Next Version Of Parhelia

[Byock]

Originally posted by ElDonAntonio


I think that a 256 version with fixes and upgrades would be Matrox's only chance. If they throw it away, they are total FOOLS. That 64MB card will get laughed at.

Most people were not that impressed with the P. Dont get me wrong, I like it, but the reviews and FPS people, slam it constantly. This will just add fuel to their fires.

[leech]

I like my Parhelia. Sure, it has a few problems, that just need a driver update. But all in all it's a great card, and once I can get a few more monitors it'll definitely be the envy of my friends. And for working on the desktop, you can't beat three screens.

Leech

[103er-Fan]

I also like my Parhelia, sometimes I am thinking that I might have been better with a Radeon 9700, but otherwise I am satisfied with my P.

What I find pretty stupid is that 64mb Parhelia idea, why did they put out FAA? I can only think it might be mainly for workers/graphic designers and so on, they need tripple head much more than FAA, for them the new Parhelia might be ideal, but for gamers it is not an option I guess, because you can get a GeForce4 Ti4400 for that price, and that will be definatly faster than that crippled Parhelia

[Pace]

The point is, would anyone buy a 64MB Parhelia? FAA is a major selling point in my view. If it didn't have this, I would not buy a Parhelia 128. $250 for this? The G550 must be close to $100 now, OEMs aren't going to use this.

Anyone who wants triple output might just get it, but then the stretched mode is limited. More likely they want DualHead or greater than 3 screens with MMS.

Possibly the daftest thing I've heard about Matrox so far. HeadCasting was laughed at a lot, but the G550 was still a great card - you can laugh at silly features all day, but you can't laugh at features a card doesn't have.

P.

[The_King]

This has advantages and disadvantages. The casual gamer is looking for a cheap parhelia because it offers the best faa (the only) around. Yes this is useful for a little while, but even at the limited cost of performance the faa delivers, soon there will be games that would only run on a 64mb parhelia with faa off. Yes I probably would have bought a 64MB version and the only reason I would choose matrox over ATI would be the FAA. However, I see where they are going by keeping triplehead. This was probably the plan all along: release the 128MB to draw the crowd then release a faa crippled 64mb version for OEM market. Still I think price is too high for that, even for Matrox who are always expensive.

[diewlei]

Originally posted by Pace
you can laugh at silly features all day, but you can't laugh at features a card doesn't have.

P.

no offence pace but why not.
we all laughed at the celeron with mendocino core that doesnt have L2 cache.

but seriously, i would get the 64mb if the price is right. if matrox were to actually target this variant as a "low-cost" solution a la the MXes, LEs or non-pros of other competitors, they should bank on features that users of this group wouldnt use ie. TH. if you choose a low cost card, you probably dont have $ for 3 monitors.

logical to assume i suppose.

as for FAA and stuff that can be enhanced through drivers, they SHOULD keep it on the cards.

[Supernaut]

I think they made a good decision to get rid of FAA. I am assuming that FAA has to be fixed in hardware or else they would have done a better job with it . If this is true, I dont want FAA since its broken.

Triple head works great so they kept it in KUDOS. But for just 150 more, you can get the P128 and if you plan on using 3 monitors im sure the extra money wont matter.

I would be happy with a dual dvi, 64 version of this card that ran at close to the same spead as the P128 without those distortions problems that the current P128s have right now. I would gladly pay 200 for it, although I think it would be a steal for 150.

WHOS WITH ME! I just need something 2 tide me over until the P256 comes out or the next version of the parhelia core comes out! FREE FROM DISTORTION AND MEDIOCRITY!

[thop]

what i'd like to see:

64MB
traditional DUAL HEAD/DVI, no HEAD CASTING
FSAA (preferably not borked)
P128 speed
everything thats is borked in P128 fixed
170$

64MB
SINGLE HEAD
no FSAA, no GIGACOLOR
P128 speed
everything thats is borked in P128 fixed
125$

i'd buy any of them.

[Dr Mordrid]

For the pro market I'd keep triplehead. It's a major advantage for video editing etc.

What really pissed me off was that you can't get an analog output while using triplehead. They should have put at least and S-Video connector on the plate so you could get analog out with all vga modes.

