The future of processors!!!

[-StrY-]

http://www.fudzilla.com/index.php?op...2934&Itemid=38

# Quote-

A group of MIT boffins think that an experimental material called graphene may be able to boost the clock speed of computer chips into the 500-GHz to 1,000-GHz range.

According to a report issued by the MIT news office, Tomas Palacios, assistant professor in MIT's Department of Electrical Engineering and Computer Science (EECS) has been doing groundwork on using Graphene instead of silicon. Graphene is a form of pure carbon that was identified in 2004. It's only an atom thick.

Palacios, Jing Kong; Han Wang and Daniel Nezich, have constructed an experimental graphene chip capable of multiplying electrical signal frequencies. They say that frequency multiplication technology generate noisy signals and require significant filtering and power.

However the graphene chip uses only a single transistor and its output is clean, and needs no filtering. The chip has been clocked at speeds of between 500-1000 GHz. Palacios believes his chips will be in the shops in a year or two as soon as they can work out a way to grow graphene wafers suitable for use in electronic manufacturing.

Modern CPU's use 150 - 350 Million transistors to achieve max 6.5 GHz. With Graphene they have achieved speeds of 26GHz from a single transistor.

Imagine the possibilities! 0.0

[Wesley]

Imagine the possibilities! 0.0

Imagine the price! I can assume only enterprise servers will be able to benefit from this technology in the interim.

[Gambit]

http://www.fudzilla.com/index.php?op...2934&Itemid=38

# Quote-
Palacios believes his chips will be in the shops in a year or two as soon as they can work out a way to grow graphene wafers suitable for use in electronic manufacturing.

I believe 'Palacios' is ****ing idiot.

[KFC]

Im sorry but I have high hopes that in 10 years we will get it on the market.
BUT: For what is this much power used???

[Signor 65]

It just struck me. At first I was like "Wow no way! That's really awesome!" but then I saw the date of origin... 1 April. I'm pretty sure this is no coincidence... happy April Fools day everyone...

[Zen]

It just struck me. At first I was like "Wow no way! That's really awesome!" but then I saw the date of origin... 1 April. I'm pretty sure this is no coincidence... happy April Fools day everyone...

Not likely, seeing as IBM has already demonstrated a chip hitting 500GHz a few years back. Sure it was cooled by LHe, but the point remains, this feat was achieved (somewhat) before.

[Wesley]

I believe 'Palacios' is ****ing idiot.

And I believe you are not thinking straight. Consider the huge jump in computing power since the release of the Pentium 4 in 2004 - We've more than doubled performance in games alone, and at the rate we're advancing, the idea of having those chips sitting in a shop is suddenly not too far-fetched. However, they would still be so expansive that only companies running massive servers will require them.

Im sorry but I have high hopes that in 10 years we will get it on the market.
BUT: For what is this much power used???

Google, what else? :p Having this much power in one chip alone would easily be enough to run half of Google's servers. Imagine NASA with just ten of these babies - if their performance scales with clockspeed, everything would theoretically compute more than five times faster (if all their servers were running 3.2Ghz Nehalem chips).

One of these chips would be able to map the human DNA in months, not years. And, of course, if you try running Crysis on it via Ray Tracing, you'd probably be able to run it on the highest settings possible, with some good frames to boot ;)

Not likely, seeing as IBM has already demonstrated a chip hitting 500GHz a few years back. Sure it was cooled by LHe, but the point remains, this feat was achieved (somewhat) before.

I cant find the article now, but a while back Intel super-cooled a Pentium 4 using L2N and a vacuum tube, hitting 341.2Ghz stably (god please give me the motherboard that ran it!!!!!!!!).

And yes, IBM did in fact hit 500Ghz in June 2006. Completely stupidly fast, if you ask me. Oh, by the way, the chip they used runs at 350Ghz, room temperature. Anyone want one?

HANG ON!

