Any cancelations for the parhelia?

[Fenrir(AVA)]

I think it is one thing for a hardware review site to talk bad about the card, but when John Carmack gets on it.....*sighs* On paper this card looks tremendous, but what happened? Matrox if you read any of these posts, do a revision quick and beef up the speed before no one buys the card I have until the 14th to cancel my card, which I might do, unless something drastic changes with the Parhelia. I hate to say it but, thats how it has to be, for me.

[Afronaut]

Martha Stewart Cancelled hers today.

[zandor]

I'm thinking about it. This is a real drag. I'm really quite disappointed. Basically, I want a new card now to get me out of 640x480 hell on GTA3 (looks like crap on a 21"), but I'm a bit hosed since I can't find anything acceptable out there.
ATI doesn't write Linux drivers. (Matrox has at least promised 2D drivers soon, and delivered on the G series) Then I can't find a review of various GF4 boards telling me which ones have decent 2D IQ. My understanding is that some are ok, and some board manufacturers skimp on components and end up with blurry pictures.
The other part of my little problem is that I don't want to be stuck in 640 hell again in 6 months when Doom III comes out. Honestly I probably won't like it all that much, but I'll probably pick up a copy just to see all the eye candy.
I guess I'll just have to either gamble or suck it up & wait.

[isochar]

I'm in the almost the same boat. I was really hoping that the Parhelia would be "be-all and end-all" of high-end gaming cards.

[The_King]

I remember after it was found out the G800 was no longer (at least the G800 we were expecting) almost everyone was saying they just need all the features with a decent 3D speed. Thats obviously not the case now. It seems that people want a video card for bragging rights....

I think the biggest mistake when releasing this card was the overhyping. Matrox did attempt to explain that it wasn't a Gforce4 killer, but all the hype on these forums is what did it, raising the expectations of everyone on here. If the card was not known about then suddenly released, almost everyone on here would be purchasing one.

Well theres my rant.

[crow8k3a]

I disagree.

i believe no one intentionally mislead, matrox or other matrox users.

it was the sheer hardware specs of the card that looked so impressive on paper, it just naturally led those to believe it to be "faster" than the current chipsets.

[tmservo]

Part of the problem seems to be that this card is not the card originally promo'd by matrox. There is no doubt about that. The core is the same, true, but it is originally in their PR as 375Mhz Clock.. whereas it is now running at 275Mhz clock. 100Mhz difference, that's huge significance. It appears as though the initial impressions - which touted a greater then 20Gb bandwidth were primarily based on the earlier clock projections.

At it appears as though Matrox had to "crank it back" in order to gain stability - which tells me that the Parhelia has a lot of room to grow with a die shrink, and may be a prime candidate for overclocking if the fan were to be replaced, etc.

But it looks as though Matrox started with one idea and had to step back, back, back to maintain stability and keep the quality of components it offered. I think this not only because of their own PR, but also because there is no other justifiable reason to use 300Mhz memory if you're not even going to use it.

Now that several reviewers are saying the 256MB version will be "clocked higher" it makes you wonder if this isn't a chip that should have been always run at .13 instead of it's current build; and that Parhelia 256MB (or a Pro or a Max) couldn't be in the waiting with a die shrink and an increase to the core speed.

Because, as it stands, it just walks, talks, and looks like it was initially intended to run at the initial proposed speed and not 275Mhz.. I'm wondering what overclockers will make of it

[Fenrir(AVA)]

I could agree with you, since I also noticed a lot of reviewrs saying the 256mb might be clocked higher. I really hope this card can be improved tremendously. Right now only thing the card has going for it is the surround gaming, which is the only reason I have decided to get the Parhelia now.

I really do hope the 256mb is beefed up, if matrox has read what people have said about the card. I hope they are commitied to the GPU and continue its evolution, and produce a top of the line card, which I hoped this one would be. Looks great on paper but something went wrong....I agree completly with tmservo however.

[rugger]

Originally posted by The_King
I remember after it was found out the G800 was no longer (at least the G800 we were expecting) almost everyone was saying they just need all the features with a decent 3D speed. Thats obviously not the case now. It seems that people want a video card for bragging rights....

I think the biggest mistake when releasing this card was the overhyping. Matrox did attempt to explain that it wasn't a Gforce4 killer, but all the hype on these forums is what did it, raising the expectations of everyone on here. If the card was not known about then suddenly released, almost everyone on here would be purchasing one.

