Curiouser and curiouser ...
I've had a look at the Sparkle GeForce FX 5800's specs:
Quote: "
SPARKLE GEFORCE FX 5800 128 MB AGP GRAPHICS CARD FOR PC
Format: PC
AGP GeForce graphics cards for PC
• nVidia NV30 (GeForce FX) chipset
• AGP 8x bus connection
• DirectX 9.1 hardware accelerator
• 128 MB DDR2 Memory (>1 GHz Data Rate)
• 400 MHz core clock speed
• 800 MHz memory clock speed
• 8 rendering pipelines
• 16 texture units per pipeline
• 12.8 GB/sec memory bandwidth
• TV out
"
And then I looked at the POWER COLOR ATI RADEON 9800PRO's specs:
Quote: "
ATI 9800DV PRO chipset
• AGP 8x bus connection
• DirectX 9.1 hardware accelerator
• 128 MB DDR
• 380 MHz memory clock speed
• 340 MHz core clock speed
• 19.8 GB/sec memory bandwidth
Up to 8x FSAA (Full Screen Anti Aliasing)
• Up to 16x Antistropic Filtering
"
The GeForce is £130, and the Radeon is £150. The GeForce has more than double the memory clock speed of the Radeon, and has a better core clock speed. They both have 128mb of ram, but the Geforce's is DDR2 (a lot faster). It seems that you get a lot more bang for the buck with the GeForce. The Radeon's only redeeming features are its memory bandwidth, and its FSAA and antistropic filtering capabilities. I don't know if the GeForce has these capabilities, and if truth be told, I don't actually know what they are.
Is the memory bandwidth all that important, or could it be sacrificed a little for money, core speed, and memory speed?
[edit]
It turns out that the GeForce 5800 has something called "intellisample", which has antialising and antistropic filtering. If the memory bandwidth isn't too important, then I think I may have found my card!
[/edit]
