[VTA tutorial] Matrix multiplication

While running on simulation, I get all the results successfully (MM, Convolution, getting_started, …) But whenever I run with my PYNQ device, only the Matrix Multiplication example gets a mismatch on the result. Where could the fault possibly be? please anyone let me know…

I tried to re-install tvm on both host and device sides, re-build everything, re-downloaded the default python file from tutorial but nothing solves this problem. Any advice?

++ Even the gemm benchmark works. Doesn’t anyone have any problem running the current VTA-Matrix Multiply Blocking tutorial with PYNQ?

Someone who is running this tutorial with their FPGA boards, do you get successful results?