Aye, I understand what you are saying, Catlin, but it was the Motherboard/Processor/Ram upgrade that actually made the difference. Same with you Pauljs... The way to test it is to, obviously, change only the graphics card. On one of my systems, an AMD 1.5, we tested it with three cards, on my personal benchmark file. The file consists of only four different objects with four different textures, running the spectrum of transparency, reflectivity, volumetrics, and blah blah... The three cards were 1) Graphics Blaster TNT(1), PCI, 16MB. 2)Chaintech Geforce 2, AGP, 32MB. 3)Asus Geforce 4 MX 440, AGP, 64MB DDR. There was 1 second difference between the TNT and the Geforce 4. On a 30-minute render time. These cards have absolutely nothing to do with Bryce, except when you are in Open GL or Direct X preview modes. Even the Nvidia Quadro card will do absolutely nothing for your render times. Bryce was made to run on any machine with enough RAM, be it a P-2 200, or a AMD 2400+, or whatever! Which totally sucks. If future versions of Bryce cannot incorporate at the LEAST, and Unreal 1-quality preview mode, or even better rendering support for hardware acceleration, I fear I'll just keep Bryce 5 and not give Corel any more of my hard earned money...