IBM and AMD to intro 90nm SOI process for up to a quarter bo...

[MacBytes]

Category: PowerPC
Link: IBM and AMD to intro 90nm SOI process for up to a quarter boost in performance. (http://www.macbytes.com/link.php?sid=20041213134526)
Posted on MacBytes.com (http://www.macbytes.com)

Approved by Mudbug

[shamino]

Cool! This tech is finally making its way into a product.

It was invented back in 2001 (http://www.research.ibm.com/resources/press/strainedsilicon/). I was wondering what would become of it.

[warcraftmaster]

Scientists at IBM have discovered a breakthrough method to stretch silicon, the fundamental material at the heart of microchips, that can speed the flow of electrons through transistors, increasing semiconductor performance and decreasing power consumption in semiconductors. :eek:

does any one know what this means this is mac rumors right? this is like when toshba told every one that they have a 60gb(and the 80gb).well they say a 25% in the higher clock speed . less power consumption then can clock it even higher becuse of less heat. they also say up to 35% in the link shamino gave us. well 25% of 2ghz is makes it a 2.5ghz :D and the 2.5ghz=3.125ghz so ibm can make 3Ghz now!!!!. some people where saying that ibm could of make a 3.0ghz but they couldn't get it upto production when the 2.5 came out then if they could have (3.0ghz=3.75ghz) a 3.75ghz and now if it was up to 35% higher clock it would be 4.05ghz. now if they could make a 4ghz g5 then cann't they make a g5 laptop soon? :confused:





well am just pointing out some facts think what ever you want

ps i think there going to let out a 3ghz+ g5 and jan 5 05(am getting my stock tomorro :p )

[ddtlm]

warcraftmaster:

It takes quite a while for discoveries to make it to shipping product.

[neut]

<snip>

uh, engineers jsut figured out how to use diamond instead of silicon to defeat the heat barrier... but i don't think we'll be seeing diamonds in PCs for awhile.

too much coffee? not enough schooling? too much internet me thinks.


peace.

[shamino]

does any one know what this means this is mac rumors right?
I would assume that this tech (strained-silicon) will make its way into future IBM chips (including various PowerPCs), just like silicon-on-insulator did a few years ago.

It probably won't affect the current PPC 970/970FX chips, but it will allow future chips to either run at higher clocks, perform faster at the same clock, or run cooler at the same performance. Or mabe some combination of them all.
well they say a 25% in the higher clock speed . less power consumption then can clock it even higher becuse of less heat. they also say up to 35% in the link shamino gave us. well 25% of 2ghz is makes it a 2.5ghz :D and the 2.5ghz=3.125ghz so ibm can make 3Ghz now!!!!. some people where saying that ibm could of make a 3.0ghz but they couldn't get it upto production when the 2.5 came out then if they could have (3.0ghz=3.75ghz) a 3.75ghz and now if it was up to 35% higher clock it would be 4.05ghz. now if they could make a 4ghz g5 then cann't they make a g5 laptop soon? :confused:
Don't expect to see new Macs in the short term. IBM expects to use this tech on chips that will ship six months from now. Once Apple starts getting engineering samples, figure at least another three months before they are integrated into a product and manufactured in sufficient quantity to sell.

So if everything goes perfectly, I wouldn't expect to see Apple reap the benefits before September.

I expect the initial batch of these chips to run at the same speed as existing chips, but consuming less power and generating less heat (much like when the PPC 970FX was introduced.)

This might (and I stress might) just allow the chips to run cool enough for a notebook application. But now we're into wild speculation.

The tech should also allow for higher clock speeds. If IBM gets a 3.0 or 3.5GHz chip out by December 2005, we could see them in PowerMacs by around April 2006.

Assuming things go well, of course. As we know from previous "G5" chips, things don't always go as planned. IBM may hit snags in developing these new chips. There may be problems making the whole tech work, and once they get it working, there may be problems manufacturing them in large quantities.
ps i think there going to let out a 3ghz+ g5 and jan 5 05(am getting my stock tomorro :p )
I wouldn't expect a 3GHz system for at least six more months. But we'll find out when the MWSF announcements are made.

[Mav451]

Strained silicon is the reason the the FX-series Athlon 64 continues to scale incredibly well, even on the "aging" 130nm process. It is the reason they have gone from 2.2, up to 2.6Ghz with relative ease. Rremember that these Clawhammer cores, the original Opterons started at 1.6Ghz. A 1Ghz boost on the same core? Not bad.

I can only imagine the results on a smaller process (which means more per wafer, which means more higher-clocked chips) wheee.