Your answer is simple, that file doesn't exists and
curl just save a HTML document as a
You can check it by opening this link in your browser. As you see, Page Not Found!
Also you can check it by command line:
curl -Is https://developer.nvidia.com/compute/machine-learning/cudnn/secure/v7.0.5/prod/9.0_20171129/cudnn-9.0-linux-x64-v7.tgz | head -n 1
HTTP/1.1 404 Not Found
As you can see, it has the same output.
So when you run this command:
curl -O https://developer.nvidia.com/compute/machine-learning/cudnn/secure/v7.0.5/prod/9.0_20171129/cudnn-9.0-linux-x64-v7.tgz
Curl just download this web page and save it as
.tgz file. So you can change
.tgz suffix to
.html and open that file with favorite web browser like firefox, because its not in gzip format it is just a HTML document.
Also when we check
cudnn-9.0-linux-x64-v7.tgz file format, as you see it is
HTML document and it's not
gzip compressed data:
cudnn-9.0-linux-x64-v7.tgz: HTML document, ASCII text, with very long lines, with CRLF, LF line terminators
cudnn you can visit this page, and after logging in, you can find many versions of
cuddn that you can download them.
And this is exactly the same file with which one, you want to download it before (But now many new versions and easy install files like
.deb are available and you can download them instead).
Because you said, you can't access web browser, I got this download link and write this command:
curl -O http://developer.download.nvidia.com/compute/redist/cudnn/v7.0.5/cudnn-9.0-linux-x64-v7.tgz
So you can use it by command line without web browser or logging in.