When choosing a pod offering, there are two priorities that stand above the rest; How fast is the GPU, and how much data can I shove into its memory (does my model fit?)

Unfortunately the cloud sub-command shows the DRAM value as "Mem GB" instead of VRAM, the latter which users naturally would expect the "Mem GB" field to represent for this kind of compute provider, leading to users having to either google search the VRAM amount for a given GPU, or use the web console.

It would be nice if the VRAM was provided in addition to DRAM, so the two main questions when choosing a pod configuration can be answered by the CLI.