Duo Volari V8 preliminary results

[Ailuros]

http://www.tomshardware.com/graphic/20031107/index.html

Ze fliezzzzzzz are coming *ducks*

[1badjedi]

ok for the less informed, is this some of the same folks that brought us teh kyro?

[opus512]

The V3, V5, and multi chip solutions :D

[Ailuros]

ok for the less informed, is this some of the same folks that brought us teh kyro?

*kicks jedi in the nuts*

XGi = former SiS graphics department (Xabre) + Trident.

[1badjedi]

*kicks jedi in the nuts*

XGi = former SiS graphics department (Xabre) + Trident.




well i dont know who people are :P

i know hardware names not manufacturers :O

[Ailuros]

well i dont know who people are :P

i know hardware names not manufacturers :O

If you wouldn't be such a lazy bastige you would have read through the preview. :D

[Jack_Tripper]

Great...another card that does good in benchmarks and shitty in games...just what we needed

:evil:

Jack

[Ailuros]

Great...another card that does good in benchmarks and shitty in games...just what we needed

:evil:

Jack

*cough*bringsbackfondmemoriesofblurrytexturesorAA4 free*cough*

[1badjedi]

If you wouldn't be such a lazy bastige you would have read through the preview. :D



yeah well, thats unavoidable http://1badjedi.freewebspace.com/New%20Smileez/Other/sleepy03.gif........:D

[OtakingGX]

Did nVidia "optimize" their drivers for Halo?

[Jack_Tripper]

*cough*bringsbackfondmemoriesofblurrytexturesorAA4 free*cough*
:)

[Ailuros]

Did nVidia "optimize" their drivers for Halo?

Yes and those are real optimisations, killing most unnecessary FP32 render targets.

[Ailuros]

Not that anyone is really interested but Wavey just confirmed that is supports only Supersampling (2xSSAA per chip).

Might be an option for you SSAA freaks out there *shrugs* :D

[WedgieMan]

$300 for that? :ugh:

[Randell]

why are Toms still the only site that consistently shows the FX5600Ultra as faster than the 9600Pro, the 5700Ultr as faster han the 9600XT and the 5900 as faster than the 9800Pro?

Some of his numbers are totally out of sync with say [H]OCP's

[Blighter]

why are Toms still the only site that consistently shows the FX5600Ultra as faster than the 9600Pro, the 5700Ultr as faster han the 9600XT and the 5900 as faster than the 9800Pro?

Some of his numbers are totally out of sync with say [H]OCP's
Yep. I've noticed that as well. Something's fooked up over there.

[marianozz]

http://www20.tomshardware.com/graphic/20031107/images/xgi-comp1.jpg

ROFL!!!!!!!!!!!!11 :bigthumb:

i wonder how many investors they catch with this kinda accurate projections... :lol:

[3dfxlover]

i wonder how many investors they catch with this kinda accurate projections...

If they manage to fix their drivers, you realize you might eat your words...

Not that anyone is really interested but Wavey just confirmed that is supports only Supersampling (2xSSAA per chip).

Where did you see Wavey saying anything about 2xSSAA.
As far as I can tell, he ONLY mentionned about Volari using supersampling, which we already knew. Tridam is the one who talked about 2xSSAA per chip, not Wavey.
So either you're mistaken in your quoting, or Wavey is another one who suffers from split personnality...

[marianozz]

If they manage to fix their drivers, you realize you might eat your words...
lol... ill buy you a beer if that happens... :rolleyes:

Should they be successful, XGI's cards may well become serious alternatives to ATi's and NVIDIA's offerings, especially considering their comparatively low prices. XGI is aiming for a street price of about $300 (plus tax), which is rather inexpensive. This makes XGI's top model a good $100-200 less expensive than the flagship models of NVIDIA or ATi.
btw, these people should lay off the crack pipe.... "rather inexpensive"... if im already blowin 300 bucks on this POS, i might as well blow 100 more and get something better :dunno:

[Ailuros]

If they manage to fix their drivers, you realize you might eat your words...

"Fix" what? For heaven's sake how the f*ck can someone screw up BASIC functionalities like trilinear; it should be just one bit per register to go from bilinear to trilinear. LOD is out of line, no anisotropic functional as not a single sign of anti-aliasing. There would be an excuse for some let's say rendering errors here and there, slow shader performance in selected cases or whatever else, but there's not A SINGLE EXCUSE why an IHV would miss basic functionalities like trilinear. Ironically it hasn't been any different since 1998 and if they continue to make the same mistakes they can kiss their supposed market leadership in 2007 goodbye.

