Question for those who use (or have used) the Quadro K620 2GB
I’m putting together a low-cost workstation and the amount reserved for the video card can not exceed the value of k620 or a GTX750Ti.
Never had experience with quadro nvidia cards, the K620 has less attributes compared to GTX750 but is designed specifically for programs that will use (Sketchup, AutoCAD and others for architecture)
I’m afraid to buy the k620 and regret if it is too weak, have no money to buy a higher quadro, in this case I buy a GTX750ti? I want to be able to navigate without locks on my projects.
My question is: GTX750Ti or 2GB Nvidia K620
For those who have used the k620 would be helpful opinion.
I may not qualify to talk about this, but I am going to attempt anyways.
I currently own GTX750Ti at home and Quadro 4000 at work.
They both perform quite well given that the model isn’t managed poorly.
(unnecessarily large geometry count, too many edges shown or large materials)
I am in architecture industry, and my model contains masterplanning (landscaping and paving) information to balustrades detail at times. Average file size is 50MB to 200MB, but that doesn’t mean much on how your computer will perform handing the model.
SU isn’t designed to make advantage from the extended OpenGL capabilities provided by CAD cards resp. drivers as of the nVidia Quadro or AMD FirePro series. The recent SU v2015 requires OpenGL v2.0 only (or higher of course).
Sketch3d_de is right. The K620 is really meant for things like traditional CAD 2D drafting. Get a “consumer” card if you don’t absolutely need a high-end professional card (K4200, K5200 etc…) for some application that supports them.
the CAD series are typically targeted to OpenGL-based mid-range/high-end 3D modelers as e.g. Catia, Parasolid/NX, ProE/Creo, SolidWorks, Inventor etc. often in connection with dedicated drivers certified by the maker of the CAD.
2D CADs reguarly do not profit by an OpenGL acceleration and therefore mostly run fine with common consumer cards or even integrated solutions.
SU does of course profit of the tessellation performance of the GPU but the performance of the CPU obviously need to be on par for getting a balanced system.