Hello, I have a GTX980 GPU that has been working well with GPUGRID and crunching away, but each time that Milkyway@Home downloads a work unit to work on that is Cuda based, it aborts immediately. Is there any special requirements other than the normal OpenCL stuff that is required to make it work properly?
Thanks.

Further information on this problem: It would appear that the program (Milkyway@home) can't recognize my GTX980 graphics card for some odd reason. Though it downloads and attempts to run, it fails with the error that the GPU does not exist. Bear in mind that this is the only program that has this error running on my GTX980. I have GPUGRID running perfectly fine right now and it's been running for about an hour without fail.
I would recommend reviewing the GPU identification code in the programs as it would seem some of the optimizations may have broken something to do with identification of features/chipset.
This applies to all GPU related programs from Milkyway@Home. I could be blasting out these WUs right now. If possible, I could take a look at the OpenCL code if you're willing to allow.