Which two solutions should the architect recommend to meet these requirements?

An organization’s data scientists are executing a plan to use machine learning (ML). They must have access to graphical processing unit (GPU) capabilities to execute their computational models when needed. The solutions architect needs to design a solution to ensure that GPUs can be shared by multiple virtual machines.

Which two solutions should the architect recommend to meet these requirements? (Choose two.)
A . NVIDIA vGPU
B . AMD MxGPU
C . vSphere DirectPath I/O
D . vSGA
E . vSphere Bitfusion

Answer: A,E

Explanation:

https://blogs.vmware.com/apps/2018/07/using-gpus-with-virtual-machines-on-vsphere-part-1-overview.html