Slashdot Mirror
Spintronics in your Future?
slugfro writes: "Do faster boot times and RAM that maintains memory after shutdown sound interesting? This article by a Science Magazine author details the study of utilizing the spin of electrons rather than just the charge in electronic devices (hence the name 'Spintronics'). Anyone out there researching this or have more info?" We do a story about MRAM every four months or so, and each time commercial development is a few years in the future. :)
I don't know, but I'm getting dizzy just thinkin about it.
Um, this is my sig.
hey hey! I'm Jesus! And what would you like for Christmas, little boy?
a subset of quantum computing to me. Mabey it will get us one step closer to quantum computing being a reality for joe luser.
I will bend your mind with my spoon
The cheap solution for persistant RAM is to provide a constant power source for your RAM from either some form of battery or just a power source that is always on.
Where this stuff is actually useful is for better Hard Drives. In fact it's already being used for that, and has been for a while.
An Intro to Spintronics - Univ. of Maryland
Article on Unisci about research into electronic spin in electronic devices
Just a clarificaiton. Technically its Science Magazine not "a Science Magazine" as stated on the main page.
Science is the Journal of the American Association for the Advancement of Science (AAAS) and along with "Nature" (its british counterpart) it is one of the premier general science journals.
That article seemed to get a little wacky toward the end. I thought the paper posted at the beginning of this month about Why Quantum Computing Sucks or whatever it was called made the point that measuring spins is a hugely hard and unsolved engineering problem?
I only have a doubt about this RAM that maintains memory after shutdown. If a virus gets into the memory, how does this affect the system. The same goes to memory managment.
Now I know I may be wrong or have the wrong idea since I ain't into memory allocation.
Writing: no longer done with the fountain pen, now done with an eraser.
This would be so great for portable computers.
Imagine keeping a laptop/handheld with you wherever you go, but only having it on when you need it. You can instantly close the lid on it, shutting off it's power. This would preserve battery life to possibly many days at a time for a traveler. Then when you need to use the computer, just open the lid, and it's exactly where it was when you closed it.
Desktop computers would also be nice to instantly shut off and turn on at the same state, but the portable computer market seems much more useful, especially for conserving battery life.
www.Beyond7.com Insane modern art water sculpture.
ram that retains what's in memory only really comes into it's own when it's either running on a machine that is:
- running a stable os that doesn't need flushing
- being used with a limited power supply
it'd be nice to save the environment too maybei tend to leave my computer on all the time - it's got enough to do without me using it (downloading stuff for me, curing cancer, cracking codes etc...). the only time i turn it off is when it needs it OR:
a laptop running on a battery which you'll be booting much more often (saves time) because it's off (to save power).
free (as in mp3s) electronic music
So .. when your disk drive needs repair .. take it to a quantum mechanic...
A feeling of having made the same mistake before: Deja Foobar
What happens when MS shoots me a BSOD? How the heck am I supposed to reboot if the ram is going to remember what crashed My machine in the first place?
Heh - aside from that, it would be way cool to power off the system, come back 24 hours later and have it boot up in seconds. Heck, Technically, it should be able to freeze a moment in Unreal Tournament , allow you to go get a beer, then comeback and pick up where you left off.
_ _ _ Go for the eyes Boo! GO FOR THE EYES!
More of this here: http://www.physics.umd.edu/rgroups/spin/intro.html
Uhhmmm.....
That's exactly what my Sony Vaio does actually right now, except for running for days, that, not yet!
The new technology call photocanonic storage is sort of along the same lines as this.
With photocanonic's the bits of data are fired via a microscopic light canon toward the reading head. The head does not move at all and no mechanical moving parts are required. The idea I guess is to bring the data to the head instead of using the head to scan for data.
The Cambridge Z88 mostly used RAM for storage. Changing the batteries, you had about 30 seconds before you started losing data. Of course, there was still a low-tech solution for the slow: plugging it into the mains while you changed the batteries.
Besides, I don't know about you, but I don't have to change my CMOS battery very often...
The boot up time of current computer systems takes a long time because it is doing a lot of different things: 1. Power On Self Test 2. Memory Test 3. Waiting for Devices to power up and "settle" 4. Finding first available device 5. IPL from first available device 6. Setup the processor and address spaces 7. Switch to protected mode ( Intel ) 8. Search for more devices 9. Wait for devices to settle / initialize 10. Start initial program loader 11. Run startup scripts 12. Run network startup ( wait for dhcp ) 13. Initialize Graphics System A reboot would still have to perform many of these steps whether or not the Ram remembered the previous state. Since whats in Ram is dependent on what devices are present and what addreses are assigned, you would probably have to wipe the ram on each boot anyway.
As to the spin and orbital properties of an individual atom, particularly the risk which comes with such a density of memory, what could easily and commonly affect these properties? Strong RF? Magnetic fields? Xrays? (note: laptops currently survive airport Xrays, but I'm clueless as to how this happens and whether it's something that eventually will catch up with data stored on hard drives.)
A feeling of having made the same mistake before: Deja Foobar
So you're saying i can use MRAM in my portable dildo?? great
Tom.
Oh arse
I submitted a story just recently about this story in SpaceDaily detailing a breakthrough in magnetic semiconductors, crucial to M-RAM technology. This group created a semiconductor that shows magnetic properties at room temperature and beyond. The excitement is palpable.
What the hell is photocanonics? There is no such thing!
Hmmm...
Left-pos, Left-neg, Right-pos, Right-neg.
Four-states per bit (quit??)
65,536 states per byte (quyte??)
This is computing raised to the power of two...
Everyone knows that an electron is in
either a spin-up or spin-down state.
Then there are some boundary
conditions, and a forward operator.
Besides a few other un-niceties
thats all!
We need to ask why these researchers havent delivered to us Peta-flop home computers
yet. Just what have they been spending they're research grant money on!
Under capitalism man exploits man. Under communism it's the other way around.
Does this mean that when your windows machine crashes and you reboot it will still be crashed???
is there anything faster?
The technology works real nice. I did technology development project management for the Air Force in the late 80s on this and other wild-ass NVM technologies. MRM is very sound, just too pricey for commercial use yet. The problem is that the manufacturing economics combined with market perceived risk keeps it from prime time. Flash (the next nearest alternative) is more mature, more familiar and comfortable (derived from EPROM and E^2PROM technologies which are 25-plus years old) dispite its speed and endurance shortcomings.
RAM that maintains memory after shutdown
I usually reboot my machine to clear the memory. While I there may be some benefits to keeping the memory state, like quick bootup, I boot in 20 seconds now ( 20 seconds from pressing the power button to the win98 hourglass disappearing) so that's not much of an issue for me. I would rather have the expected effect of a clean system startup on reboot than a 1 second boot time. Having the system remember bugs etc after a reboot would be awful, IMHO.
rooooar
I see lots of talk about this in the comments but does anybody really know what it is?
Why does that sound like the name of a cheesy early-nineties techno/hip-hop group? "Tonight on In Living Color, Jim Carey does something incredibly silly, and the musical stylings of SPINTRONIC!"
Error: PANTS NOT FOUND. Press <F1> to continue.
A bunch of people have posted silly questions about what happens if they get a crash or a virus or something.
All you'd have to do is to have something on the motherboard that resets the memory.
Then you'd have the normal shutdown/reset buttons, and then another button (or combination thereof) that would also wipe the memory in case of bad stuff.
There, now how hard was that...
There are alot of uses for this .....
.... and have alot more storage....
Just think about handhelds that don't need battery backup's
Also , if memory by this system is cheap enough, could it replace smaller hard drivers , on systems which access time is critical. Eg. Large Scale web sites , (search engines..etc..)
Cruise TT
I work in the quantum mechanics field and have been studying the spintronic model for a number of months now. My take on it is that it is not a feasible solution to any problem in the near future. It might seem that the tables have been turned on Einstein. The very argument he used in EPR to show QM must be incomplete requires that hidden variables models have explicit nonlocal operations. However it is experiments and not theoretical arguments that now must decide the issue. Although all experiments to date have produced results consistent with the predictions of QM, there is general agreement that the existing experiments are inconclusive3. There is no conclusive experimental confirmation of the nonlocal predictions of QM. If these experiments eventually confirm locality and not QM Einstein will be largely vindicated for exactly the reasons he gave in EPR. Final vindication will depend on the development of a more complete theory.
Most physicists (including Bell before his untimely death) believe Photocanonics is correct in predicting locality is violated. Why do they have so much more faith in the strange formalism of P.C. than in basic principles like locality or the notion that observations are produced by objective processes? I think the reason may be that they are viewing these problems in the wrong conceptual framework. The term `hidden variables' suggests a theory of classical-like particles with additional hidden variables. However quantum entanglement and the behavior of multi-particle systems strongly suggests that whatever underlies quantum effects it is nothing like classical particles. If that is so then any attempt to develop a more complete theory in this framework can only lead to frustration and failure. The fault may not be in classical principles like locality or determinism. Their failure may only be in the imagination of those who are convinced that no more complete quantum RAM is possible.
To the people who are worried about their system getting sullied by "traces" of viruses or somesuch nonsense: watch out! The Illuminati have already planted an evil "persistent storage mechanism" in your computer right now, cunningly disguised as a so-called "hard drive"! For your protection, you should wear a tinfoil hat. If you want maximum security you should locate and remove the "hard drive" from your computer put it in a microwave oven (careful to wear your tin foil hat!). This will destroy the hard drive's subspace transmitter that is being used to send information about you to a race of alien overlords who use this information in the planning of their military campaign against Earth.
Pushin' 'n dealin', shovin' 'n stealin'
I'm just throwing this out as an idea...
But could this be combined with trinary computing? You could consider whether something is not charged, and if it is, there is the spin, which would give you 3 states, perfect for trinary computing.
Is the basic phenomenon analogous in these cases? It seems like this was also reported here on /., but I can't find it right now.
A lot of people are complaining that they only reboot their machines to "clear" the RAM. And they seem to think that if Windows crashed, this new MRAM stuff will cause their computer to be in a permanent crashed state.
Well, obviously, computers making use of MRAM will have some way to purge the memory. And maybe the OS would set a flag on a normal shutdown that would tell the BIOS (or whatever it would be when this stuff comes out) that it can go ahead and just jump right to the OS (and the OS would clear that flag as it's first order of business). If the flag didn't exist, it would go through a boot sequence which involved loading the OS off a hard drive or whatever.
But let's look at the advantages of having persistent RAM. If you have a journaling file system, the journal could be kept in memory without fear that it would be lost on a crash. When the system comes back up that data would be in memory and could then be used to repair the file system. Also, disk writes would be extremely fast because they could be cached and when the system is idle or when the disk is not busy, they could be written at that time instead of having to be written to a log that is physically on disk.
Maybe, programs that are running could survive an OS crash because their state would be perfectly perserved in persistent memory.
And if CPUs had persistent registers... recovering from a power failure would be seamless.
Just some thoughts.
--
"What do you want me to do? Whack a guy? Off a guy? Whack off a guy? Cause I'm married."
See, Taco? This is the kind of thing that you get when you allow annonymous posting.
The only reason you feel safe spewing that sort of crap is because you're hiding behind the "Annonymous Coward" tag. I bet you wouldn't be spewing such shit as a logged in user, where I could track you down and KICK YOUR ASS !
I mean, as things stand, one can at least be confident that a value stored in volatile RAM is irretrievably gone if the power gets cut.
Would it be feasible to have the OS shred all memory granted to a process when the process exits?
WARNING: Anti-Microsoft comment below. If you hate such commentary, stop reading now. I don't want to hear the needless whinning about all the Anti-Microsoft commentary. Guess what? This is Slashdot! It's part of the culture here. Do go preaching Atheism in church! Don't go accusing Slashdot of being unbiased! Simple right?
---Commentary begins---
My knee-jerk reaction to this is that this type of RAM will not compatible with any version of Windows. Microsoft Windows, by Microsoft's explicit advice, often requires a reboot (memory clearing) of the machine upon which the operating system is loaded. Very often, powering off is used as opposed to a simply "reboot" or "reset" since memory is sometimes retained after an instantaneous restart. Cold booting is often preferred and recommended. (Interestingly, Microsoft also recommends periodic re-installation of the operating system after formatting the drive... apparently, it's not only system RAM that is prone to corruption.)
While it can be said that this is true of all PC operating systems, it can more accurately be said
that it's "less true" of other PC operating systems.
Essentially, since Windows relies on the "clean machine" approach to efficient running, this memory technology is unsuitable for use with Windows without a "clear memory" switch.
Pushing forward with the notion of "clean hard drives" wouldn't it be great if Windows Hard Drives came with an instant re-format feature?
You have not truly experienced the 'ring of fire' until you have tasted some fine New Mexican quisine! Green chilé will make you want to shove ice cream up your ass!
A teaching assistant (grad student) at my University was hired by the National Institute of Standards after graduation to work on quantum computing methods. According to him, they have a pretty complete theory, and have obtained some fairly large grants to actually put that theory to work. Some of the things he talked about sounded really incredible.
t io ns.html
I found a Scientific American article from 1996 which details some of the their work, as well as the work of others, plus a lot of useful links.
http://www.sciam.com/explorations/091696explora
I remember first reading about using electron spin in quantum computing way back in 1994, in a NY times science edition.
There's a company called Micromem Technologies, Inc. that is nearly ready to produce the non-volatile memory that we're all talking about. Follow the link and read their yahoo profile.
I've heard that practicing spintantrics is a good way to relieve stress.
