High Resolution graphs and Imaging best route on a Quadro 2000D and a Tesla C2075

Hello Everyone! (apologies in advance for waffling like a kid in a candy store trying to get the best tasting sweet)

I have been tasked with improving a product that is pretty dog slow, all its graphics rendering is done in GDI+ within Winforms.

So straight away my OpenGLES 2.0 background was itching to come out and play :).

However imagine my surprise when I take note of the hardware and its intended use (medical high resolution imagine and big data processing).

I realize the current hardware is being woefully underused and instead of going with my default setting of OpenGL I wanted to dive into the NVidia developer world and learn how to best make use of this hardware.

So a couple of questions:
*Where should I start? the standard Nvidia SDK?
*The Quadro information says it can handle 10M images, how? DirectX? OpenGL? Specific APIs?
*The current solution is C#, am I going to need to write some C++ wrappers or is there already existing C# support? (I was going to run with SharpGL originally)
*I believe the TESLA card is being utilized for its CUDA capabilities but im not sure how well, whats the best way of reviewing its utilization?

Apologies for all the questions, I am rather excited by capabilities of this hardware :) :) (I am used to Mobile GPUs)

I appreciate any feedback!!

The site http://www.nvidia.com/object/product-quadro-2000d-us.html says “support for 10/12bit grayscale monitors up to 10Mpixels” means you’d need a monitor to support that. Many greyscale monitors have 5MPixels (2560*2048).

10/12 bit display is normally done via OpenGL by selecting the proper pixel formats.
The OpenGL example code to select and use a 30-bit color format or 10- or 12-bit grayscale monitors can be found on this site in the “Whitepaper, Sample Code, Demos” section:
http://www.nvidia.com/object/quadro-product-literature.html

The driver installation comes with an nvidia-smi tool which allows to query information about your installed GPUs including usage at runtime. If you directly want to profile and debug CUDA, try Nsight: https://developer.nvidia.com/nvidia-nsight-visual-studio-edition