super fast CPU's tested- 250 times current speed

Fallschirmjager

from the new york times....but dont exactly hold on buying yer cpu..

---

Researchers Say New Chip Breaks Speed Record

Article Tools Sponsored By
By LAURIE J. FLYNN
Published: June 20, 2006

Researchers at I.B.M. and the Georgia Institute of Technology are set to announce today that they have broken the speed record for silicon-based chips with a semiconductor that operates 250 times faster than chips commonly used today.

The achievement is a major step in the evolution of computer semiconductor technology that could eventually lead to faster networks and more powerful electronics at lower prices, said Bernard Meyerson, vice president and chief technologist in I.B.M.'s systems and technology group. He said developments like this one typically found their way into commercial products in 12 to 24 months.

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.

At 500 gigahertz, the technology is 250 times faster than chips in today's cellphones, which operate at 2 gigahertz. At room temperature, the chips operate at 350 gigahertz, far faster than other chips in commercial use today.

Mr. Meyerson compared the achievement to the development of the chips used in Wi-Fi networks. It was not until the semiconductor technology used in those networks was produced with silicon that wireless networking become affordable for consumer applications.

Dan Olds, a principal at the Gabriel Consulting Group, a technology consulting firm in Portland, Ore., said the development was significant because it showed that the chip industry had not yet reached its upper limits. "There's been talk that we've started to hit the physical limitations of chip performance," he said. "The news here is that we're not coming anywhere near the end in what processors are capable of."

Mr. Olds cautioned, however, that the technology was far from finding its way into commercial products any time soon, considering the performance leap it represents. Today's performance-hungry computer buyers, for example, are buying machines operating at about three gigahertz, he said.

John D. Cressler, a professor in Georgia Tech's School of Electrical and Computer Engineering and a researcher at the Georgia Electronic Design Center, said the work "redefines the upper bounds of what is possible" using silicon-germanium.

The research group included students from Georgia Tech and Korea University in South Korea, and researchers from I.B.M. Microelectronics. The results will be reported in the July issue of the technical journal IEEE Electron Device Letters.

---

end


you can see the 999 calls already...howya can you get me an ambulance, i just broke the helium coolant in me pc and snapped me arm off...

djmarkus

Im cynical as you know, ted..

DemonOfTheFall

Fallschirmjager wrote:

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

Didn't take me long to realise that it wasn't worth reading...

g5hn710m4xpdwy

go home :rolleyes: now just imagine if they actualy produced em' what would they cost?

Stark

Using cryogenics to speed up chips is nothing new. They simply cryogenically freeze a man, who wakes up in a futuristic world 150 years from now, buys a ****load of cheap CPUs, then steals a time machine to travel back in time with them.

L31mr0d

at the rate that they release new technology into the market, it will probably 20 years before we even see this kind of technology. But like the article said, whats the point of having 350Ghz processors if there is no real hardware to support it at the moment, we'll have to wait until the entire market for computers catches up.

I wonder how they achieved this leap though, is it a new architecture or form factor, it must be a lot less than 15nm to achieve speeds like that at room temperature.

Stark

It wasn't at room temperature. They had to cool the chips to 9 Kelvins in order to achieve the speeds.

B00MSTICK

But 350ghz at room temp is not bad.

Nukem

Stark wrote:

They had to cool the chips to 9 Kelvins in order to achieve the speeds.

Well some one is telling prokies. Absolute Zero is 0.15 K or -273°C and thats hard to reach so i reckon that s a fudged report too

SouperComputer

One step forward and 350 back. All talk of GHZ and nothing about efficiency or actual processing power. Pointless.

gline

doesnt say what architecture the cpu was using.. in fact, doesnt mention anything about the cpu at all really :rolleyes:

Stark

It was probably an AND gate

po0k

Interesting from a physics pov.

For practical, near-future advancements look to germanium as a silicon alternative/additive, rotary CMOS clocks and better controllers for the core(s)-memory operations - particularly on-die cache.

Speed when defined as clock rate means absolutely nothing.

Fnck 'speed'. Far far wider packages (16-32 cores), low clock speed, low voltage, low capacitance, unified cache and unified main memory, on-package power regulation and rotary clock gens.

Rethink how performance is engineered, and particularly, focus on performance per watt.

