PDA

View Full Version : This Has to be in my next computer



gjjackson
06-20-2006, 08:07 PM
250 GHz at room temperature (http://www.nytimes.com/2006/06/20/technology/20chip.html?ei=5090&en=215511bacfc970b5&ex=1308456000&partner=rssuserland&emc=rss&pagewanted=print)

Seen this and was amazed at what may be on the horizon.

Meaty
06-21-2006, 08:08 AM
Pretty neat, but does that mean these chips will be sold with liquid helium cooling components? Sounds kind of.... volatile!!! :hey:

lilrayray77
06-21-2006, 08:32 AM
Umm, so how big is this thing? And where would you get a crygenic freezer? Sounds pretty amazing.

Martin Adams
06-21-2006, 09:05 AM
No, no, forget 250GHz - its actually 350GHz at room temperature and 500GHz when overclocked and cooled to about 4 degrees celcius above absolue zero.
http://www.reghardware.co.uk/2006/06/20/ibm_overclocking_feat/

5 house points to the company I work for :)

lots
06-21-2006, 10:58 AM
Just thought I'd add something to the mix. IBM's recent advance is in transistor speed. The thing is able to hit 500GHz at 4C above absolute zero.

This does NOT translate into CPU speed. This is just one transistor. IBM does not say how big this thing is, if it can fit into modern day CPUs (built on the 65 nm and smaller manufacturing processes) or if it needs more space than that (in its current form).

Secondly, information traveling through a CPU's pipeline has to get to certain stages in order to complete a CPU cycle (which is the clock speed of the CPU). So modern CPUs already have transistors that are faster than the overall clock speed of the CPU, but this is because a CPU is many complicated parts. A transistor is one of the basic building blocks of these complicated parts.

Third, comparing a transistor speed to CPU clock speed is like comparing apples to oranges. :P

Just for kicks, here's a press release from Intel, in June 2001:
http://www.intel.com/pressroom/archive/releases/20011008tech.htm

Here's a quote:

In June, Intel scientists unveiled the world's fastest transistors, running at a blistering 1.5 Terahertz (1,500 Gigahertz), and featuring structures as thin as three atomic layers.

Cageman
06-21-2006, 11:26 AM
What is the difference between a semiconductor and a transistor? These things are somewhat confusing to me. :) *lol*

Captain Obvious
06-21-2006, 02:24 PM
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.
:screwy:

Yeah, good luck with that whole science thing... :rolleyes:

lots
06-21-2006, 03:00 PM
A transistor is an electrical device that does a variety of tasks (depending on the type of transistor, BJT or FET) These functions include switching, signal modulation, amplification, etc. A transistor is a solid state semiconductor.

A semiconductor is a material with an electrical conductivity that is intermediate between that of an insulator and a conductor. By that I mean, at low temperatures, a semiconductor acts as an insulator, at high temperatures, it acts as a conductor (albeit, not as effective as a true conductor).

go look through wikipedia.org for more information on these subjects :)

Signal to Noise
06-21-2006, 04:16 PM
It'll be tough going lugging a freezer on your back for that 500GHz laptop! ;)

newtekker04
06-21-2006, 11:20 PM
Superconducting is the way of the future.:D Should only take about 1 sec to render a hi-res scene with radiosity, with HDRI. :D

RedBull
06-21-2006, 11:41 PM
I get sick of these announcements.
Motorola announced a 70Ghz CPU a few years back.
http://www.geek.com/news/geeknews/2001sep/chi20010904007687.htm
Never to be mentioned or heard of again. It's due in 2005.... ROFL........

These are technologies displayed and designed to temporarily boost share prices, when bad forcasts are shown in quartely reports.
The actual validity of any of them never amounts to anything IMO.

I ask you if Intel could sell you a 5Ghz, a 6Ghz and a 7Ghz processer they would..... They like the small incremental speed increases, they make far more money from bogus theoretical laws like Moores Law.

Do you ever think Fuel Companies, and Car manufacturers like Ford, GM would allow a car that runs on water.....?

Well semiconducter makers, would never commercially release a technology that has the potential to ruin so many large money making schemes.