Component would have been even better, even if it took an accessory plate.

Dr. Modrdid

[K6-III]

You can get a VGA-Component converter box for appx $70....

[Reckless]

I hadn't upgraded my G400 as I was hoping for a little more speed for a *little* less cost. So, I thought instead of buying something else I'd wait and see what the 64MB unit price was but WHAT NO FAA, LESS speed What are the marketing people in Matrox doing

I've been buying Matrox cards for years - partly because of the Matrox brand and support but mainly 'cause they at least started off good and took a while to devalue themselves. The Parhelia has certainly seen me change my attitude and it looks like someone else may now have to fill the AGP slot

This 64MB rubbish just sounds like the G450 when compared to the 'original' G400

[Kurt]

P64=market research offspring

It probably makes sense for them to offer a TH, no FAA version.

This is NOT a gamer-oriented card. IMO if you're a gamer you shouldn't consider a P (unless TH is the killer feature for you).

If their graphics unit is in trouble, it makes perfect sense for them to release another card for the multimonitor/analysts/financial market.

It will eat up their MMS-series' sales but you have to consider the cost of development of the P and it ROI. Don't forget that TH is not 4 screens (QH ). The ones who really need QH are probably not many anw will spend the money on MMS cards. OTOH with a 250$US price-point they might expand their MM market.

It might be to the P what the GF4 4200 is to the nVidia line -i.e. it sells like hotcakes (all things relative).

As far as I'm concerned, they should scrap the 256MB version and shoot for a game only card with the P2 on an accelerated time-frame (I don't mean they should release it next month, maybe next fall). In the meantime they need to cost-reduce the actual line -which they're doing with the P64.

My target prices would be 279$US for the P and 179$US for the P64 by spring/summer next year which would open up the 349/399$US spot for a P-II (or P-256 if they insist ).

The main problem with introducing a P-II next year is that if it can only focus on the high-end gaming market (at 399$US), which is very small, the ROI won't be good enough, unless they scrap the P line except for the P64 (P64-LE?) which should sell for 129$US with the low-end P-II (128mb?) at 249$US and the high-end at 399$US (256mb?). That'd be a dream line-up if they can pull it. The question is will there be a P-II by that timeframe? Will they have the resources/balls to scrap the actual P line in favor of a newer, better one?

[bic]

VigilAnt must be wrong! His information must be wrong! No one in their right mind would make such a ridiculous decision. Look at www.pricewatch.com, the Matrox Parhelia is the Highest price Video card right now ($322), even more expensive than the RADEON 9700 Pro ($311) which just came out. If the crippled P64 comes out at $250, why would anyone want to buy it when you can get a GeForce4 TI 4600 ($212) with double the ram and double the speed? And by the time the P64 is out the price will probably drop even further. It doesn't seems right.. VigilAnt can you confirm your data again... and is there an eta on the P64?

If it's true.. then RIP matrox.

BTW.. recommendation to prove/disprove sales for P64.. someone setup up survey as who is willing to buy the crippled p64 for $250US and send this result to Matrox marketing. They need better data from actual targeted consumers.

[Nuno]

Crippled 64 Mb P: Big Mistake

Issues to correct on a new Parhelia rev:

- Get out 2 texture units on each pixel pipeline. 4x2 is more than enough. More than 2 texture units by pipeline was already a proven failure with original Ati Radeon. Much better than that is the ability to render at least 6 textures on one *geometry* pass.

- Redesign memory controler to 4 64 bit independent buses with bandwidth saving features.

- Some sort of z-buffer conpression/occlusion culling

- More efficient texture caches and definitly the ability of doing a trilinear filtered pixel on each texture unit (the need for two passes trilinear is killing parhelia right now)

- Get some 2.0 pixel shaders with all the transistor space that´s left after removing all those useless texture units

- Improved chip design/layers/yelds to do at least 300 Mhz core

- Independent resolutions on 3 monitors, migh as well be a driver hack in order for the OS to see 3 video cards, who cares.

- $250-300 price range (quality doesn´t come cheap)

If properly optimized, with a 300 Mhz core and a 256-bit memory bus, it could do near Radeon 9700 performance.

[Novdid]

If you want 9700 performance you have to have 8 pipes, 4 ain't gonna cut it.