There are a couple outfits making M.2 AI accelerators. Recently I noticed this one: DeepX DX-M1M 25 TOPS (INT8) M.2 module from Radxa[1]: https://radxa.com/pro

There are a couple outfits making M.2 AI accelerators. Recently I noticed this one: DeepX DX-M1M 25 TOPS (INT8) M.2 module from Radxa[1]: https://radxa.com/products/aicore/dx-m1m https://radxa.com/products/aicore/dx-m1m If you're in the business of selling unbundled edge accelerators, you're strongly incentivized to modularize your NPU software stack for arbitrary hosts, which increases the likelihood that it actually works, and for more than one particular kernel. If I had an embedded AI use case, this is something I'd look at hard.