TVM on CPU with ARM Neon support and QNX OS

Hello,

I wanted to know if there is any existing branch for QNX on ARM CPU support for TVM inference? Use case is I just want to run a simple neural network for a short demo but the device doesn’t have Linux support. I checked in the forums and saw some support for Zephyr RTOS but didn’t find anything on QNX. If there isn’t existing support available, what would be needed to add such a support? Will the TVM runtime need updates or just the CodeGenerator for LLVM IR?

TVM supports ARM CPU and neon instructions. You may find the following thread helpful: Do we need to explicitly enable NEON for ARMv8 devices? - #10 by ramana-arm

ok, thanks, will check. Does it also support QNX?