Matrox without ATI?

[Topha]

Just wondering: If ATI hadn't released the 9700, Matrox would probably have been a lot more successful with the Parhelia, right? Considering that its not that slow compared to the GF4, whose sucessor isnt available yet. So maybe big M, just like nVidia, was just not prepared for ATI to becomes as big a competitor as they are?

Or maybe the 9700 made them panick and release the Parhelia before it was finished?

[The PIT]

I may have changed my 8500 for it was doubtful.

[TdB]

could be, parhelia looks like it could need a few extra revisions, to really shine.

just imagine how it would be if the 16xfaa was completely bugfree, (and working anisotropic filtering) so that they wouldn´t have a reason to include the normal 4xfsaa.
if that was the case, then the reviews would have been a lot less concerned be the lack of raw speed, the recent trend in GPU-reviews is that there is put alot more focus on high-quality-tests, with fsaa and anisotropic filtering.
when viewed like that, then the parhelia ain´t bad at all, it would clearly be superior to the ati 8500(it already is, IMO) and probably compare well to the lowend gf4ti´s.

of course if it wasn´t for the r300, then nvidia wouldn´t be ashamed of releasing a 400mhz nv30 4 months earlier(they probably could, atleast review samples), but matrox would have hit a window where the parhelia would be a good deal for gamers, who would want good 2d(assuming all the bugs would be fixed in a later revision).

[GT98]

Rumor has it that the Parhelia was ready to go in the Fall of 2001, but numerious problems delayed it to the Spring of 2002. And those problems have been well documented here.

[borat]

Originally posted by GT98
Rumor has it that the Parhelia was ready to go in the Fall of 2001, but numerious problems delayed it to the Spring of 2002. And those problems have been well documented here.

that figures, would have been a beast back then.

[UtwigMU]

Well according to 270 core overclock posted here, and explained in detail on www.3Dcenter.de, Vigilant's rumorus

It is fast. It will have a new memory design, different from all current video cards. Memory speed will be faster than the current fastest (a GeForce4). Faster by a huge margin of victory (~19GB/s bandwidth).
4. It is fast. It will be an all new chip. It will not be 256 bit and will not be 128 bit. Expect clock speeds around 300 Mhz. Also expect varied product offerings from this core after the initial products are released.

about 300MHz clock speed, huge heatsink

on alpha board, parhelia was targeted for 300-350 clock speed.

If they made it they would land in the 4600 - Radeon9700 range. Were the AF not broken and with low penalty and glitchless FSAA (I heard it's much better than on start now) this card would look much different.


Uppon reading original Millennium manual (I'm getting it tomorrow to power my 2nd head), Parhelia's philosophy is closer to Millennium than GXXX cards: Fast, Heidi drivers for Autocad and drivers for Bentley Microstation.

Well i really wish/hope/want to believe they come up with a card that for once matches what was planned (G450, G550, Parhelia).

Do you think Sundog/Parhelia 1.5 could have fixed glitches and reach Parhelia's target clock speed?

Radeon is doing 380 core on 150nm process.

[Kruzin]

Sundog will use the same chip (they still have a large stock that they need to get rid of), and it will be clocked even slower than the current Parhelia, and stripped of some of P's features (coughnoFAAcough).
Don't expect anything better than the current P till they exhaust the existing batch of chips, and start producing the 8x variant.

[bsdgeek]

How long do you think that will take?

[Kusa]

Is this gonna be a Sundog?

http://www.tenlinks.com/NEWS/PR/matrox/022803_ndes.htm

[dZeus]

Why do people keep thinking that the samples ran at 300 MHz hence the huge heatsink? If that was the case, then why do people have such troubles to reach high clockspeeds with their modded cooling and even core voltage mods?

Perhaps the huge heatsink on the alpha/beta boards was because of something else? Maybe the chip just got VERY hot because of a bug?

[UtwigMU]

Well they (o'cers) reached 270 with voltage mods. 300 as a target (I'm not sure they (Matrox) reached it but they might have planned it.) is possible.

That might be the 64MB version for entry level CAD. If they pull it off it might be good: cert. CAD drivers and a nice price for a professional card + a workstation deal with premier tiers could move a lot of cards. It wasn't the MAX's that sold the bulk, it was the G400 SH's and OEM G450's most likely.

As i heard Sundog is a different core, so they likely won't release that untill they get rid of stock.

Sundog = Parhelia 8x?
Pitou = Parhelia II

Haig, if you see this, (how) would you comment above linked article?

Edit: Clarification

[Marshmallowman]

The alpha was an engineering part, it may well have run quite well at 300mhz, but the current one sure can't.

Poor matrox, they copped so much shyt because they could not supply enough g400 maxs to fill demand, so they make sure their next big release has plenty of parts for imediate release...and plop no demand. If the swapped the release quantiyys g400<-> parhelia eg heaps of max's and short supply of P's they may not have had to lay off so many people.

I quite like the P in current form, but in oz it is still 40% more than BBA 9700pro, Feck me if I going to pay that much for it, they have to start dumping these chips pretty fast or MGA will end up as a sub depatment of their video group pretty soon

This horrible over compensations screams very bad management, I would have thought when releasing a completly new lline a bit caution would have been appropiate, especially since they seem to have cut short their beta tesing. Now they won't make the hard decsion to clear old stock.

I have horrible feeling I have already bought my last matrox card

[Novdid]

Originally posted by Marshmallowman
I have horrible feeling I have already bought my last matrox card

Sadly I have the same horrible feeling...

[chhfchhf]

the Geforce4 ti4600 under locks to 220MHz/550MHz,the Parhelia can defeat it!
the GeforceFX proof that the frequecy of core could not decide directly the capability of a video card.
i am pride for my parhelia

[Kurt]

It doesn't matter that the P is faster clock for clock than the 4600. Fact is the 4600 is faster. The reason why it is faster is not important