• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Was Dave right?? Check out Anandtech's new Voodoo 5 5500 review incl. HSR

decent reviewage, tho they ignored quite a bit of stuff we recommended in the thread, most notably

1) depht precision
2) Geometry Assist INSTEAD of T&L enabled in MDK2 (christ almighty, that's a doubling of performance...ARGH!!)
3) No glide scores in UT
4) No overclocking

but at least they showed some of the perfomrance increase, despite missing out on the biggest/bestest feature (depth precision)
 
Man i know that dood hardware is sheeeeting his pants now. 🙂 Damn 3dfx is getting so close i only wish they would concentrate on this alittle more.

Yeah what happen to the GA ,OCing and glide scores?

It is proven now that the V5 5500 is powerful no doubt, I hope these drivers are updated first week of Jan, beta or WHQL build. But lets hear what dave has to say! 😉
 
I just looked at the Q3 benchmarks... nothing more.. I'm REALLY ready to bitch because the scores on the V5 are laughably low.... it REALLY pisses me off actually because I KNOW the V5 is faster than that. Are they putting the V5 in a system that already had a GTS in it without formating? That is EXACTLY what it looks like, because those are the numbers the V5 gets when I do that (ie. doing so lowers the score).


sigh... sad...
 
Hey guys. Glad to see that you had time to check out the review. Let me address a few points.

RoboTECH: I believe that you will find the answer to questions 2 and 3 by reading the Test Explained page. We did not alter the depth precision settings because we found that changing this setting adversly effects image quality. Overclocked scores were not included because the overclocked performance of the card was already addressed in our full
Voodoo4 4500 and Voodoo5 5500 reviews. If you check out our last driver review, NVIDIA's Detonator3 drivers, you will see no overclocked numbers were included here as well.

Adul: Please see the Test Explained page for an explination of why the Glide API was left out of this test.

DaveB3D: As with all of our tests, any hardware change is accompanied by a full reinstall of Windows. The only card used in the Voodoo5 5500 and Voodoo4 4500 tests were the Voodoo cards. All NVIDIA tests were performed on a fresh install of Windows.

I hope this answers your questions.
 
Let us look at a few things.

First, I ran benchmarks on a P3-733 on a V5 with the betas.

First, entirely default driver settings. Q3 at max detail (I dunno what they run it at.. can anyone tell me??)

10x7x32- 73.3
12x10x32 - 44.7
16x12x32 - 27.5

Now enabled refresh optimization

10x7x32 - 74.9
12x10x32 - 50.7
16x12x32 - 30.2

Now we set refresh optimization to faster (an no, unlike what some people say it does not go to a 16-bit Z-buffer when you do this)

10x7x32 - 78.4
12x10x32 - 50.7
16x12x32 - 33.5

We see a drastic change.

So at first glance it does appear that the GTS is still faster. But is it really? The law of averages comes into play here. The GTS has T&L, giving it a higher peak frame-rate in lower fill-rate situations. Now this mostly holds true for 10x7 and 12x10. Consider what happens when we move to a 1 GHz T-bird.

10x7x32 - 83.8
12x10x32 - 53.7
16x12x32 - 35.1

We see that the performance goes up in even upper resolutions. We could expect the same thing to remain true with a faster CPU to an extent (how much I don't know). See it all comes down to averages. You can record the instantaneous frame-rate with something like Intel's Graphics Performance Toolkit to see this as well. It is pretty neat to actually see the performance of every single frame that you display.
 
We did not alter the depth precision settings because we found that changing this setting adversly effects image quality.

I play with this on Faster all the time. I am a sucker for Eye Candy so if it hurt the picture any it wouldn't be one.

Would you say it adversly effects the image quality more or less with than the GTS with Texture Compression...

It is obvious that you are partial to NVidia and were a poor choice to conduct this review.
 
No joke.. I'll admit that in 16-bit you'll find a few image issues, however in 32-bit there is NO reason not to set Depth precision to faster.
 
heheh, yeah you have to turn off the GTS' texture compression because that really screws stuff up.
 


<< I dunno what they run it at.. can anyone tell me?? >>


Maybe you should read the review as it is right there explained in plain black and white 🙂


<< Now, when it comes to game configurations, we test each card with the same set of settings. In Quake III Arena, we test all cards with the &quot;Normal&quot; settings, only modifying the color depth, texture quality, and video mode (resolution) to the desired resolution. For example, at 1024x768x32, we set the video mode to 1024x768 and both the color depth and texture quality are set to 32-bit. >>

 
IBMer, you beat me to it. I agree about the texture compression issue. No one seemed to mind the Geforce image quality problems with texture compression. Seems to be a double standard.
 
I thought the HSR results looked very promising, and there has been a definite trend of performance improvements in the driver updates.

Of course none of that matters now because nVidia are going to purchase 3dfx. 🙁
 
Why does it seem like everyone here is a 3DFx or Nvidia zealot? Zealot being the key word. Some of you guys act like this is a football game or something. Go Nvidia, go! Geez.
 
so lemme get this straight...you guys disabled V-sync, which is known to degrade image quality...yet you don't want to fiddle with depth precision because it degrades image quality

😕

Paul
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top