This is why no card for hardocp.

Status
Not open for further replies.
From anand.


The hardware we were sent is final hardware running RC1 drivers; we expect final drivers to be in our hands by next week. Here are the final specifications of the hardware:

Matrox Parhelia (128MB Retail Version)

- 220MHz core clock
- 275MHz DDR memory clock (17.6GB/s of memory bandwidth)
- $399 Estimated Street Price

Matrox Parhelia (128MB OEM Version)

- 200MHz core clock
- 250MHz DDR memory clock (16GB/s of memory bandwidth)
- ~$300 Estimated Street Price?

The first thing you'll notice is that the retail Parhelia has a fairly low core clock. This means that unless you're really stressing its memory bandwidth or texture unit advantages, the Parhelia's performance will be proportional to the clock speed difference between it and the GeForce4 (since they both have four rendering pipes).

The core clock is a bit lower than what Matrox expected, with 220MHz being even lower than our original 250MHz estimate. You have to remember that this is a very large chip and thus clock speeds will be limited. The price is a tad lower than originally anticipated as is the memory clock (presumably to keep the price under $400).

Rest assured MOST if not ALL Nvidia friendly sites are going to rip this card a new A@#h#$%&.

At 220mhz its goint to be about as fast as the gf4 ti 4200 in most head to head benchmarks (no aniso or FSAA). Review sites are going to have a field day as the GF4 ti 4600 puts the smack down on the parhelia. I wonder if the GF4 Ti 4600 will have faster FSAA... It might.

Even with aniso and FSAA this card is going to be a tough sell at 399$. I have a feeling that the Radeon 8500 128 will beat it in Aniso benchmarks. Hardcore gamers are simply *not* going to be turned on by this card imo.
 
I will wait before making any claims, and like I said before ...more players =better innovation =more competition = better prices =happy wives :LOL:
 
Jb,

Look even anand tell you stright up (he has the card)... Expect performance to be proportional to a GF4 of the same clock. Becuase its still a 4 pipe architecture. The GF4 may have lower raw bandwidth, but it has a host of bandwidth enhancing and saving features. We will soon see the results. But come on..

The writting is on the wall.
 
I too would wait before making any claims as to who puts "the smack down" on any one.

As has been discussed many times before, we all know that the Parhelia will be a bit slower, maybe alot, in standard benchmarks (no AA, no aniso).

But, turn on ALL the details, to maximum level, and I think its safe to say that the bandwidth of the Parhelia will pull it ahead of any currenlty available card.

Fuz
 
Lol...

This thread is actually the reason why Matrox is not sending Hellbinder a card. ;)

The writting is on the wall.

IMO, the writing says what most people are predicting. Sub Geforce4 Ti 4600 performance in non AA and aniso, but as good or better perfformance, potentially much better perofmrance, (& with better quality) with AA and aniso enabled.

I don't think that Geforce4 even offers any edge AA mode that can begin to compare, quality wise, to Parhelia's 16X FAA.

The trade-off being that Parhelia's FAA won't work in all cases, so it will be interesting to read about the success / failure rate of FAA in games. I'm also interested in seeing if Matrox supplies a "force anisotropic" option for D3D and OpenGL, or if they just leave it up to the application.
 
I guess there's no way on earth Parhelia will use a different texturing approach, texel fillrate being 1.5 times higher than on a 4200 and over twice it's memory bandwidth is completely irrelevant too.

If one strips it down assuming Parhelia is in fact about as fast as a 4200, yet manages to yield significantly higher framerates with IQ improving features enabled, where's the problem here? The wide majority of PC enthusiasts constantly straves for higher performance with AA and/or aniso enabled, more samples here, less aliasing there etc etc.

I can't play a game without AA/aniso enabled anymore. Hello am I alone here? :rolleyes:
 
Al,

Im only saying what I think many review sites are going to do. Also you can do 100 Textures per pass, It only matters if the game takes advantage of it, or is specifically designed for it.