Palacios believes his chips will be in the shops in a year or two as soon as they can work out a way to grow graphene wafers suitable for use in electronic manufacturing.

Did he say grow? Organic computer chips? My my, this keeps getting better and better.

[wir]

Did he say grow? Organic computer chips? My my, this keeps getting better and better.

Wraith technology, here we come. Just need some Iratus Bugs now.

[Xm@niac!]

This is only the beginnig.

Excuse me, I need to go and drool all over the specs of the Core i7...what a thing of beauty. ( until this Graphene thing arrives)

[Wesley]

This is only the beginnig.

Excuse me, I need to go and drool all over the specs of the Core i7...what a thing of beauty. ( until this Graphene thing arrives)

Gah, the Core i7. Just look at the price of the budget i7 processor: R4070 ! I could build a decent budget gaming computer for that price! (no, really, I've done it before) I could easily build an entire computer that'd keep up with the i7 965 for the same price of the procesor!

i7 is nothing but another way that Intel can show the world that it could totally dominate the market if it wanted. Its a daunting display of power, and AMD is trying very hard to tone it down. And it costs so much! Sheesh, I could fix up my Bug nicely with sixteen odd grand ;)

[Frozenfireside]

I could easily build an entire computer that'd keep up with the i7 965 for the same price of the procesor!

Pics or it didn't happen.

[Chevron]

Pics or it didn't happen.

Depends on how you define, "keep up".

[Gambit]

And I believe you are not thinking straight. Consider the huge jump in computing power since the release of the Pentium 4 in 2004 - We've more than doubled performance in games alone, and at the rate we're advancing, the idea of having those chips sitting in a shop is suddenly not too far-fetched. However, they would still be so expansive that only companies running massive servers will require them.

I was not referring to the fact that that speed can be achieved but the statement that they: "Palacios believes his chips will be in the shops in a year or two as soon as they can work out a way to grow graphene wafers suitable for use in electronic manufacturing." :/ look closely before posting a smart comment.

[Wesley]

Depends on how you define, "keep up".

Well, in games at the very least, it'll be able to keep up with the pace of a i7 965 ;-) In other apps the i7 will run away from just about anything.

[X-Gamer]

When they say "grow", its simply a Manufacturing process, not Organic.

And they only got ONE single transistor to run at 25Ghz, what they gonna do? Turn an LED on and off REal fast?

Anyway, going from one transistor to the millions required in a processor will take some doing.

[Xm@niac!]

Gah, the Core i7. Just look at the price of the budget i7 processor: R4070 ! I could build a decent budget gaming computer for that price! (no, really, I've done it before) I could easily build an entire computer that'd keep up with the i7 965 for the same price of the procesor!

i7 is nothing but another way that Intel can show the world that it could totally dominate the market if it wanted. Its a daunting display of power, and AMD is trying very hard to tone it down. And it costs so much! Sheesh, I could fix up my Bug nicely with sixteen odd grand ;)

I only said I am going to drool over one, not go out and buy one!!!! I agree, the price of a core i7 is waaaay past my budget. Can get more than a great card for that price, or buy a computer youre right. But hey, it is still a thing of beauty now matter how much it costs.

[Wesley]

When they say "grow", its simply a Manufacturing process, not Organic.

And they only got ONE single transistor to run at 25Ghz, what they gonna do? Turn an LED on and off REal fast?

Anyway, going from one transistor to the millions required in a processor will take some doing.

You do realise there's absolutely zero mention of what process was used to make the chip? Consider this: take the E1400, slap the design onto a graphene layer using a 45nm process, and do the magic with the transistors to make it run at +300Ghz at room temperature. Load up Folding@Home, and watch your points streak past everyone else's in a matter of minutes - and thats not even using all the chips power. Hell, its not even a powerful design - try the same thing with an i7 on the stuff and watch those pigs fly past your window.

Whats the record, one billion plus transistors on a Core i7?