I have been trying to use the gpu_selector() for the queue after selecting the proper node in the devcloud that has iris_xe_max but it keeps on giving the error that no such device exist (i.e. the CL_RUNTIME error saying that no gpu exist). Can you suggest how to compile the dpc++ code using the gpu_selector for the queue?
Hi Nikhil ,
Thank you for posting in Intel Forums.
Could you please share the devcloud node name you are using ?
We suggest you to use generic method to login to available Iris_xe_max nodes that will guarantee your workload execution.
qsub -I -l nodes=1:ppn=2:iris_xe_max -d .
Also, please run this command and see whether GPU is available on the node you are using :
And let us know the updates.