问题
I'm trying to run code on my GPU, and firstly I clashed with the problem of having a working CUDA on ubuntu 18.04. After some pain I was suggested to work in a Docker Image, where I have (I presume) a well working CUDA version (I'm starting learning pytorch, and I was able to obtain a torch.cuda.is_available() = True).
After that I tried to run a simple code copied here, in order to see if effectively my GPU was able to do its job.
Unfortunately not... I receive the message:
NotImplementedError: bounds checking is not supported for CUDA
and I really wasn't able to understand a proper way to fix it.
This is what I obtain if I run nvidia-smi
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 440.33.01 Driver Version: 440.33.01 CUDA Version: 10.2 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GeForce 940M On | 00000000:0A:00.0 Off | N/A |
| N/A 36C P0 N/A / N/A | 242MiB / 2004MiB | 1% Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
+-----------------------------------------------------------------------------+
And this is what I obtain if I run nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Sun_Jul_28_19:07:16_PDT_2019
Cuda compilation tools, release 10.1, V10.1.243
回答1:
Seems like the community is dealing with some bugs
So I tried some hacks and it actually worked!!!
Just changed the line
@jit(target ="cuda")
to
@jit
来源：https://stackoverflow.com/questions/60012614/notimplementederror-bounds-checking-is-not-supported-for-cuda