HL2 performance boost for nV owners?

[Keysplayr]

Originally posted by: nitromullet

Originally posted by: DAPUNISHER
Congrats on your pending Nuptials :beer:

Thanks.

Normally I'd say to run like the rabid dogs of hell are chomping at your heels, but, congrats.

 

[LocutusX]

Someone at the Beyond3D forums has "benched" forcing the game to FP16 on a FX5900:

http://www.beyond3d.com/forum/...rder=asc&start=160

1024x768 (all high except "reflect world", vsync off, trilinear filtering)

Radeon 9800 Pro - 73,7 fps
FX5900 @ DX8.1 - 62,44 fps
FX5900 @ DX9 forced FP16 - 31,45 fps
FX5900 @ DX9 full p. aka FP32 - 19,46 fps


1280x960 (same as above)

Radeon 9800 Pro - 52 fps
FX5900 @ DX8.1 - 45,18 fps
FX5900 @ DX9 forced FP16 - 23,31 fps
FX5900 @ DX9 full p. aka FP32 - 14,4 fps

So much for this "tweak" making DX9/FX5900 at the same level as 9700/9800...

 

[nRollo]

Originally posted by: LocutusX
Someone at the Beyond3D forums has "benched" forcing the game to FP16 on a FX5900:

http://www.beyond3d.com/forum/...rder=asc&start=160

1024x768 (all high except "reflect world", vsync off, trilinear filtering)

Radeon 9800 Pro - 73,7 fps
FX5900 @ DX8.1 - 62,44 fps
FX5900 @ DX9 forced FP16 - 31,45 fps
FX5900 @ DX9 full p. aka FP32 - 19,46 fps


1280x960 (same as above)

Radeon 9800 Pro - 52 fps
FX5900 @ DX8.1 - 45,18 fps
FX5900 @ DX9 forced FP16 - 23,31 fps
FX5900 @ DX9 full p. aka FP32 - 14,4 fps

So much for this "tweak" making DX9/FX5900 at the same level as 9700/9800...

Can't disagree with that!
An impressive performance boost (50%) nonetheless, although apparently at the cost of IQ in this case.
Looks like nV3X owners are better off sticking with DX8.1 and R300s retain their Dx9 lead.

 

[Dethfrumbelo]

Too bad we can't see how the nV4x's would stack against the R4xx's under identical precision, either ATI at FP32 or nVidia at FP24.

Apparently, converting to FP24 causes some performance loss for ATI, but still probably less of a hit than doing full FP32.

20 fps -> 30 fps at 1024, respectable I guess, but the FX5xxx's just suck at DX9, no way around that.

 

[Drayvn]

Originally posted by: Dethfrumbelo
Too bad we can't see how the nV4x's would stack against the R4xx's under identical precision, either ATI at FP32 or nVidia at FP24.

Apparently, converting to FP24 causes some performance loss for ATI, but still probably less of a hit than doing full FP32.

20 fps -> 30 fps at 1024, respectable I guess, but the FX5xxx's just suck at DX9, no way around that.

Huh? ATi do it at FP24 anyway...

 

[James3shin]

im guessing he meant ati goes the fp32 path