Sound interesting? Hmmmm, no, actually it sounds just like the static memory in my BBC model B (cerc. 1982), ok so I only had 32K of SRAM (and 32K of DRAM), but same idea (kind of).
As for booting I also remember the BBC B could boot really fast. Using a trick that most slashdotter might find really interesting and have never heard of before, it's called putting the operation system on a rom. Neat yeah?
M0571y H@rml355.
And thinkpads already have that feature.
Dumps the ram on a file.
There is also a "safety net" mode where the memory is saved even if the battery life is exhausted. "The tasks are restored from the hibernation file as soon as the AC power or charged battery is installed"
Nouvelles de jeux et technologies en français. TC
Today's computers rely on silicon-based microchips to process data in a binary form -- which allows only for "on" and "off" states. Quantum computers, however, will be able to examine data using spins, which has can have many different states.
Support Texas Troops use TXGoogle
By opening and closing programs etc. you fragment the RAM (HEAP or something?) just like your HD.
I think I read that there are HD filesystems that do not fragment is this true?
I'm guessing that the way this works is some kind of permanent background defragmenting.
In any case, MRAM would need defragmenting because it would get fragmented more than normal Volatile RAM between every complete PC shutdown.
Are there OS/utilities that defragment RAM out there? How do High-Availability servers do this?
Also, MRAM without ECC would be useless right?
How about using an etch-a-sketch for storage? Black dots mean 0, clear spaces mean 1. If you get nailed with a virus, you can erase the storage by picking up your computer and giving it a good shake. Oh, I guess that means simply moving your computer could create some hassles. Nuts... have to works the bugs out of the idea.
File under 'M' for 'Manic ranting'
I already have one of those... It's called a PowerBook. Close the lid, and it goes to sleep in a fraction of a second and can remain that way for at least 2 weeks. Open it up and push any button, and 1 second later you're precisely where you left off...
Spintronics is actually old knowledge. The foundations come from the Stern-Gerlach experiment in 1923. If you guys want to know what's going on, read up on that. The whole deal is that electrons are fermions (half integer spin) particles. They therefore have a antisymmetrical permutation. This is why no two fermions can be in the same state at once and hence form singlet states. A simple magnet is the best example of how the spin of electrons play a role in our world. A magnet is composed of iron and other materials that has the magnetic moment of the electron spins all aligned in the same direction. If you heat that magnet, the spins will fall out of alignment and you'll loose your magnetism. Another important aspect is that parallell and antiparallel spins have different "resistance". What IBM and others do is use the difference in resistance due to spin alignment as a means for information. Anyway, It's much more involved than just this but I can't go into depth like this so you guys will just need to go read up on the Stern-Gerlach experiment and possibly IBM's research page on GMR (giant magnetic resistance).
Sounds like it'd be nice to have suspension in ram, but getting rid of memory resident nasties could be a lot harder.
:-)
Today (with Windows) most people are safe because they have to reboot every other day.
Any connection between your reality and mine is purely coincidental.
Good idea! I'd like to shove some nice, cold, mushy ice cream up my ass!
and I haven't eaten any spicy food this week!
I'm laughing. Because you're no where near first post.
What's so special about getting a first post, anyway? Does it make you feel all warm and fuzzy inside?
I'd recommend you find something in life other than make vain attempts at a first post, which you will probably never acheive.
I, personally would like to see more content-relavent and insightful posts than the crap the spews forth for other fellow anonymous cowards such as yourself.
So, why bother wasting your time trying to get that elusive first post? Life is short. Don't waste it on first posts!
Danged fool, its not that hard!
All you need to understand this is
Set of coupled PDE's that describe everything.
Boundary conditions
PDE solver
Then just integrate the equations forward
in "time", and all will be revealed.
Import that data into powerpoint, and
make animation of reality.
The irony in this is just too good to ignore!
While you're berating an anonymous coward, you, yourself are posting as an AC..
How ironic is that?
Can you say "memory-resident virus"? Thought so.
Yes, there are people who are clueless or ignorant enough NOT to have up-to-date virus shields. A virus that survives major system upgrades and disk sweeps would not be an impossibility. I shiver at the thought.
There is no such thing as good luck. There is only misfortune and its occasional absence.
If you don't understand look up James Blish.
Warning: this article may contain humor, sarcasm, parody, and perhaps even irony. Read at your own risk.
Hi, I'm slashdot, the host of many "fight for your right mediums" that will let you post anything you want [which is good] but will mod down anything they disagree with [is wrong just on principle]
Why do you guys even host a feedback mechanism if all you're gonna do is disagree with all competent user feedback?
Someday, I'll have a real sig.