Fnck nVidia in their swollen complacent, marketing department-driven asses with quad-SLI 300W cores. And shame on ATI for traipsing down the same route.
Smell that? That'd be the smell of singed ass hair.

SouperComputer

For practical, near-future advancements look to germanium as a silicon alternative/additive

Hasnt germanium been in use since the 60's? I know it was used in diodes and early transistors anyway.

po0k

Not in the manner or scale it's being implemented now.
It's far more expensive a material to use compared to Silicon.

g5hn710m4xpdwy

if it was true, it could be 3.5THZ

youcancallmeal

Its more chemistry than it is electronics these days

masteroftherealm

They not for PC's, they comms chips, used for frequency generation

secret_squirrel

That was a very poorly written and researched article.

The same story is covered on theregister.com in much better fashion.

For a start it was a comms equipment chip, that you would get in network hardware not a cell phone chip.

I would be stunned to find out any currently available cell phone cpu runs at 2Ghz. Even the ARM chips in PocketPC's only get to around 500-600Khz.

Very poor journalism.

Cuddlesworth

Stark wrote:

Using cryogenics to speed up chips is nothing new. They simply cryogenically freeze a man, who wakes up in a futuristic world 150 years from now, buys a ****load of cheap CPUs, then steals a time machine to travel back in time with them.

Thanks for the new sig material

SouperComputer

masteroftherealm wrote:

They not for PC's, they comms chips, used for frequency generation

Well that makes sense, Cellphone = 2.4Ghz

conzy

Yup, It has nothing to do with faster CPUs, but it means better, faster Wifi!!

W00t:D

Dooom

Stark wrote:

Using cryogenics to speed up chips is nothing new. They simply cryogenically freeze a man, who wakes up in a futuristic world 150 years from now, buys a ****load of cheap CPUs, then steals a time machine to travel back in time with them.

Demolition Man! Oh..wait..

masteroftherealm

SouperComputer wrote:

Well that makes sense, Cellphone = 2.4Ghz

Yup exactly, just realllllly badly written article making it out to be a microprocessor, not a processor like it is ;p

Its all in the details people ;p

po0k

Cellphones (GSM) are not 2.4Ghz. That's 802.11/b/g, ISM spectrum.
GSM is 900, 1800, 1900Mhz bands.
Processor clock speed and radio frequency are rather different things.
Even if the wagon who wrote the article was talking about 2.5/3G/WiMAX handsets, she's still giving the wrong impression.

This is what happens with mainstream media try and cover tech. They fnck it up because they don't use people who know what they're talking about.

secret_squirrel/masteroftherealm, thanks for clarifying. I just buzzed through the article, hadn't checked thereg yesterday.

I assume this would find applications in the generation-after-next of optical networking? 100Gbps+ down a single fibre???

PogMoThoin

http://news.bbc.co.uk/1/hi/technology/5099584.stm

They might believe ya now.

masteroftherealm

PogMoThoin wrote:

http://news.bbc.co.uk/1/hi/technology/5099584.stm

They might believe ya now.

Wha ya mean?

SouperComputer

Cellphones (GSM) are not 2.4Ghz. That's 802.11/b/g, ISM spectrum.
GSM is 900, 1800, 1900Mhz bands.

I sit corrected, I was thinking of DECT cordless phones.

masteroftherealm

For all u hugh speed junkies.
http://www.intel.com/technology/silicon/micron.htm#tera

Intels demo of a terahertz.
Clock speed is nothing on simple chip design.

masteroftherealm

http://arstechnica.com/news.ars/post/20060622-7117.html

This guy explains it better than me ;p

Capt'n Midnight

SyxPak wrote:

For practical, near-future advancements look to germanium .

That can't be right ?

Gallium Arsenide is the technology of the future.
Always was. Always will be

http://www.pnas.org/cgi/content/full/98/2/410

The cyclohexadiene/hexatriene electrocyclic ring opening, occurring in less than 100 fs, may be considered as a prototype of a molecular ultrafast switch.

That would equate to a switching speed of 10 THz

A four-level scheme that represents the ultrafast rearrangement during a enol-keto isomerization. The transfer times are experimental (19). The Tinuvin molecule shown (2-(2'-hydroxy-5' methylphenyl)benzotriazole) is special among the systems exhibiting proton transfer in the excited state in that it undergoes the complete cycle within less than a picosecond.

Ok the complete cycle is only 1THz, but there are four distinct states..