Cuda driver version is insufficient for runtime version
![cuda driver version is insufficient for runtime version cuda driver version is insufficient for runtime version](https://user-images.githubusercontent.com/65749168/111619795-ef4c0600-8820-11eb-8c82-175548b1f0c3.png)
I am unable to trace the line of reasoning being followed by this error message (specifically that CUDA driver version is insufficient for CUDA runtime version). This prints False and raises: Check failed: e = cudaSuccess || e = cudaErrorCudartUnloading = false: CUDA: CUDA driver version is insufficient for CUDA runtime version The relationship that 450.80.02 >= 410.48 is true, therefore, the system is in order.Īfter compiling TVM from source with CUDA support enabled, I attempted to run the following debug code: import tvm
![cuda driver version is insufficient for runtime version cuda driver version is insufficient for runtime version](https://user-images.githubusercontent.com/50136603/56967057-12626380-6b7e-11e9-89da-27bd31099408.png)
The toolkit-to-driver-relationship chart in NVIDIA’s docs indicates that runtime version 10.0.130 requires driver version >=410.48.
![cuda driver version is insufficient for runtime version cuda driver version is insufficient for runtime version](https://user-images.githubusercontent.com/35773560/49597379-4dcb7e00-f9a2-11e8-8804-277c8b086fb8.png)
Nvcc reports runtime (toolkit) version 10.0.130: $ nvcc -versionĬopyright (c) 2005-2018 NVIDIA CorporationĬuda compilation tools, release 10.0, V10.0.130 Nvidia-smi reports that my current environment (an NVIDIA T4 on AWS) has driver version 450.80.02 installed: $ nvidia-smi