Maybe they should think of another company name change in the meantime, they might need it after such a pathetic showcasing. Even with bilinear the performance in custom timedemos is more than just slow; where does the excuse of beta drivers exactly fit it? Oh wait they haven't been able to fine tune performance for those yet or what?


Where did you see Wavey saying anything about 2xSSAA.
As far as I can tell, he ONLY mentionned about Volari using supersampling, which we already knew. Tridam is the one who talked about 2xSSAA per chip, not Wavey.
So either you're mistaken in your quoting, or Wavey is another one who suffers from split personnality...

I used a parenthesis on purpose (should you need an explanation of the term or what role it might play in proper syntax let me know).

Neither did I use a straight quote, I don't seem to be unable so far to handle simple UBB tags.

IHVs that deliver SSAA only in dx9.0 chips deserve to be shot IMO. There's not enough fillrate for shader functions and reducing fillrate by 1/2 for 2x samples and 1/4 for 4x samples is engineering suicide, not a viable sollution.

[opus512]

Tom's a kn0b.

[3dfxlover]

how the f*ck can someone screw up BASIC functionalities like trilinear. there's not A SINGLE EXCUSE why an IHV would miss basic functionalities like trilinear.

Tell that to nVidia first :D
I hear trilinear doesn't work in many D3D apps with a GeforceFX... Just check Beyond3D if you don't believe me, there are sufficient reports of well informed people over there claiming that.
Let's not be one sided, shall we? :)

[Ailuros]

Tell that to nVidia first :D
I hear trilinear doesn't work in many D3D apps with a GeforceFX... Just check Beyond3D if you don't believe me, there are sufficient reports of well informed people over there claiming that.
Let's not be one sided, shall we? :)

The only that's one sided here is you. If you would have bothered some months ago I was quoting and linking to articles that exposed the "brilinear" (or hybrid bi-/trilinear) modes appearing since 51.75 Detonators to be exact in extensively analyzing articles from 3DCenter itself.

However the GFFXs shipped with fully operational default trilinear and so did all their products for years now, one thing I have yet to see with anything coming from the original SiS origin/today XGi since 1998.

About 1.5 year ago, does it look any familiar? We had a quite similar discussion about it back then too and yet what did it turn out to be?

http://www.xbitlabs.com/articles/video/display/xabre400.html

October 2002. 3DCenter exposes that Xabre doesn't antialias with 4xAA in reality, but rather applies a 2*2 block blur filter:

http://www.3dcenter.de/artikel/2002/10-20_a.php

Since Volari was unable to run any application apparently with trilinear and LOD was way out of place, Weinand should have actually run all cards in just bilinear and with a positive LOD offset of +0.5 to put things on more or less the same level. If you're satisfied with the showcasing at Tomshardware then it's fine by me. I personally don't see anything to have changed so far, albeit I'd really like to see it for a change as unbelievable or obscure it may sound to you.

I'm all for extra competition or to be more specific for a real 3rd contender; sadly enough I don't see any of the kind happening not now and not in the foreseeable future.

I check B3D more than once a day, since I'm an active member over there.

[Ailuros]

As soon as the first results from several tests from the german PCGH pop up we'll have a few more detailed results about single and duo V8.

So far it's doing fine apparently, single V8 gets in Halo 1024*768*32 an average of 6 fps.

[sj_hurst]

As soon as the first results from several tests from the german PCGH pop up we'll have a few more detailed results about single and duo V8.

So far it's doing fine apparently, single V8 gets in Halo 1024*768*32 an average of 6 fps.

So what you're really saying is... V8, like its veggy juice namesake, leaves a bad taste in your mouth. :puke:

:D

[Ailuros]

So what you're really saying is... V8, like its veggy juice namesake, leaves a bad taste in your mouth.

That's a one sentence post. Too short :evil:

[Blighter]

As soon as the first results from several tests from the german PCGH pop up we'll have a few more detailed results about single and duo V8.

So far it's doing fine apparently, single V8 gets in Halo 1024*768*32 an average of 6 fps.
lol :lol:

that's pretty good. :bigthumb:

almost on par with the 5600 u :evil: