Researchers at Imperial College London claim to have developed technology that would enable the creation of three-dimensional microchips, which will dramatically increase their memory capacity. According to their press release:
"by using nanotechnology it is possible to reproduce the key functions of semiconductor electronics in microchips using only the 'spin' of electrons, which is responsible for magnetism, rather than the more conventional 'charge' that traditional microchips use."
Post a Comment
Member since:
2005-07-06
Member since:
---
Nano tech is a fraud, in that we won't see any benefits from it in the next 30-50 years. These headlines are designed to increase venture capital investment into companies who will not actually produce a sellable product or IP for decades. Not exactly appetizing in these current market conditions.
Member since:---
Member since:2005-09-11
Member since:---
Computers, specifically mechanical ones, were first used to easily calculate artillery firing angles. Low tech gear systems were used to calculate the angles.
Nano tech robots fighting a virus, are you kidding me? This is such crap, we'll sooner see a full sized robot fight crime.
Be sure to invest heavily.
Member since:---
That's a stupid thing to say. Nanotech is shaping up to be a lot like advanced physics research in the 1930's-1960's. A lot of money went into the experiments that ultimately lead to the understanding of quantum mechanics we have today, and you can bet there wasn't immediate benefits to a lot of that work. Today, much of our prosperity (at least, in the developed work), comes from technologies (eg: semiconductors), that would not be possible without that research.
Advanced research is costly, yes. It doesn't necessarily have immediate benefits, true. But it's also the only way of ensuring that progress continues. Without it, the human condition will get worse, not better.
Member since:---
http://www.hi.is/~hj/QuantumMechanics/quantum.html
Quantum Physics is quite real and was quite useful AND TOOK AT LEAST 30 YEARS TO PRODUCE AN INTEGRATED CIRCUIT.
What I think is a fraud is the way nanotech is portrayed. No little robots are going to fight cancer for you, get over it. Complex behavior is difficult enough to achieve with normal sized robots and it sure ain't perfect now.
We will all be long dead before any useful nanotech is applied to everyday life. But by all means, please invest heavily.
Member since:---
Member since:2005-07-24
Member since:---
Member since:
2005-07-17
Typical that the first thought of application is further expansion of bells and whistles on mobile phones. How about something useful:
Right now, if someone calls me and I don't pick up, the caller gets diverted to (costly!) voicemail service. Many folks, including me, have this 'feature' disabled because they don't want to call back an expensive service to hear any recorded messages. Only the time and caller's number (if given) is recorded by the mobile phone.
Desired behaviour: answering machine built into the phone. Makes a lot more sense, more user-friendly, cheaper to use, and carriers needn't provide voicemail anymore (yeah I know, those carriers wouldn't see that as positive ;-).
Needed for this: simple audio encoding (MP3?), and enough memory to hold a decent amount of messages. BTW - current memory tech is already NOT a bottleneck for this application. Your phone has this built in? Please let us know make/model(s), that would be a must-have feature for my next phone.
Other applications: ehhhmm... PC's? Low-power, silent, small, shock-proof storage? Digital camera's? Digital VCR's? Countless possible applications, mobile phones among them, but certainly not a first thought.
Member since:---
Right now, if someone calls me and I don't pick up, the caller gets diverted to (costly!) voicemail service. Many folks, including me, have this 'feature' disabled because they don't want to call back an expensive service to hear any recorded messages. Only the time and caller's number (if given) is recorded by the mobile phone.
Desired behaviour: answering machine built into the phone. Makes a lot more sense, more user-friendly, cheaper to use, and carriers needn't provide voicemail anymore (yeah I know, those carriers wouldn't see that as positive ;-).
And what happens if your phone is turned off? Personally, I'm sick of all this cell phone-mania. 99% of the people that use them don't even need them.
-bytecoder
Member since:2005-07-09
> And what happens if your phone is turned off?
Actually, it can take a cue from the call centers -- include a "standby" mode so you can leave the phone if you don't want to talk. A more serious issue is what would happen if you're out of range?