Bog
06-22-2006, 03:09 AM
Hum. I notice you can get a Dell box doing 4.some-odd GHz at the moment, shipped as overclocked and with some fairly sincere cooling on it. I grok that transistor speed <> CPU speed and all that... but is it just me, or has the actual increment we've been seeing in CPUs plateau'ed somewhat? Seems like we've been tooling along at 3.2GHz as the "fastest sensible CPU to buy" for about two years now.

zapper1998
06-22-2006, 05:19 AM
Hey what about the new "CELL" Chip Technology?????

Bog
06-22-2006, 05:29 AM
Take a look at the memory read speed in this article (http://www.theinquirer.net/?article=32171).

Cell doesn't look so good for anything demanding, for example, performance.

Captain Obvious
06-22-2006, 05:32 AM
Seems like we've been tooling along at 3.2GHz as the "fastest sensible CPU to buy" for about two years now.
Well yeah, but a 3.2GHz single-core Pentium 4 is hardly comparable to the upcoming 3GHz dual-core Woodcrest.


Oh, and the Inquirer should be taken out back and put down. It's a worthless news source. Seriously, their idea of omg NEWS!!11 is "the other day at Home Depot, I overheard two people I'm pretty sure work for Intel talking about the new chips."

gjjackson
06-22-2006, 05:56 AM
A transistor is an electrical device that does a variety of tasks (depending on the type of transistor, BJT or FET) These functions include switching, signal modulation, amplification, etc. A transistor is a solid state semiconductor.

A semiconductor is a material with an electrical conductivity that is intermediate between that of an insulator and a conductor. By that I mean, at low temperatures, a semiconductor acts as an insulator, at high temperatures, it acts as a conductor (albeit, not as effective as a true conductor).

go look through wikipedia.org for more information on these subjects :)

A transistor IS a semiconductor. That's why the amplification properties.

I remember when it was thought manufacturers would maybe be able to develop a 100Mhz computer by the year 2000. Back in the 8Mhz days. Never know.

lots
06-22-2006, 12:05 PM
I never said it wasnt :P :) if you go look at the wikipedia entry for transistors, you will see that it is a type of semiconductor

gjjackson
06-22-2006, 01:08 PM
I glanced more than I read. I don't put much reference in Wikipedia considering the way that site is set up. There have been some flagrant mistakes, if not outright misinformation in some areas. I would put more trust in a more reputable source.

mattclary
06-23-2006, 05:55 AM
A couple of helpful articles:

http://electronics.howstuffworks.com/diode.htm

http://computer.howstuffworks.com/question307.htm

mattclary
06-23-2006, 06:29 AM
More info:

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

lots
06-23-2006, 07:20 AM
Heh good matt beat me to it :P

mattclary
06-23-2006, 07:41 AM
Heh good matt beat me to it :P

That's OK, you got me on CGTalk. ;)

Did you see that dude's mention of quantum dots in the PS3? Is it just me or is he smoking crack?

lots
06-23-2006, 08:32 AM
well the wiki you posted mentioned quantum dots used to create the blue lasers used in blue ray ...

That much of it makes sense I suppose. Its pretty difficult to get a blue laser ;)

mattclary
06-23-2006, 08:44 AM
well the wiki you posted mentioned quantum dots used to create the blue lasers used in blue ray ...

That much of it makes sense I suppose. Its pretty difficult to get a blue laser ;)

Ahhhhh.

lots
06-23-2006, 08:59 AM
Though I have not actually looked into how blue ray is set up, so I am just going by what was posted in wiki.. which can or cannot be right :)

gjjackson
06-23-2006, 11:23 AM
I remember years ago there was talk of an "organic" CPU. Evidently nothing ever came of it. But we Do need people with Imagination to take things to the next step.

lots
06-23-2006, 11:47 AM
Actually, some of that "organic" type thought did make it into CPU technology. In the passed, CPUs were attached to a cyramic (spelling?) tile of sorts. These days they're placed on an organic tile that is more resiliant to heat and has better properties advantagious to CPU construction and performance.

I've not looked too deeply into the subject, so the exact situation is a bit fuzzy to me :)