saw others with the same issue in another thread, would it be possible to provide a .reg file to modify the registry to correct the problem with 3.6.1197?
The new version i.e. 3.7.1202 is out yesterday and it has fix for the issue you are currently facing.
You can download it from the following web link:
Please note that on Intel platforms, only "Time Based Profiling" feature works.
Let us know if you face any other issue.
Great, downloading 3.7.
I am not clear on the features yet.
Does "Time Based Profiling" give me timing info for inside the kernel, including utilization per core?
Other tools I have tried do not have any timing info for the code running on the GPU, they only time between OpenCL calls. This includes NVIDIA Visual Profiler, AMD APP Profiler, and AMD gDEBugger.
In "session settings" dialog, check "CPU Utilization" option to get utilization per code.
If you run your kernel on CPU, CodeAnalyst does provide timing information for the OpenCL kernel.