qt and opengl lib issue?

Accelerated Computing CUDA CUDA Setup and Installation
1

Using 5.5qt. I encountered this error when plotting 3d opengl using nividia GPU K20

QGLShaderProgram::attributeLocation( vertex ): shader program is not linked
Vertex shader for simpleShaderProg (MainVertexShader & PositionOnlyVertexShader) failed to compile

do I need install some API extension or driver etc? seem like I’m missing something

2

when remote running openGL app on different machine, does my local machine need to have GPU?

3

that’s an interesting question, because some sort of remote execution will utilize local libraries for drawing and running the code and other means will utilize remote machine resources, in my opinion.

4

check for reference https://devtalk.nvidia.com/default/topic/1021756/jetson-tx2/remote-desktop-control-if-jetson-tx2/

5

Your local machine needs to be able to run whatever is the local client associated with your remoting solution. This typically does not require e.g. 3D OpenGL support, but does require some sort of graphics capability to render the GUI. Normally, the ability to run a VNC-like client is sufficient:

[url]ubuntu - How to run CUDA/OpenGL interop (particle) sample from a remote machine - Stack Overflow

6

As I understand, two cases can be distinguished:

  1. code [for example cuda] is located at remote side and is executed at remote hardware with remote libraries ie cuda

  2. the code residing at remote computer is executed at local hardware with local cuda version etc.

7

so I can run non-openGL GUI remotely but not the openGL verision remotely(run fine locally). which make me think my local GPU need to be nividia GPU.

8

In my opinion nomachine based x2go will provide you with more options of running a remote code or getting it delivered.