Profcrab said:
MikeB said:
Profcrab said:
MikeB said:
HappySqurriel said:
MikeB said:
What? Within a couple of years PS3 games will run circles around what most currently sold PCs are able to achieve by that time. Cell and Blu-Ray are going to make a long term difference. |
The Cell processor will have similar performance to processors that were released using a similar process and have a similar number of transistors |
Plenty of research documents show you are plainly wrong. In terms of raw processing power the Cell can perform multiple times faster than commonly sold x86 desktop CPUs today at all optimised software tasks. The x86 CPUs carry around lots of legacy garbage and non-crucial (note I am not saying useless to everyone) features. I look at the Cell as a processor adopting the philosophy of achieving elegance trough simplicity. |
The cell processor was designed to be a next generation workstation/server chip by IBM. Gaming applications are altogether different. Games are not designed the same way as most office and commercial apps. The true statement is that most of the power of the Cell is untapped. The ugly underside of that true statement is that games don't need most of what the Cell has to offer. Sony further handicapped it by giving a small amount of memory dedicated to it. Games on the PS3 will improve as developers use it more efficiently but what must be understood is that it is not a gaming processor. The games will not improve as dramatically as you believe. This vast untapped potential it has is better used in a server farm. This is why people have looked at the PS3 for those applications. |
There are two kinds of Cell naysayers I see dominant on forums. Those who claim the Cell is optimised for multi-media applications and thus is useless for serious stuff and other like you who seem to claim the opposite. The Cell chip is a multi-media powerhouse, excellently suited for gaming purposes as well as for building a new unrivalled supercomputer. There are however various considerations, that's why I stated back in 2005 it's going to take a long time before developers will get the most out of this CPU relating to redesigning legacy game engines. |
Cell is not the Swill Army Knife of processors. Take your nose out of Sony's and IBM's marketing announcements for the moment and look at the actual results. Sony's own devs are having a difficult time making games that look better than 360 games. Even when they do, it's not a dominating improvement. Meanwhile, IBM is selling the Cell for what it is designed, server applications. Have you seen IBM's product offerings for the Cell? Cell Blades, Cell Servers? In what way are server applications and gaming applications similar? IBM developed the tech for the Cell and it does exactly what IBM needed it to. Does it work as a gaming chip? Sort of, but IBM is the big winner in that deal. They got the server chip they wanted and Sony and Toshiba got to help pay for it's development. Go go American bullshitting! U-S-A! U-S-A! |
Actually the CELL is a mult-purpose processor. It's ability to crunch numbers are insanely high, and it's ability to stream media at a constant speed is very beneficial to gaming. With the ability to crunch numbers it can allow for more physics related actions to happen on screen at once. The ability to stream media allows detailed graphics to be constantly refreshed at really good speeds. Given the the SPEs are similar to RISC processors but run at a more demanding way, they allow up to three threads of processing per SPE which means that each SPE can be dedicated to a single purpose or even multiple purpose.
Studing the CELL's architexture, it is easy to conclude it can handle all purposes and has been used for all purposes from medical imaging to servers and in the future in the consumer PC through an add-on card.
