http://www.anandtech.com/video/showdoc.html?i=1620

[Slougi]

Originally posted by VJ
Slougi: Haig mentioned the 0.13 micron in one of his the threads on the Matrox forum, claiming that it is not ready for production yet. I'm not sure if I read it here somewhere or on Matrox-forums, but someone hinted at Matrox producing this card in 0.13 by late fall...

Jörg

Yeah I heard about that as well..... Can't remember where either, although I think it was one of the G1000 threads at the matrox forums.

[rugger]

Re: Re: All I can say is

Originally posted by Slougi


ALL GF 4 cards have dualhead

Thanks for correcting me. I really don't read Geforce reviews anymore, simply skim them to see where the performance bar has been shifted to.

Most Nvidia reviews are VERY boring, even the Geforce 3 generated boring review.

[Slougi]

Originally posted by Jazzz
VJ, Someone said that on a single monitor the Parhelia can play Quake3 @ 300-400 fps (we are not sure yet) and can play it @ 100fps on three monitors, this is not confirmed it is only a rumour.

This would put performance approximately on par with the ti 4600.

Thanks for correcting me. I really don't read Geforce reviews anymore, simply skim them to see where the performance bar has been shifted to.

Most Nvidia reviews are VERY boring, even the Geforce 3 generated boring review

No problem.

[Jazzz]

Originally posted by Slougi

This would put performance approximately on par with the ti 4600.

yeah anand was saying that seeing that the Ti4600 drivers are quite Quake3 optimized the parhelia won't be able to beat it because Quake3 doesn't use much of the Parhelia's features, however switch on FAA there won't be much of a drop of performance with the parhelia but there will be with Ti4600

[Marshmallowman]

on par with a ti4600, which is a evoltionary card in the geforce series...hence the drivers are quite optimised.
Whereas parheila is totaly new silicon design, no real optimisation..but it is on a par with 4600 *Now* that sounds good to me.

in 6months it will be faster still...and remember 10bit colour, its going to look a #$%# load better straight up..

Add a few new game that use dx9 features shaders etc....and it just gets better..

[Kosh Naranek]

Yes .. but what about the lack of Z-buffer compression ???
And the fully-programmable FPU pixel pipeline issue ?


Why didn't Matrox go all the way, why did they have to "nuder" these two important features ?

[rugger]

Because They can be added to the next version of the card for even better performance.

Plus those things are too hard to implement on a Parhellia card when it is on the affordable 0.15 micron chips.

[Slougi]

*sigh*
Now for the wait till E3 Doom 3 runnning on Parhelia anyone?

[Electric Amish]

I hope when the got to .13 they add in some Occlusion Culling. That's the only feature I'm kinda dissappointed in not seeing and it's all due to not having enough room on the die.

amish

[Parhelia Pro256]

quote:
--------------------------------------------------------------------------------
Originally posted by Jazzz then Slougi

VJ, Someone said that on a single monitor the Parhelia can play Quake3 @ 300-400 fps (we are not sure yet) and can play it @ 100fps on three monitors, this is not confirmed it is only a rumour.


This would put performance approximately on par with the ti 4600.
--------------------------------------------------------------------------------
Really? I didn't know nVidiA had a triplehead card

[Slougi]

Originally posted by Parhelia Pro256
quote:
--------------------------------------------------------------------------------
Originally posted by Jazzz then Slougi

VJ, Someone said that on a single monitor the Parhelia can play Quake3 @ 300-400 fps (we are not sure yet) and can play it @ 100fps on three monitors, this is not confirmed it is only a rumour.


This would put performance approximately on par with the ti 4600.
--------------------------------------------------------------------------------
Really? I didn't know nVidiA had a triplehead card

They don't but a ti 4600 gets around 300 fps on a single screen, divide by three and you get 100 fps. Also it is doubtful how many people will use this, as only a minority have the money for three monitors. Maybe if it is possible to use two small monitors on the sides and one big in the middle, but there is no indication of this thus far.

[Jazzz]

also another thing is that Anand has seen the nv30 in action, i think it has been posted before but ill post it again

I must say that right now the video card sector is by far the most interesting of them all in this industry. Matrox has a pretty big announcement tomorrow that I'll be covering, 3DLabs just recently announced their powerful VPU, NV30 is amazing, and Matthew and I will be up to see ATI later this week.

nv30 is amazing!!!! Seeing how he probably has seen the nv30 in action i tend to believe he was kinda comparing the two chips towards the end of his article, and he said it had potential to be the best card from when it is released till the nv30 comes out!! Quite a statement there. I think the parhelia won't be far behind it but i seriously doubt it would beat it, nvidia will look at the specs and counter it with something. the fact that ps2 memory management are not on the chip and the lack of Z-occlusion technology doesn't help either.

[Electric Amish]

NV30 might also be able to get in on the .13 die size as well, which will give it an advantage.

amish

[Jazzz]

Originally posted by Electric Amish
NV30 might also be able to get in on the .13 die size as well, which will give it an advantage.

amish

I believe it has been confirmed by the CEO of nvidia that the nv30 will be made on .13, so im sure ATI would probably also be making it on .13 as well

[Greebe]

Yes it's been know for a while now that the NV30 will be on 130nm process and using copper interconnects, but let me remind you all the reason nV canned the NV25 GPU was because of their fear of what Parhelia will bring... and it's huge... for them to sidestep the NV25 and jump on the NV30 with the addition of a new manufacturing process that's only just now feasible by their fab TSMC. They both have alot of work to go with getting this chip made.

You guys make me laugh in thinking that developing a chip is a month or two deal and or moving to another process is a simple transition. It most certainly is not.