Post your NV40 preview predictions

I think the upper right hand corner looks better on the XT (has a nicer color), but I also think that the power lines look better and more defined on the 6800. Very close IMHO, and in game the differences would be negligible.

 

ATI's AA does look superior to me about 99% of the time; however, these next gen cards are probably going to be running @ 1600x1200 oftentimes, and as resolution increases, the advantage of ATI's AA over NVIDIA's diminishes.

 

For me there is a significant difference between the two but limited to that area of the images only. The gamma compensation is making a big difference, my opinion. Would be nice to see the % hit for the different AA and AF modes. Good thing is that many more reviews (including HardOCP in full) will be available tomorrow .

 

I also noticed that the powerlines looked a lot less defined. Because of the gamma correcting ?

 

It looks to me like they do a similar job strictly in terms of removing the jagged appearance of the power lines, just that NVIDIA's screenshot seems to show fuller lines than ATI's. That is, everywhere the lines appear to be a little jaggier than they should be on the NVIDIA screenshot, they appear to be a little fainter than they should be on the ATI screenshot. I think this has more to do with just a different philosophy on how to weigh and blend the samples, rather than one technique being better than another.

The major image quality difference is in the brightness of the scene, but again it's hard to say if that's due entirely to gamma correct FSAA on the Radeon, or if there's also a difference in the base gamma settings for the cards themselves.

 

Sure those charts make things harder. Nv has made some improvements:
- New AA up to 8* (almost on par with Ati but still a little behind i would say)
- 16* AF
- Bri/trilinear is now an option
Wait for others previews now (i'm not impressed right now, like almost everybody, perhaps it will change with the other reviews)

OT: Congrats Humus!

 

Now that you mention it, it does look like Nvidia method enlarged the power lines due to AA. Now in motion I believe that the ATI method would blend better then Nvidia's probably due to the gamma compensation. No where else do I see any significant difference I might add.

 

radar, that is totally inappropriate, as well as being totally wrong.
The fact remains that the game is capped at 60fps...
The extra bolding you did merely explains why.

 

Count me in on that. A fellow Luleå citizen working for Ati, how about that.

Grattis Humus !!

 

Oh no it isn't wrong.

As I said in my first post, experiment in doom/doomsday (jdoom).

The command you need is settics, it defaults to 35. Try setting it different. See what results you get.

 

wot about the bit of corrugated iron roof at top right below the power lines??

 

Why oh why, in the year 2004, are we still forced to put up with JPG images? For gods sake, post some PNGs. I don't care how long they take to load, it's better than swift loading meaningless lossy nonsense. As the image quality between the two gets closer, the lossy images become even more useless than they already were.

 

Saw this at nvnews: http://www.hardwareanalysis.com/content/article/1708.1/ Old, fake or the real thing.

edit: Hopefully a fake because that must be almost the worst review I've seen ever.

 

Amen to that, at last some reason here. I will conclude Joe's post into a single point:
1) Use the "maximum playable settings" comparison as additional usefull info of a normal review, maybe two pages at the end. Otherwise it's completely inappropriate for comparing relative performances. To everyone out there who's dissapointed by those results, wait until we see some real reviews. Why ? it's so simple...you do that kind of reviews when the video cards you are comparing are KNOWN QUANTITIES i.e. there have been numerous "regular reviews" out there which have painted the piture pretty well and so you can then look at [H]-type reviews and have usefull info on maximum playable settings. However, benching a new video-card only that way (unknown quantity) is pretty much retarded, no-offence intended. You only get people seriously confused, not to mention some fail to see the different settings cause the look at the graphs hasty, and of course how the hell do you define "maximum playable setting" which is completely subjective ??? i tried playing FarCry at average 35 fps and i almost vomitted...
Noone is able to tell what the relative performance hit for each video-card is when "upping the res and AA one-notch" as someone else said.

 

I thought that the bold lines here explains why you still can measure framerates even though the game tic is capped to 60 Hz.

 

Strange "preview". No FSAA or AF , and, on the UT scores, the 6800 seems to have better FPS at 1280x1024 than it has at 1024x768. Also, um, how many here have P4 EE's?

Weird.......

 

http://www.firingsquad.com/hardware/nvidia_geforce_6800_ultra/

 

oops dp

but let me paste their conclusion

 

man you are fast I was there 5 mins ago (and now) and nothing still on the front page

someone copy this one

Far Cry comment

We created our own custom demo for testing with FarCry, Crytek’s latest shooter that takes advantage of 2.0 pixel and vertex shaders, and with the latest patch, 3.0 shaders as well. We also made sure that all GeForce FX cards were running the same 2.0/1.1 shader mixture as the GeForce 6800 and RADEON cards. All image settings were set at their highest. Since gunfire isn’t accurately reproducible in FarCry’s benchmarking mode, we simply ran through a level and recorded along the way.

Under this environment, the 6800 Ultra is CPU-bound at 800x600 and 1024x768. In fact, it doesn’t really begin to pull away from the other cards until 1600x1200, where its superior memory bandwidth really allows it to shine.


@ 1600*1200

NV40 60.2 fps
9800XT 42.6

and with 4xAA

NV40 43.8
9800XT 26.2

Now that is 66% difference plus for NV40

 

Aquamark3 on 9800XT : 19195 ??? That's about half of the actual score it gets normally.
And there they go with bloody BOTMATCH numbers, just ridiculous.