a good sign 65nm tech for cell ??

[kyetech]

If this has any weight, could this be a good signal for 65nm tech being ready for latter 2005 for cell? Its certainly inline with what sony wanted to achieve.


if intel can do it, maybe IBM can too? or then again it is the inq !

http://www.the-inquirer.com/?article=22041

 

[Acert93]

Complexity of the design is relevant also. This is one reason GPU makers hold off on moving their large parts onto smaller processes, but instead test their simpler mainstream cards on new processes to get a good handle on them so yields will be good when they move their high tech stuff down.

And while IBM has some advanced fabs, comparing one company to another is risky because advancements do not always come at the same pace. All companies have issues and hurdles at times, so Intel pushing forward with 65nm may not tell us about how TMC or IBM are doing.

That said, we should keep our ears open to see how IBM is doing on hitting their own projections.

 

[jvd]

aye i don't think a p4 will compare complexity wise to a cell chip

 

[Fox5]

aye i don't think a p4 will compare complexity wise to a cell chip

Transistor wise doesn't a dual core P4 have more than Cell?

 

[Npl]

aye i don't think a p4 will compare complexity wise to a cell chip

Transistor wise doesn't a dual core P4 have more than Cell?

Transistor-wise 32MB Dram has more than Cell :?

 

[ShootMyMonkey]

A massively long pipe, OOOE, self-scheduling engine with both scalar and vector pipes like P4 would probably rank as being several times more complex than CELL. CELL is made up of some pretty simple parts. The only thing from a manufacturing complexity standpoint is the fact that most of those ~240 sq mm are logic on CELL. Whereas with P4, most of the die area is cache.