About the phone-mania, I agree. I've never much cared for all-in-one devices that try to do everything, but end up doing nothing particularly good. Not only do these all-in-one devices work poorer than their more specialized counterparts, they provide a single point of failure, and reduce a person's flexibility. Mainframes lost their edge to PCs for precisely these reasons.
If companies could focus on getting single-purpose devices smaller and more convenient, few people would mind carrying dozens of them around any more than people have a problem carrying dozens of cards on in wallets or dozens of keys on their key chains. I don't know if this tech can do it, but a scaled down cheaper version of it with less capacity might.
Member since:2005-07-09
Actually, it can take a cue from the call centers -- include a "standby" mode so you can leave the phone if you don't want to talk. A more serious issue is what would happen if you're out of range?
Where I live it's just as expensive to call voice mail as it is to call the person (which you would have to do to leave the message on the phone).
I think putting voicemail inside the phone would make as much sense as letting everyone run their own e-mailserver. It's just unreliable.
The voice mail is there to take your calls if the phone is off (intentionally or out of battery), if you are out of range etc.
The service should be free though, at least for the one who's calling. (like it is in some countries)
One thing that I think would be great though is if every single message was sent in an MMS to my phone with a text telling me from which number it is, how long it is and perhaps even the amount of speech in the message. That way I wouldn't have to spend a lot of time listening to a lot of "..........click"-messages. Also I could save important messages in my phone for later.
Member since:2005-07-06
"Personally, I'm sick of all this cell phone-mania. 99% of the people that use them don't even need them."
You are right, generally speaking, but a bit over the top.
Mobile phones company would go bankrupt if everybody used them as much as I or my parents do. However there have been times when they have been extremely useful to us: in my case, for instance, at times of car breakdowns, or plainly when I had a need to communicate and no other phone was available.
However, on certain condition, mobile phones could replace traditional ones altogether:
1)They shouldn't be detrimental to your health as they are now (and I know what I am talking about)
2)Good reception should be guaranteed almost everywhere
3)The cost of a call per minute should be the same as for a land line
Even so in this country (Italy) many people have already replaced their land line for a mobile.
The reason is that the Italian Telecom "tax" (monthly standing charge) is one of the most hated in the country.
Member since:---
Member since:
2005-06-29
Member since:
2005-07-06
That means a revolution for gaming. Some 10 GB Ram on your video card is definitely more than 128 MB :-)
Never swap anymore. Put all your applications in RAM when the computer starts - so each application you need is already in RAM when you need it.
Speech recognition and visual recognition will improve, but throwing more RAM to the beast will only result in slightly better recognition. Better algorithms is what is needed in this field. Why can an insect with a very simple brain outperform a supercomputer when it comes to pattern recognition? I quess improved neural networks will do the trick.
Is depending on a "spin" reliable? I once read that the spin sometimes just changes without reason, so we really need a control bit. That means 1/8 of the storage will be lost to the control bits.
Member since:2005-07-06
Why can an insect with a very simple brain outperform a supercomputer when it comes to pattern recognition?
http://www.quantumconsciousness.org
Quantum Neural Network.
I found a few links on QANN's
http://www.cic.unb.br/~weigang/qc/aci.html
http://virtual01.lncc.br/~giraldi/qc/Quantum-Neural-Nets/Research/R...
Member since:---
Member since:---
Member since:2005-07-06
We will never reach the stage where everything is happily loaded into ram. Sadly the more power our computers gain the more we waste it..
10 years ago had someone told you 1GB memory would be standard you would have laughed at them "you'll never NEED that much memory" yet despite the fact that you could load an entire OS and office suite into memory other 'features' stop this from being a reality
Member since:---
Member since:2005-09-03
Yes speed (latency) is important, but loading ALL textures on the video card means less disk access after the load, and less Memory -> AGP/PCI-E traffic. Everything being local to the GPU makes processing the texures faster, even if the latency of the RAM is not better. Basicly the vedio RAM goes from being a cache for the Main memory (which in turn is often used to cache the harddrive) to being primary texture storage.
What excits me is the benifits this technology could have the SRAM (CPU cache) 100 MB of CPU cache would be incredible. Although returns for more cache become expodetialy less profitable.
