m_hael... I see you are professional developer in game industry then... Great!
Thanks for fixing up the mistakes in my post.
Can you give details please?
In the hacking section of this site (post related to efuses) we got info about the on-the-fly encryption of every data coming in or out of cpu cores with a key that is different in each core and reset (recreated) at will in each core.
So I've assumed that if a push buffer was to be prepared then updated somewhere in memory, one core (the one who wrote it) could work on it but not the two others. Can you comment this theory? I'll be happy to hear it's totally crap, but then I don't see how it can work...
About the PS3 octo-core, I've read on our favorite site homepage, that a professional developer said which CPU and GPU was the best. Ok for GPU. But for CPU he rather liked xbox360CPU, not because it was more powerful but because it was easier to program. As developper you need everything that helps to make it in time before the deadline, so I understand. But now, let's say you have plenty of time... The difference between XBOX1 and PS2 shows that more dma routes and co-processors helped to do great things on PS2 even if processors frequencies were lower than on XBOX1. Don't you think that, with time, and courage... some developpers may really take adventage of 1CPU+7SPU? Once again, I will be happy to hear it's crap, I'm not a fanboy of either side. I just want to understand how it works really.
One more precise question for you... Is your final code running as real code or is somehow interpreted by a virtualized processor? That would mean you can't really be sure you have pushed the system to 100% of its power because of all the efforts put to secure security of console...
Thank you a lots for any details you can give.