Power Consumption Solved: Magnetic Processors

[elcentral]

in mobiles phone this culd get realy interesting, you culd run it for weeks if your screen is inactive most of the time

 

[ojas]

[citation][nom]caedenv[/nom]I think the point was that the HDD would screw up the proc, not the other way around[/citation]

no read that guys post again: keep the electromagnet well away from the HDD.

And for all the skeptics, i just read in the CPU magazine (computer power user):

"Initial discussions of Tri-Gate transistor designs occurred in the mid-1990s, through work done at the University of California, Berkeley. Intel made its first Tri-Gate announcement in 2002, and it published a few research papers over the next several years. In May 2011, though, the company was ready to make the transition official, announcing that all of its future transistors would feature the Tri-Gate design."

Wait 15 years to see this happen then...

 

[fir_ser]

The magnet processor is interesting, hope to see an implementation for it in the near future.

 

[KonstantinDK]

The question is - can it run Crisis?

 

[bjaminnyc]

[citation][nom]elcentral[/nom]in mobiles phone this culd get realy interesting, you culd run it for weeks if your screen is inactive most of the time[/citation]

Just keep the phone out of your pockets or your credit cards will be rendered useless.

 

[f-14]

" The magnets are also vulnerable to random "fluctuations from thermal effects, stray electromagnetic fields and other kinds of noise."

sucks if you live near high transmission power lines, radio or television stations, cell towers and microwave towers, cell phones perhaps (idkfs)
defintely not going to work for laptops and on electric trains/streetcars unless they can sheild the whole machine from this interference.

 

[FloKid]

Make it out of atoms. You can test the charge on any atom right? Well if you remove an electron, that's a 1, and if you have an electron that's a 0. No moving electrons, no giganto magnets, and it's the smallest you can make... There I just made the new supercomputer!!!

 

[WyomingKnott]

If I remember correctly, the closer that you get to the Landauer limit, the longer any calculation takes. It becomes a random walk towards the final state, and you spend a lot of time going backwards.

Scientific American had something on this in July of 1985, co-written by Landauer. It is very briefly available online for free: http://www.scientificamerican.com/article.cfm?id=the-fundamental-physical-limits-of-computation

 

[bebangs]

nice... finally, someone really understood how magnets work.

 

[dalauder]

I'd like to see something about how much this costs. If they can make parts at the 100nm x 200nm level, we're already where we were at with transistors in 1997.

I imagine if the cost wasn't prohibitive, this would be great to just build a CPU the size of a cellphone and sell it as a "5W Server Processor". Heck, you could put it in a laptop and just use the space a battery takes up and promise 20 hours battery life (how much does a screen use?). If they get the costs down--this could be awesome! Size isn't completely limiting considering that if my CPU was 10x bigger, it would still take up less space than my heatsink.

 

[dalauder]

[citation][nom]FloKid[/nom]Make it out of atoms. You can test the charge on any atom right? Well if you remove an electron, that's a 1, and if you have an electron that's a 0. No moving electrons, no giganto magnets, and it's the smallest you can make... There I just made the new supercomputer!!![/citation]Yeah, you can say something cool. Gene Roddenberry did that with Star Trek. But can you build it? Because they can build simple versions of these at UC Berkeley.

 

[eddieroolz]

At the rate technology is evolving, I have hopes for this.

 

[demonhorde665]

I could see nasa geting interested in this , , think about it , currently nasa uses 40 year old computer tech in all new space exloration plans , becuase the tech is relable doesnt over heat easy , and can stand up to "gorrila" engineering (aka some one beatign on the computer). these mangnets certainly soudn mor durible tuan today's cpu's (which are totally destroyed if you run your finger over the chip with any pressure). they certainly run cooler , and use WAAAAYYY less pwoer that conventional chips.

 

[stevo777]

This is nothing more than an academic curiosity IMO. Photonics and graphene hold much more promise.

 

[Guest]

MY QUESTION TO YOU IS. WILL THIS WORK ON A MAC. GOD BLESS

 

[Guest]

Yeah - to the people who made comments about it ruining your HDDs. It's more likely the other way round - the magnetic fields in the processors would be very tiny, very local, and *very* sensitive. I imagine most of the packaging for these CPUs will be shielding.

 

[new003]

[citation][nom]ojas[/nom]Yup it's good to dream, and it's great to keep working towards them and making them a reality. It's also good not to bee too skeptical about new tech. They may sound impractical now, but hey, 200 years ago if you told anyone about a transistor the size of a large bacterium, you'd be put in a mad-house. Actually even 70 years ago it wasn't really thought possible.Ultra low power processors may be the future to a sustainable economy...imagine comps running off solar panels, and powering Crysis 8 maxed out in full 4D?[/citation]

i dont think we can hope to live till crysis 8 lol well who knows maybe we will all become immortal and turn into androids that keep changing our internal parts. See u online then

 

[cypeq]

Very interesting project let's wait with our hopes until they have functioning memory module. Performance of this parts are still unknown.

 

[Guest]

if you use a tesla switch you can probably make current processors consume 1/1M the power, and your hard disk also!

 

[Fennecbutt]

Please don't forget that power consumption = heat. A magnetic processor could run at maximum stable speed at room temperature, no need for cooling.