Jump to content

CUDA / cudNN on GIGABYTE GTX 1050 TI GV-N105TD5-4GD


arathorn2nd
 Share

5 posts in this topic

Recommended Posts

Hello there!

 

I'm considering updating my trusty nVidia GT 640 for a newer pascal card. I've found one that seems to be recommended for its good support with nvidia web drivers:

GIGABYTE GEFORCE GTX 1050 TI 4GB GDDR5 128BIT, GV-N105TD5-4GD

 

The question is, does anyone knows if CUDA and cudNN works with the web drivers? I'm currently working on neural networks using TensorFlow and that would be a requirement for me.

 

Anyone using CUDA with a pascal card?

 

Thanks a lot!

Link to comment
Share on other sites

Hi !

i'm using 1050. cuda driver is loaded ver. 8.0.81.

kextstat | grep -i cuda
  121    0 0xffffff7f82865000 0x2000     0x2000     com.nvidia.CUDA (1.1.0) DD792765-CA28-395A-8593-D6837F05C4FF <4 1>

i think it works,

but i'm getting this error when try to compile samples from cuda toolkit.

on sierra for 8.0 cuda. xcode 8.3

nvcc fatal   : The version ('80100') of the host compiler ('Apple clang') is not supported

Link to comment
Share on other sites

  • 2 months later...

I would check for cuda sdk (mac) updates first. Cuda will work but dev in CUDA (using SDK) is other thing - Xcode version must fit !!

https://developer.nvidia.com/cuda-downloads

Look, after selecting Mac, OS X version in the realease notes about what Xcode is supported. Often the newest and never beta version are not supported.

June 2017 : CUDA Compilers. The CUDA compiler now supports Xcode 8.2.1. = No higher version support yet!!!

For cuda dev you must downgrade Xcode.

Link to comment
Share on other sites

 Share

×
×
  • Create New...