I think that the price is brilliant. While comparing the Parhelia to ATi and nVidia consumer-level cards makes it seem a bit expensive, it is a multi-head workstation class card that is useful for much more than gaming. It should be compared to Quadro 4s and Wildcat IIIs.
For gaming, the Parhelia offers a slew of features that make existing games look better and run better now, not at some undefined point in the future. Case in point is the pixel/vertex shading features of NV20 and NV25, which are not being used much at the moment, and are about to be made obsolete with NV30, which will have features that few games will use until 2004. Also, I think that we can trust Matrox to write drivers that don't fudge up image quality for the sake of Q3A benchmarks. I would be looking forward to the R300 if I were able to trust ATi at all...
Another subtle, but nice thing about Matrox cards is that you only have to deal with one vendor, which makes troubleshooting and support much easier. Plus, Matrox isn't going to go belly up anytime soon.
Going forward, Matrox will have to introduce cards with programmable vertex and fragment processors as well as support for Cg. Occlusion culling also needs to be implemented in the next gen Matrox cards. It's perfectly acceptable to be without those capabilities now, but not in a year or so from now, especially if Matrox wants to move into the 3D workstation market after securing its hold on the 2D workstation market while making inroads back into the gaming market.
For gaming, the Parhelia offers a slew of features that make existing games look better and run better now, not at some undefined point in the future. Case in point is the pixel/vertex shading features of NV20 and NV25, which are not being used much at the moment, and are about to be made obsolete with NV30, which will have features that few games will use until 2004. Also, I think that we can trust Matrox to write drivers that don't fudge up image quality for the sake of Q3A benchmarks. I would be looking forward to the R300 if I were able to trust ATi at all...
Another subtle, but nice thing about Matrox cards is that you only have to deal with one vendor, which makes troubleshooting and support much easier. Plus, Matrox isn't going to go belly up anytime soon.
Going forward, Matrox will have to introduce cards with programmable vertex and fragment processors as well as support for Cg. Occlusion culling also needs to be implemented in the next gen Matrox cards. It's perfectly acceptable to be without those capabilities now, but not in a year or so from now, especially if Matrox wants to move into the 3D workstation market after securing its hold on the 2D workstation market while making inroads back into the gaming market.
Comment