Installing Warp on AWS

[Cryo-EM in the cloud]

I am very interest in Dr. Dimitry's #Warp# software. Limited by the lack of a suitable windows host, I tried to compile it from source code on HuaWei Cloud environment other than AWS. The hardware configuration of mine are followed:
1. Intel(R) Xeon(R) CPU E5-2690 v3 @ 2.60Ghz
2. 16.0 GB RAM
3. 1x Nvidia Tesla GRID M60
The operate system is Windows Server 2012 R2 system and the version of the system's own Nvidia Driver is 369.71, I have tried to upgrade the driver but not success, cause of the cloud environment' virtuality. Other prerequisites are similar to what Dimitry's recommend. However, Compilation was always unsuccessful, everything else goes smoothly until came across tensorflow:), I tried to skip the compilation of TFUtility by directly using the libtensorflow.dll provided by Dimitry so I could temporarily ignore the tensorflow. Finally a binary file was generated, Unfortunately, when I tried to open it, a error window popping up saying could not load GPUAcceleration.dll. I'm sure the GPUAcceleration.dll has been built and located at where the binary is, attached is a screenshot of that popping information, so I would like to ask whether you have some ideas with this problem? It would also be very grateful if you could give me some advice about how to build #Warp# on Could environment. Thanks very much in advance!

Hello - 

For me in US-East-2, I used the following base AMI as a starting point: 

"Windows_Server-2016-English-Deep-Learning-2018.05.09" - ami-a0102dc5  

After booting this up, I follow installation instructions for: 

.NET Framework 4.7 or later

Visual C++ 2017 Redistributable

Latest GPU driver from Nvidia (Tesla M60)

And then download Warp & install. 

See if that helps your installation problems.

Mike