Well theres my rant.

Well, I wasn't expecting a Geforce 4 killer either.

But I also wasn't expecting something that can barely handle new games either. 28 fps in UT2002 at 1600x1200 is barely fast enough to play, and soundly beaten by most of the competition.

The 256bit bus was supposed to power the Parhellia through these tough benchmarks, but it doesn't seem to have made any difference what so ever.

I was hoping that the Parhellia would be within 10-20% of the Geforce 4 4600 in current games, and be able to beat it into a bloody pulp in new games.

[Maniac]

Originally posted by tmservo
Part of the problem seems to be that this card is not the card originally promo'd by matrox. There is no doubt about that. The core is the same, true, but it is originally in their PR as 375Mhz Clock.. whereas it is now running at 275Mhz clock. 100Mhz difference, that's huge significance. It appears as though the initial impressions - which touted a greater then 20Gb bandwidth were primarily based on the earlier clock projections.

At it appears as though Matrox had to "crank it back" in order to gain stability - which tells me that the Parhelia has a lot of room to grow with a die shrink, and may be a prime candidate for overclocking if the fan were to be replaced, etc.

But it looks as though Matrox started with one idea and had to step back, back, back to maintain stability and keep the quality of components it offered. I think this not only because of their own PR, but also because there is no other justifiable reason to use 300Mhz memory if you're not even going to use it.

Now that several reviewers are saying the 256MB version will be "clocked higher" it makes you wonder if this isn't a chip that should have been always run at .13 instead of it's current build; and that Parhelia 256MB (or a Pro or a Max) couldn't be in the waiting with a die shrink and an increase to the core speed.

Because, as it stands, it just walks, talks, and looks like it was initially intended to run at the initial proposed speed and not 275Mhz.. I'm wondering what overclockers will make of it

uhhh sorry I have to be the one to break the news, but the clock speed is not even 275, but instead a lowly 220mhz. This I beleive is the biggest factor in why its getting trashed in alot of the benchmarks and reviews. Nvidia cards were running faster than 220mhz well over a year ago. If M. can get it up past 300mhz, then this chip could actually START to perform like a winner...

[tmservo]

My apologies, I meant memory speed was being used at. Haig and others had intoned it had originally been tested/idead at 300/375, not 220/275. As I said, seems to be a card crying for a die shrink and better cooling, as the clock definitely came down as the fan shrunk from the beta boards (which Haig has said had to happen to stay within AGP spec)

[CHHAS]

Haig posted on the Matrox Tech Sup forum the he had seen no indications of the 256 MB version would be clocked higher.

Link

If it has higher clocks, then yes it will be getting higher frame rates. But I am told that the clocks will be the same.

[tmservo]

I'm sure Haig is very knowledgeable, but when the Marketing VP says that the 256MB version will have higher clocks:

http://www.extremetech.com/article2/0,3973,181999,00.asp

I have a reason to think that a corporate VP will have maybe a better inside track.

Anand also notes this is what he has been told as well:

http://www.anandtech.com/video/showdoc.html?i=1645&p=2

[Cheesekeeper]

Matrox did attempt to explain that it wasn't a Gforce4 killer

I wasn't expecting a Geforce4 killer. The problem is that at current speeds, the Parhelia isn't even fast enough to be a Geforce4 wannabe. The whole point of the Parhelia was that it could maintain its speed even when quality was cranked up to full. That may be case, but when the speed it's maintaining is already lower than the speed of a Geforce4 *with* quality up on full, then that's not much of an advantage. Plus, based on some of the screenshots, the Parhelia's aniosotropic filtering quality isn't as good as the Geforce4 anyway.

The Parhelia would need to be turning in rates at least 75% of the speed of a Geforce4, before its "low performance hit" becomes an advantage. Unless they can start turning these out running at least at 300MHz, the Parhelia is destined to live a very short life............

[Marshmallowman]

Given the size of the alpha boards heatsink, it may have been they could not get acceptable cooling as a standard AGP card.

With AGP pro that would have been easier..

With a bit of luck, if you don't have a card in the next pci slot, and you whack a decent size socket A cooler on it(dare I say peltier)
you may be able to get 300 out of the core