I have a Windows calculation-intensive solution that uses NVidia's GPGPU (CUDA).
I am looking to host my application on the cloud and expose it via a Web service. The only host I found was AWS, but it does it on Linux CentOS.
I want to find an online host that supports:
1 - NVidia GPUs, specifically Tesla (Tesla M2090 or Tesla M2075).
2 - Using Windows Server 2008 R2 or 2012.
3 - Virtual or Dedicated.
Are you aware of any host that fits my 3 points above?