By using this site, you agree to our Privacy Policy and our Terms of Use. Close
dahuman said:
slowmo said:
dahuman said:
Onibaka said:

Cell is not THAT good in GPU. They went from the 2xCell PS3 in 2004/2005 to the  Cell RSX in 2006. They realized that Cell doesn't surpass even a cheap GPU in some tasks. Its a waste for a CPU, with can process complex codes, to only perform simple tasks that only requires brutal force, like a GPU.

What I find funny is that all the experience that IBM gained by developing the Cell was used to make the Xenos(x360's processor). There are even rumours about some secrets of Cell being revealed for the development of Xenos, like while Sony IBM Toshiba were developing the Cell, some workers from IBM worked in both projects. Xenos have 3 PPUs almost identical of the Cell one. Both have 3.2GHz. I think that Microsoft wanted to make a console easy to make ports from. By releasing X360 early, MS pretty much assured that would have a very good support for their plataform. For the very loyal developers, there is nothing that money can't buy...

 

Actually, I think that if the PS3 didn't have a Blu-Ray, but instead a 8800GTX as GPU, they could have outsold Xbox360by a larger margin at today.

You have one thing in reverse, The 360 Xeno came before the Cell in the experimental cycle, it's just a Tri-Core PowerPC, it's no different than any multicore CPUs we see today in it's basic design form. The current Cell and Power PC CPUs are on a completely different level in the computing world of course. Cell is really a different design and I can't say it was the best option as a dev friendly gaming only CPU but it's really efficient at pretty much everything if you can code for it correctly, fortunately, we are already there after years of criticism. HD consoels suffer the same problems, not enough memory, and outdated video chipsets, though PS3 is not suffering as badly if you make games exclusively for it in an overall sense.

The Cell was in the works for over a year before Microsoft went to IBM requesting their chip design.  In the end the core they decided upon for their CPU was the top secret one being designed for Sony.  Microsoft then started working alongside IBM on the quiet as IBM knew Sony wouldn't be happy finding their technology in the enemies hands (although IBM as partners over the rights had every right to use it).

The 360 is indeed similar to a Power PC CPU but it still isn't "just a Tri-Core PowerPC", in performance terms it buried every PowerPC CPU made up to that point in terms of performance and power consumption. 

Both chips were finished around the same time ultimately and both could have been even faster if some small contraints weren't made in the core that occasionally benefitted one design more than the other.

wait wait, it's a tri core Power PC with 2 threads on each core so how's it not "just a tri-core PowerPC?" o_O;; I'm not saying it's a bad chip since PowerPC CPUs are very effective, but it wasn't anything crazy new as far as the actual tech went by the time it came out. Anyhow, I suppose the Xenos is more based on the design of the main core from the Cell in a tri core setup with some mods, they did beat Sony on the actual manufacturing part, as far as research goes, Cell was about 1 year ahead though the end product had a more time difference. Cell really was a pain in the ass if you think about it, it was a good idea, but it's more a workstation or milti-purpose CPU, using it purely for gaming was just not a smart idea since it upped the cost of the PS3 too much at the start, shitty business decision if you ask me.


No that is factually wrong, the chip was miles ahead of any other Power PC core at the time including the best cores IBM had in the server market.  When Microsoft approached IBM in the summer of 2003 for a CPU they looked over all IBM's prospects and the Cell's core was years ahead in every category, performance and consumption.  The chip had been in development between Toshiba, Sony and IBM for 2 1/2 years at that stage too I might add. 

I suggest you read The Race For A New Game Machine by David Shippy and Mickie Phipps, it comprehensively covers the production of both the Cell and Xenon from 2003 onwards.

By the way in terms of having the core finished and running code in the lab, they were a couple of weeks apart.  Sure manufacturing onto a board there was a gap but the development ended on the actual processors themselves around the same time.

 

*It is of course technically a Power PC processor, my point was that it wasn't JUST a power PC processor.  It was quite unique at the time.