no idea.. maybe. I haven\'t given much though to it, so perhaps that would be possible. As for how fast the Cell is...
Hmm, I don\'t want to get hyped up about it, but I think it\'s fairly powerful. Last I heard, one Cell could be able do 1 TFLOP (compared to the EE\'s 6.2 GFLOPs). I also heared that the PS3 might be just about 1000 times more powerful than the PS2, although that can mean anything. 1000 times the polygons capability? If this would be the case, we\'d have roughly
20 million * 1000 =
20 billion Polygons/sec.
LOL. No way.
Honestly, no matter how powerful the Cell may be, I doubt it can do that many polygons. But perhaps 1 billion to 6 billion polygons might be fairly realistic?
Well, the PSX was able to do 360\'000 polygons/sec (I think raw). The PS2 can do 66 million (raw) polygons/sec. That\'s a 183 times increase in 5 years. If the PS3 manages to beat out the PS2 by 183 times in polygon performance, we might be looking at 12\'078 million (raw) polygons. That\'s 12 billion.... not bad. The PS2 is able to do about 20 million in-game, so we might be looking at about
3+ billion polygons/sec for PS3.
Also, if the 1 TFLOP figure is correct, that\'s a 160 times increase over the PS2\'s 6.2 GFLOPs, so that might just underline our polygon number further up.
While we\'re at polygon numbers. 3+ billion polygons/sec sounds mighty impressive but this is still at least 3 years off and I fully expect the PC industry to come a long way in this time. I must say though that I hope that Sony/Tosh/IBM really max out the Cell and design something that not even the PC industry can dodge for at least a few years. I think it\'s possible, seeing that the PC industry still has the biggest disadvantage of having a bottlenecked architecture. The other thing is, the Cell is said to be built with parallelism in mind (like connecting various Cells to gain much more power). Does that mean the PS3 might be powered by multiple Cells maybe? If so, we could have a significant increase in performance...