the big parhelia review conclusions thread

[thop]

Originally posted by Sinistral
- Yeah but does the Parhelia offer great 2D quality? Triple head and True Dual Head? 10-Bit Gigacolour? Etc... -

a freudian slip if i ever saw one

[Sinistral]

- I edited it! . No one is perfect... -

[Mumrik]

Originally posted by Dr Mordrid
You're ignoring the fact that the low frame rates are mainly due to the absense of an outdated culling methodology that is used in these older products.

Since Matrox has issued a DirectX8 SDK for using Displacement Mapping and Distance Adaptive Tesselation new games can start using it right away.

Also: DirrectX9 will be using DM and DAT so Parheia's best performance lies in the future. Unless NVIDIA and ATI use this Matrox-licensed technology they will be the ones stuck in the past

Dr. Mordrid

Explain to me why oclussion culling is outdated. It seems Matrox "brute force" method of providing more memory bandwidth isn't doing the job...

Btw, if nVidia and ATi don't use the Matrox-licensed tech Matrox will be the one stuck. Rather few game developers are willing to do special programming that only one card supports... especially if that card isn't a performance leader...

[xortam]

Originally posted by Dr Mordrid
You're ignoring the fact that the low frame rates are mainly due to the absense of an outdated culling methodology that is used in these older products.

Since Matrox has issued a DirectX8 SDK for using Displacement Mapping and Distance Adaptive Tesselation new games can start using it right away.

Also: DirrectX9 will be using DM and DAT so Parheia's best performance lies in the future. Unless NVIDIA and ATI use this Matrox-licensed technology they will be the ones stuck in the past

Dr. Mordrid

I feel where Matrox failed is not getting these developers on board and ready to demonstrate these features at product launch. There just isn't enough SW currently available to exploit the unique features of P. Who is the market for these cards as they stand now? The drivers don't appear to be mature enough for anyone especially given the price point of this product. I hope eventually the drivers and applications will show the true power of P but the competitors may be able to surpass those capabilities by then.

[MrGaribaldi]

mallowman> I think the problem with UT2k3 is that the beta boards were running at a higher clockspeed than the retail boards does...

Granted this is pure speculation, but hear me through...

• Haig has commented that they couldn't set the clock to 300, as they'd need the heatsink from the beta board, which violates the AGP specs...
• It takes time to sample the chips, so changing the core in 2-3 weeks are unlikely has happened
• Haig has also commented the P should be very easy to oc

mikeul> Well, the V5 is getting quite old now... (I should know I still use it in my main system..)
and the fact that the P has so many features, and a company that still develops drivers kinda makes telling someone to buy a V5 kinda hard...

Xortam> When did you last see a product launch with this many features that were supported by SW? The only thing that springs to mind is FSAA with the Voodoo, and that's about it...
Not to mention the list that Matrox has put together with upcoming games to support some (if not all) of these features is quite good...

aldamon> Well, what about those of us who actually has 3 monitors? Or how expensive is it to pick up a couple of used monitors from a IT company going bust?


.02$

[xortam]

MrGaribaldi ... Matrox needs some compelling reasons for users to buy this card today and I don't think they've quite delivered on that. They could have addressed this by making sure there was SW available to show off the P's strengths. Lacking that, they could have assured that they released the cards with solid drivers to run current SW. The price is too high for most users, especially professional users, to purchase this HW currently on the promise of what's to come.

[Pace]

The question I wanna know is how it ran UT2003 *so* well, yet struggles here.

Performance in UT and Q3 is very important to me, it will be great to finally run these at top resolutions (I hope), but UT2 and Doom3 is just as important - if Parhelia runs Doom3 as well as the G400 runs UT I'll be happy...but I have my doubts.

P.

[Mumrik]

Pace, did you read the Carmack quote on Doom3 performance?

[The_King]

Speaking of quotes...from Anand's conclusion....

"...they can't compete with NVIDIA's driver team on the very first major hardware release they've had in the past two years, if they were able to then NVIDIA would have been off hiring members of their driver team long ago....."

[GT98]

Originally posted by The_King
Speaking of quotes...from Anand's conclusion....

"...they can't compete with NVIDIA's driver team on the very first major hardware release they've had in the past two years, if they were able to then NVIDIA would have been off hiring members of their driver team long ago....."

More like stealing them...GRRR

[Marshmallowman]

The synthetic tests showing high fill rate & better hi res FPS are a good sign.

But the low fps at low res can indicate two things.

1. The drivers are not efficient (yet to be optimised)
2. the geometry engine is limiting the FPS

I think both are true, in current games I would expect at least 20% improvemnet to come with driver optimisation.
It could even be up to 50%(or more) in some extreme casese.

The limitations in geometry throughput are a know fact (matrox says it won't be a FPS king), but I think there is enough raw grunt there for good gaming perfomance.
A good culling algorythym(and DM) will certainly improve performance a lot, since the card is geometry limited not fill rate limited.
We may see some extra perfomance boosts in current games, when developers patch their games to use the extra pixel/vertex and texture hardware.

New games which know and use the new hardware features should really pump along.
Displacement mapping is a BIG thing, it will be a real boon for a lot of games when they use it, flight sims will just hum along.

Yes it is expensive, but it does have a totaly unique feature set, things that a lot of people will purchase the card for, professionals not gamers.
no other card has 10 bit colour, triple head,higest rez available for dual digital flat panel and the best signal quality for the best output quality.
The glyph hardware is another pro feature that people overlook, CAD, desktop publishing and graphic artists will se nice perfomance and quality improvements.

Do a search for the P10 card, I just read a review and it was near half the speed of a quadro, why?. No applications support for the strengths of the hardware.

P10 and parhelia are revolutionary cards and as you have read, the drivers have been written from the ground up.
So don't expect optimal perfomance until we hit version 2.x or 3.x of the drivers.

and I can't aford one either....yet!

[EvilTypeGuy]

What I can't believe is the review here:

http://www.neoseeker.com/Articles/Hardware/Reviews/parhelia/6.html

I want that Parhelia!

[Pneumatic]

From that review:

I would like to warn the hardcore gamer that this is a high-end graphics card catered to the needs of Design Artists, Animators and stock analysts as a gaming card it falls short in the raw performance category. If you are a hardcore gamer who likes raw performance, I would recommend that you look at the GeForce 4 Ti4600 Series.

[Afronaut]

It's 2002 why would anyone be happy Gaming @1024?

Thats so 90's

[Fenrir(AVA)]

Only reason I went above 1280*1024 in some games is to get rid of the jaggys, make them less noticeable. Some I can not play above 1280*1024 since the text is so small and I can not read it. The P10 from the begining though did not promote like the parhelia as being a great gaming card, they said it was mainly for workstations.