500GHZ CPUs coming soon

[mcmikemc]

http://www.nytimes.com/2006/06/20/techn...r=rssuserland&emc=rss&pagewanted=print
I think I could live with the 350GHz model.

EDIT: The last link did not work but it is fixed now.

 

[biostud]

'chip' not = CPU

 

[Fraggable]

At 500 gigahertz, the technology is 250 times faster than chips in today's cellphones, which operate at 2 gigahertz.

I get this vague feeling we're not talking about CPUs.

 

[mcmikemc]

It seems they are talking about chips that are signal proccessors.

The reason why I posted this here is becasue it will have an impact on CPUs. This impact might not push CPUs to 500GHZ but it will increase the speed in some way.

 

[BucsMAN3K]

The researchers, using a cryogenic test station, achieved the speed milestone by "freezing" the chip to 451 degrees below zero Fahrenheit, using liquid helium. That temperature, normally found only in outer space, is just nine degrees above absolute zero, or the temperature at which all movement is thought to cease.

I wonder what the stable frequency for the chip was at room temperature...

I bet I could hit my processor to 100+ Ghz at 9 degrees above freaking absolute zero.

 

[dguy6789]

Originally posted by: BucsMAN3K

The researchers, using a cryogenic test station, achieved the speed milestone by "freezing" the chip to 451 degrees below zero Fahrenheit, using liquid helium. That temperature, normally found only in outer space, is just nine degrees above absolute zero, or the temperature at which all movement is thought to cease.

I wonder what the stable frequency for the chip was at room temperature...

I bet I could hit my processor to 100+ Ghz at 9 degrees above freaking absolute zero.

They said 350Ghz at room temperature.

 

[zephyrprime]

Ok, this article is extremely misleading. Whoever was the original source of this press release must not have know what he was talking about and now every news source is writing this up inaccurately.

First, this is probably a signal processor, not a cpu. Second, this is silicon-germanium, not plain silicon. Third, this is probably just a ring oscillator circuit and not a longer, slower circuit like that found in a cpu.

I don't know what the current state of sige development is so I can't say how this new record compares to prior devices.

 

[Griswold]

Originally posted by: zephyrprime

First, this is probably a signal processor, not a cpu. Second, this is silicon-germanium, not plain silicon. Third, this is probably just a ring oscillator circuit and not a longer, slower circuit like that found in a cpu.

The interesting thing is, SiGe is the same material AMD will use for the upcoming 65nm process though - developed with IBM.

But yea, this is not related to CPUs we know.

 

[Ramzie]

the difference here is these processors (the ones IBM makes for the government etc and large communications companys) work differently then desktop CPUs in almost every way besides basic CPU theory. Its kind of like back in the day trying to run windows on apple hardware, incompatible because of how it functions.

These 500ghz processors simply are too .... redicoules really to make their way into the enthusiast market.

The world in general likes to keep this technology behind by 20-30 years before it enters the hands of people like us.

It is kind of cool though to know that there are 500ghz CPUs out there even if they are QUITE different from the CPUs were talking about. Makes you wonder what kind of rendering power the government might have too! 32 400GHZ core 1300THZ memory with every board having 1TB memory in SLI! OMG! still cant play oblivion at full settings tho ;P

 

[hardwareking]

500 ghz huh.Well welcome netburst 2 then.Maybe they'll give a sub-zero cooling solutions along with each chip.(sarcasm.)
I don't see how communication chips will have any sort of impact on cpus.

 

[TecHNooB]

Isn't IBM also working on light powered processors? They can probably reach 500Ghz

 

[TrevorRC]

There are superconducting processors... only probelm is latency so they have to use optical cabling instead of copper [Copper would apparently melt at ~100/200Ghz ]

LN2 cooled, of course, but it's still cheaper than present-day supercomputers which release ridiculous amounts of heat...

--Trevor

 

[imported_Questar]

The interesting thing is, SiGe is the same material AMD will use for the upcoming 65nm process though - developed with IBM

Even more interesting, Intel already uses it an 90nm.

 

[Griswold]

Originally posted by: Questar

The interesting thing is, SiGe is the same material AMD will use for the upcoming 65nm process though - developed with IBM

Even more interesting, Intel already uses it an 90nm.

Its not the same technique. e-SiGe DSL - Intel doesnt use that.

A clicky just for you.

 

[imported_Questar]

Originally posted by: Griswold

Originally posted by: Questar

The interesting thing is, SiGe is the same material AMD will use for the upcoming 65nm process though - developed with IBM

Even more interesting, Intel already uses it an 90nm.

Its not the same technique. e-SiGe DSL - Intel doesnt use that.

A clicky just for you.

Essentially the same.
Link provided by Viditor.

"AMD and IBM last year used a dual-stress-liner approach for their 90-nm transistors, putting differently configured nitride capping layers on top of the NMOS and PMOS transistors. At the 65-nm node, the partners added an embedded SiGe layer similar to what Intel used at the 90-nm node."

http://www.crn.com/sections/breakingnews/breakingnews.jhtml?articleId=174901579

 

[Griswold]

Similar != same

 

[imported_Questar]

Originally posted by: Griswold
Similar != same

Of course it's not the same. Intel invented SiGe, and it's patented. IBM's process would have to be slightly different.

 

[Noobody]

who's cell phone chip has 2ghz chip? it is just wrong

 

[Regs]

Well I would just love to see them try that on a superscaler processor.

 

[F1shF4t]

Originally posted by: Noobody
who's cell phone chip has 2ghz chip? it is just wrong

you could prolly do that, but clock speed does not mean more performance, for example u could have a 2ghz chip in a mobile, it would just do a hell of a lot less per clock and will be very simple.

 

[SonicIce]

Originally posted by: zephyrprime
Ok, this article is extremely misleading. Whoever was the original source of this press release must not have know what he was talking about and now every news source is writing this up inaccurately.

First, this is probably a signal processor, not a cpu. Second, this is silicon-germanium, not plain silicon. Third, this is probably just a ring oscillator circuit and not a longer, slower circuit like that found in a cpu.

I don't know what the current state of sige development is so I can't say how this new record compares to prior devices.

it sounds like they're playing the telephone game.