Utilizing the two devices in a Tesla K80 on AWS p2 instances

Question

I'm using a p2 instance on AWS which has a Tesla K80 GPU with 2 GK210 GPUs, So I should be able to see and should access each of the 2 devices separately. But I can see only one. So how can I access both devices?

Answer 1

Actually, with p2.xlarge instances you get 1/2 of a K80 assigned to VM, so your 'Nvidia-smi' output is expected here, and you won't be able to access 2 GPU devices from the VM. 

Do checkout AWS Certification to learn more about AWS.