Ripped from VE. Here is what one of the guys at Shiny wrote on T&L in Sacrifice:
While there's certainly no problem playing the game without T&L, I find it makes a very noticeable difference.
There's a lot of polygons in the source data for most of the stuff in the game. Cards that have really good hardware T&L, like the Geforce cards, lets you see a lot more of that. I've seen poly counts nearly double on Geforce2 cards vs the same machine w/o T&L.
Its not just poly counts though. In some cases it can dramatically improve the framerate.
In summary: if you've already got a 3d accelerator that meets our min spec don't worry about it. But if you're planning on buying new card anyway you should consider a board with good hardware T&L.
Now I find this VERY strange.
I've run it side by side.. identical systems and identical monitors. One with a 5500 and one with a GF Ultra and the two of us looking at it could tell ZERO difference between the two. We stared at it for about 5 minutes, moved around, etc. Couldn't find a lick of difference between the two.
The guy from Shiny I talked to at E3 say he ran it at work on a P3-500 and a TNT2 (or was it TNT1?). He said that it was VERY close to GTS. He said the only place he noticed a difference was WAY out in the distance on the mountains. Also he said that a 700 MHz system and similar or quicker graphics card would look identical to the T&L version.
While there's certainly no problem playing the game without T&L, I find it makes a very noticeable difference.
There's a lot of polygons in the source data for most of the stuff in the game. Cards that have really good hardware T&L, like the Geforce cards, lets you see a lot more of that. I've seen poly counts nearly double on Geforce2 cards vs the same machine w/o T&L.
Its not just poly counts though. In some cases it can dramatically improve the framerate.
In summary: if you've already got a 3d accelerator that meets our min spec don't worry about it. But if you're planning on buying new card anyway you should consider a board with good hardware T&L.
Now I find this VERY strange.
I've run it side by side.. identical systems and identical monitors. One with a 5500 and one with a GF Ultra and the two of us looking at it could tell ZERO difference between the two. We stared at it for about 5 minutes, moved around, etc. Couldn't find a lick of difference between the two.
The guy from Shiny I talked to at E3 say he ran it at work on a P3-500 and a TNT2 (or was it TNT1?). He said that it was VERY close to GTS. He said the only place he noticed a difference was WAY out in the distance on the mountains. Also he said that a 700 MHz system and similar or quicker graphics card would look identical to the T&L version.