Unsupported layers will transparently fall back to CPU execution. You can then choose to implement those who can be as TPC kernels. It's fundamentally much less

Unsupported layers will transparently fall back to CPU execution. You can then choose to implement those who can be as TPC kernels. It's fundamentally much less flexible than a GPU. Efficiency of Habana hardware isn't that great either, but that's another story... (they're still using TSMC 16nm in 2022 notably). Where Habana has an advantage in some workloads is cost.