Has any of these same benefits made the 8500 a *better* card in the mids of many many sites? After all a R8500 puts the smack down on a GF4 Ti 4600 with max aniso...

Bottom line... No one cares. (edit: most dont care)
 
Hellbinder said:
Look even anand tell you stright up (he has the card)... Expect performance to be proportional to a GF4 of the same clock. Becuase its still a 4 pipe architecture. The GF4 may have lower raw bandwidth, but it has a host of bandwidth enhancing and saving features. We will soon see the results. But come on..

Perhaps in Quake3, or similiar DX7 games, but I'm willing to bet the Parhelia will really strut its stuff in DX8 games, and when you turn on FSAA and aniso. I'm going to wait to see reviews to test those 3 aspects before I make any judgements. It's the 3 criterias I'm looking forward to the most with the Parhelia... performance with DX8, aniso, and FSAA.

Edit: Doh, just read Joe's posts... for once we're on the same page. :D

Ailuros said:
I can't play a game without AA/aniso enabled anymore. Hello am I alone here?

I'm the same way, I play all my games with both or one enabled. One of my biggest gripes with Morrowind is that I can't enable FSAA or aniso because it chokes my Ti4600 with either enabled in outdoor areas. Otherwise, most games play fine with both on, and I can't go without either.
 
this topic really should be called "Helbinder's crusade against Parhelia" or "R300 will kix everyone's a$$."

HellBinder: yes, yes, you are always right... you never make mistakes and so never learn about those anything. I know, I know. But if you don't make any mistakes, who believes you?

I am getting enough of fanboyism. :rolleyes:
 
Hellbinder[CE said:
]Al,

Im only saying what I think many review sites are going to do. Also you can do 100 Textures per pass, It only matters if the game takes advantage of it, or is specifically designed for it.

Has any of these same benefits made the 8500 a *better* card in the mids of many many sites? After all a R8500 puts the smack down on a GF4 Ti 4600 with max aniso...

Bottom line... No one cares. (edit: most dont care)

I'd rather agree with Nappe's last post. You do seem to be on a roll on more than one sites "against" Parhelia.

Who said anything about textures per pass? Maybe this will help:

4200:

250 mhz x 4 pipelines = 1000 MPixels/sec Pixel fillrate
250 mhz x 4 pipelines x 2 TMU's = 2000 MTexels/sec Texel fillrate
250mhz DDR SDRAM/128bit bus = 8GB/sec memory bandwidth

Parhelia:

220mhz x 4 pipelines = 880 MPixels/sec Pixel fillrate
220mhz x 4 pipelines x 4 TMU's = 3520 MTexels/sec Texel fillrate
275mhz DDR SDRAM/256bit bus = 17.6GB/sec memory bandwidth

As far as the 8500 vs "any" GeForce, websites, etc etc. Each has his positive and it's negative aspects. I don't see perfection in any of those two by the way. If you stip it all down the GF4 line will win over due to it's mature and stable drivers. ATI has made gigantic leaps in the past two years. If they manage to set their software on the same level as NVIDIA's in every aspect then they'll clearly win the performance crown.

As far as aniso or AA goes. Isn't it a bit out of line to compare them with vast differences in implementation? Websites do it all the time do we really have to fall into the same trap?
 
Ail, you certainly arent alone and you know it.

Morrowind, one reason I'll wait until its a bargain priced game, I might have a card capapble of running it with AA/aniso by then :) Anyway I have enough unfinished games in my collection to never again buy a game at full price when it first comes out.
 
Well, with 16GB/s it has TWICE the memory bandwith of GF4Ti4200.
It should be bloody fast when playing Doom3 8)
 
morrowind

With Morrowind, hopefully they will get some patches out that make it go faster. Some people have played around with it, and apparently it's only doing culling at the max view distance, which is why turning the view distance down has such a dramatic performance effect. I'm actually kind of curious how the kyro would fair with it.

Nite_Hawk
 
Status
Not open for further replies.
Back
Top