|
|
 |
|
Workstation Smackdown (Nvidia Quadro DCC review) Added on: Mon Feb 04 2002 |
| Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
Is Software Heidi dependent on the video card?
Software Heidi is CPU and system dependent. It is a software emulation of the viewports, for those systems which don�t have the hardware to accelerate the displays. Video cards do not effect software Heidi performance.
Does the video card effect render time? Why didn�t you do rendering tests with the video cards?
The Video card has nothing to do with rendering performance. That is completely cpu, memory, and system based in performance.
However Maxextreme offers a hardware accelerator (Maxextreme renderer) for Max3, which allows an active shade type of display, which allows you to render a non production still, to help make choices in your design process. This was not tested as Max4 was used.
Why didn�t you use Max4�s ShowFPS=1 counter? Why did you calculate it mathematically?
As mentioned in the benchmark article, Max�s FPS counter is extremely erratic, as is the results found while using Maxscript. The only true way to get an accurate result, was to take four samplings manually, and average their results, for a highly accurate and highly reproducible result.
Why is Direct3D so much slower then OpenGL and Maxextreme? I thought Nvidia was fast in D3D.
The Direct3D in Max is not Nvidia�s doing, it is the design of its implementation within Max4. Since Max4 is the first 3dsmax release containing D3D it's expected to be a bit buggy and slow. Nvidia�s D3D drivers are top notch, and EXTREMELY fast. D3D performance in Max4 is entirely Discreet�s fault.
What�s the advantage of software Heidi if its so slow?
You don�t have to buy an OpenGL card to use Max. That�s the advantage. Heidi is inaccurate and extremely slow, though surprisingly stable. I would recommend purchasing an OpenGL accelerator as fast as possible to provide optimal performance.
Why are you running at such a high resolution?
Most 3D professionals run at a minimum of 1280x1024, with the middle range falling around 1600x1200. Many users won�t even work at a machine unless the resolution is 1900+. Sites that routinely test �workstation cards�, tend to do tests at 1024x768 and lower, which are actually unusable resolutions, as many programs (XSI for example) will not even function properly at resolutions below 1280x1024.
Now on to the benchmarks.
|
 |
|
|
