I have a machine that won't run any CUDA applications, but work fine with OpenCL ones, but the project keeps trying to send the old cuda31 work units (1.03 application).

How to tell the server to avoid them and send only the newest OpenCL 1.05 application work units ?

My other machines which are quite similar (all Windows 10 and NV GPUs) are getting these OpenCL WUs fine.

Under Project preferences you could uncheck the NVIDA box and test that. If that affects your other systems you can always recheck it.

Second question do you have the Run Test Applications checked ? if yes that will get you newer versions of some of the applications (version 105)

Bill F

---

In October of 1969 I took an oath to support and defend the Constitution of the United States against all enemies, foreign and domestic;
There was no expiration date.

It finally decided to send OpenCL work to this machine ... but I don't know why ...

Under Project preferences you could uncheck the NVIDA box and test that.

But then NO work will be send??

I also have the problem that i get these very old and inefficient CUDA3-WUs und would like to get only OpenCL.