Originally posted by: PG
I don't have any reading problems. That kind of performance might be impressive IF the card were available today, but it isn't and won't be for a little while.
My main point is that by the time these new Matrox cards are widely available it will almost be time for the NV30.
Parhelia becomes available in late june / Early july. NV30 somewhere in late august /late september (no-one knows yet). There seems to be 2 months between those two. And when NV30 get's released, it's almost time for Parhelias improved version (rumoured to be released late this year, early next year). Point is that if you delay a purchase because "New and better product is just around the corner", you will end up waiting forever. There are ALWAYS new and better things coming down the pipeline!
So, if we assume that the 20-30% is correct (I don't know that is it). We could guesstimate that with finished hardware and more mature drivers Parhelia would beat GF4 4600 by... 50%. That's not impressive?? It would give NV30 a run for it's money!
I'm just worried that Matrox does not have enough of a lead to stay competitive with Nvidia and ATI for the long term. Matrox cannot keep up with the 6 month product cycles of Nvidia so this is the card Matrox will be selling for quite some time. In the meantime Nvidia and ATI might catch and pass Matrox.
Ummmm, no. There's already a .13 micron Parhelia in the pipeline. I heard that it could be released late this year, early next year. That chip would have full DX9-compliance, hardware tweaks and propably higher MHz.
How do you know Matrox can't keep up with NVIDIA? Because they have had longer product-cycles in the past? Long product-cycles aren't necessarily a bad thing. BMW and Mercedes have pretty long product-cycles too .