Dragan Djuric
Dragan Djuric
I do it only occasionally, in cases where it is really important for that particular binding to stick out. I think that recommending that for all boolean variables would raise...
Here are the results for AMD's Pitcairn (R9 270X). I'll also upload the results for Hawaii (R9 290X), but I am getting an error during Xgemm. I'll open another issue...
Hawaii (AMD R9 290X): [hawaii.zip](https://github.com/CNugteren/CLBlast/files/244276/hawaii.zip)
And i7 4790k: [i7-4790k.zip](https://github.com/CNugteren/CLBlast/files/244290/i7-4790k.zip)
Sorry, I messed up that zip. As I do not have those files any more, I'll send them when I manage to do that tuning.
Tuning results for Nvidia GTX 1080 [nvidia_gtx_1080.zip](https://github.com/CNugteren/CLBlast/files/714943/nvidia_gtx_1080.zip)
Results for i7-4790k: [i7-4790k.zip](https://github.com/CNugteren/CLBlast/files/715149/i7-4790k.zip)
@Ulfgard I tuned those hawaii results on R9 290X. In my case, it would be impossible that performance drops 30%-40% for larger matrices, since I get (if memory serves me...
I was also talking about wall clock time in my (Clojure on the JVM) program, not ClTune results. 8192x8192 sgemm runs in 293 milliseconds on R9 290X (5.4 TFLOPS max)....
I will. Can you point me to the right function in the API (when you have time, I'm not in a hurry). I am using CLBlast via JOCLBlast, and I...