A dock-camera species tagger is a large vision network, and CPU hosts miss the ranger-app latency budget. A second, tiny tabular lot-full model on the same campus does not need a GPU. How should inference compute be chosen?

Short Explanation

The species tagger is a large vision net that misses the latency budget on CPU. The tiny lot-full model does not need a GPU. Pick GPU when the model and latency need it, CPU when they do not.

Full Explanation

Inference-host selection is GPU when the model and latency need it and CPU when they do not. That is not a Domain 2 training-instance pick and not Domain 4 post-go-live rightsizing.

×