pretty sure tht it doesn't. and as you might know it is quite abit slower than your 4600....in my opinion it is not at all worth the extra cost or trouble for it. althought the surrond gaming ting looks pretty cool.
Without a doubt, the visual quality of the Parhelia is light-years ahead of both Nvidia and ATi; however at the moment the card doesn't manage to cut it speed-wise. It is an extremely powerful card, but it's specific also - and nothing is currently programmed to target its particular strengths. In a couple of years, the games that support those features will still be playable on a Parhelia, whereas the 4600 won't stand a chance - but then how many of us will still be using the same video card in 2 yrs' time?
And no, Matrox has stopped putting any VIVO functionality on their condumer-level cards.
a13antichrist, I'd really want to believe that, but sadly that's marketing BS. It's not a case of the Parhelia being ahead of its times, it's just that the chip is horribly slow for the _present_ time. It just doesn't have have the chip clock frequency to compete with the big boys, and Matrox drivers always were slower than a snail too.
For example, given that memory bandwidth and the number of texturing units, it should zoom past any other card when anisotropic filtering is enabled. It doesn't. It just can't keep up. It just doesn't have the speed to compete. And you don't need any specially programmed games to take advantage of anisotropic filtering.
So in two years time, it will be even more POS than today. That's all. It will do you about as much good in two years time, as a TNT2 M64 bought two years ago does you now.
Plus, let's talk features in advance. Anyone remember the GF256? Right. It was supposed to be a card to last you for ever, remove the need for a CPU upgrade, and generally solve everything but cancer and world terrorism. You just wait until games start supporting those features. Only it actually took almost two years before those features actually became mainstream, by which time, the card was far from being top of the line any more.
Or even better, let's talk S3 Savage 2000. You just wait until people start supporting those features, and S3 gets their drivers to support T&L right. You'll see how it kills a GF256. Only it never did. It always remained a mediocre performing card, whose drivers eventually progressed from catastrophic to merely buggy. FFS, the card couldn't even support S3's own texture compression right, if we're talking unused features.
So your saying Matrox has potential, but they just ain't got their **** togehter, and trying to make a high performace card based on yesterdays technology (such as 0.15u die set) and slower clock speeds?
There is a Matrox benchmark, written of course to favour the Parhelia's strengths. From memory, which could be off, the Parhelia triples the scores of the Ti4600. Nvidia-optimized apps like that tree bench thing don't show anywhere near that sort of difference. The Parhelia does have power behind it, but Matrox are kidding themselves if they think developers are going to devote years of their time to include features for a card no-one is going to buy because of lacking raw performance numbers.
With 16x FSAA and ANISO all enabled, the Parhelia does outperform a similarly-handicapped Ti4600 in a lot of things. And for the people buying the Parhelia, it's with those settings that they're going to be playing.
Is a lot of it marketing BS? Sure, no doubt it. But there is some truth to it at the same time.
I will be buying a Parhelia. But I'm sure hoping they've revised the core a bit by mid-next year when I start putting together my Dual Hammers....
Actually, Matrox's 16x FSAA is also a part of what's making me _not_ buy one. It's edge AA, _not_ FSAA for a start. But more of a worry I'd put on another limitation of it: "Currently FAA will not work with any use of stencil in a game, which is one thing Epic had to disable in the UT2003 demo in order for FAA to properly work. Although the demand isn't necessarily great for UT2003 to support stencil right now, eventual support is necessary and if FAA doesn't properly work with it enabled then Parhelia will be forced into 4X supersampling mode which is no better than what the Radeon offers."
That's Anand's words. I would have said it will be _worse_ than what a Radeon already offers.
So let me get this straight, you need developpers to _disable_ features for Matrox's AA to even work at all? That's not encouraging.
And from all the screenshots I've seen, Matrox's anisotropic looks just marginally better than having anisotropic filtering disabled.
Frankly, for image quality with FSAA and anisotropic filtering on, the ones that look convincing are the ATI cards, not Matrox.
Plus, if you want to check out what happens with the performance when FSAA and anisotropic are on, you may want to have a look at this page: http://www.anandtech.com/video/showdoc.html?i=1645&p=16
Really doesn't look to me like the Parhelia is zooming past the other cards even in that setup.
Generally, every test I can find here and now, just doesn't support Matrox's claims. For example, fair enough, let's imagine that the Parhelia really needs quad texturing to shine. (Ironically, that exact same claim I've heard before for the S3 Savage 2000.) Well, I understand that UT 2003 does that, yet the Parhelia still ends up way behind. What's the problem, then?