Member since:2005-07-06
You do realise that when solid state storage gets to a decent speed, there will be no need for ram because the memory will be that fast, it can be loaded directly off the storage device rather than it being copied from storage device into memory then shunted off to the CPU for processing.
Member since:---
Member since:
---
http://osnews.com/story.php?news_id=11784
does this guys have anything to do with it? i hope they do 
--
garapheane
Member since:
2005-07-06
Member since:
2005-08-20
Particularly, does it hold it's memory reliably without power? And how does it compare speed-wise to current day ram?
It would be kind of interesting if the fastest storage space available was also the cheapest and didn't need power. You'd be looking at a fundamental change in the storage hierarchy of computers! I'm probably over-imagining things here though 
Member since:
---
Typical that the first thought of application is further expansion of bells and whistles on mobile phones. How about something useful:
---------------------------------------------------------------------- ------------
More people own cell phones than PC's. They are hitting a larger market; in essences: "Show me the money".
Member since:
---
Member since:
---
The press release says nothing detailed about the technology, how far it is away from commercial development (other than that the researchers are working with commercial developers), or anything else useful.
IF, however, their research is both valid and the process can be made reliable enough to do what is claimed at a commercially viable price, it is a very significant advance.
Nanotech will do a lot more in due time, as well. There's not much it can't do once the technigues are developed (short of dealing with subatomic or direct energy matters, which is not in its purview.)
Anybody who thinks nanotech is a fraud is an idiot.
Member since:---
Member since:
---
Member since:
2005-07-06
I know it's "dumbed down" for the readers, but I sincerely hope the physicist they talked to didn't actually say spin is responsible for magnetism. Going back to your old days of physics and chemistry, electrons exist in shells around the nucleus. By the Pauli Exclusion Principle, fermions with the same characteristics cannot exist in the same place at the same time. So how do you get multiple electrons into the same shell? One property is different, and we call that property "spin." It has NOTHING AT ALL to do with an electron spinning!
Metals share electrons in their outer shell. This sharing of electrons is what gives them their characteristics, like conductivity. Atoms tend to clump together into larger structures. Different atoms clump in different ways. Depending on the atom (and things like impurities), these larger structures of metallic atoms share their outer shells in a manner that electrons moving around the larger structure all constitute a charge in motion, creating a magnetic field. If enough of these macro structures are oriented in the same direction, you can actually measure the magnetism and use it for things like holding papers to your refrigerator.
Member since:2005-07-06
'Spin' is called what it is for a particular reason. While spin is an intrinsic property of electroncs (which can't 'spin' in the classic sense, because they are point masses), it leads to certain phenomena that a real spinning charged object would exhibit, such as a magnetic moment. To quote Wikipedia:
Particles with spin possess a magnetic moment, just like a rotating electrically charged body in classical physics. However, this magnetic moment exists even for point particles like the electron, and for electrically neutral particles like the neutron. This magnetic moment can be experimentally observed, by the deflection of particles by inhomogenous magnetic fields (as in the Stern-Gerlach experiment) or by the magnetic fields generated by the particles themselves. In fact, ferromagnetism arises from the alignment of the spins of the atoms in a solid.
To be fair, the latter information can't be found in most high-school textbooks, and is usually the subject of an introductory materials science course.
Member since:2005-08-28
Member since:2005-07-06
From a text on spin and magnetic moment of electrons:
The term "electron spin" is not to be taken literally in the classical sense as a description of the origin of the magnetic moment described above. To be sure, a spinning sphere of charge can produce a magnetic moment, but the magnitude of the magnetic moment obtained above cannot be reasonably modeled by considering the electron as a spinning sphere. High energy scattering from electrons shows no "size" of the electron down to a resolution of about 10-3 fermis, and at that size a preposterously high spin rate of some 1032 radian/s would be required to match the observed angular momentum.
The rest gets rather complicated. That ferromagnetism is due to electron magnetic moment is a hotly debated topic, given that electron magnetic moment has an unexplained "fudge factor" g in the equation. That and the fact that magnetic domains are on the order of 300 atoms of iron has led to newer explanations that don't use electron spin. It makes for fun reading.
Member since:
---
Member since:---
Member since:
2005-07-06