FIRSTUS  POSTUS  BEEOTCHAE!!!! 
I r #1! All others are #2 or lower!
The irony of some overly literal nerd on slashdot feeling the need to belabor an obviously intentional joke post troll is just too delicious to withstand.
i would not want them to ever touch it. they need to be wiped after being used. no storage ever.
There are several groups working on spintronics-related research around the globe. You can check some of the research the spin-doctors are working on by looking at the Spintronics 2001 Conference webpage. Some incredible results involved researchers injecting spin-polarized current into an LED and producing Circularly Polarized Light!!! Other researchers are trying to produced spin-transistors, to switch/amplify spin-polarized currents. Many of the recent challenges involve producing spin-polarized currents, finding materials that can transport electron-spin, and injecting spin-polarized electrons into semiconductors.
The Chien group here at JHU has been the first to demonstrate experimentally the existence of a half-metal. Crystals of CrO2 have been shown to have spin-polarization of 96%. This was measured at the superconductor/ferromagnetic interface through Point-Contact Andreev Reflection (PCAR) techniques.
I'll explain some of the current concepts of spintronics, but pardon any errors as I haven't really begun my research yet. The manipulation of electron spin is an extra degree of freedom that novel electronic devices can exploit. Spintronics has already, since 13 years after the discovery of GMR (Giant Magnetoresistance) in 1988, penetrated the technology industries (magnetic storage). It's rare for such new technology like this to be commercially available so soon after its discovery. Transistors were one such monumental achievement, the first Ge transistors were available within years of the transistor's invention.
GMR is an effect that occurs with a normal metal film that is sandwiched between two ferromagnetic layers. Depending on whether the spins of the ferromagnetic layer are parallel or anti-parallel, a significant change of resistance is measured across the structure. A more useful device which extrapolates off this concept is a spin-valve This is the standard GMR trilayer, with an anti-ferromagnetic layer on the bottom. This layer pins the spin of the bottom Ferromagnetic layer. The top ferromagnetic layer can then float, and have it's spin affected by the external magnetic field. This in turn creates a magnetic-field-dependent resistance across the device. Sensitive measurements of the magnetic field, obtained by measuring resistance, can be obtained in this manner .This magnetic-field-dependent resistance is known as Magnetoresistance. This concept, in a fundamental sense, is how the newer GMR-based read-heads on high-density hard drives operate.
Another similar device is the Magnetic Tunnel Junction . This is similar to the GMR trilayer, but an insulator film is sandwiched between the ferromagnetic layers, instead of a normal metal. Current can then tunnel through the device, again dependent on whether the spins are parallel or anti-parallel in the ferromagnetic layers. The tunnel junction is the fundamental concept at the core of the MRAM's.
Another exciting area of research with spintronics that I haven't heard anybody on slashdot mention yet is quantum computing. Electrons are spin-1/2 fermions, and hence have two distinct eigenstates of the Spin operator (the eigenstates are usually called "spin-up" and "spin-down"). This makes them perfect candidates for representation of quantum bits (qubits) for potential quantum computation. Some groups are working on this idea, by studying interactions of quantum dots for instance.
Overall, this is a budding field that has already impacted the technology industry in it's scant 13 years of existence. Expect many more interesting and potentially groundbreaking discoveries to occur. But then again, I'm spin-biased. :-)
make world, not war
Sounds a lot like the write-only drive I developed in the 80's while a physics student in college. It used the Pauli Exclusion and Heisenberg Uncertainty principles to store tremendous amount of information. Problem was, I couldn't get the informaton back out.
A guy called after reading our April edition of our user's group newsletter and asked if I had patented the idea and wondered if we could send him a prototype. Shortly thereafter, he received a bakelite case with a SCSI cable coming from it. He hooked it to his Mac. Took the guy three phone calls before he understood it was all a joke.
Can I claim prior art here?
Uptime in weeks?
[tim@cr660477-a tim]$ uptime
5:15pm up 118 days, 19:01, 6 users, load average: 0.00, 0.02, 0.00
[tim@cr660477-a tim]$ cat
Red Hat Linux release 7.1 (Seawolf)
Kernel 2.4.2-2 on an i586
</brag>
I searched for photocanonics as spelt in this thread, but came up with nothing in /. or google.
I think I may have inhaled yer brainfart
We used to call it magnetic core memory.
:)
Or rather, as the maker calls it, FRAM, for Ferroelectric RAM. www.ramtron.com. Only 256k right now, on a 0.5 micron process, but .35 and .25 micron is in the works and they expect the cost/density to become competitive with flash soon.
There will be a spintronics symposium at the upcoming Fall 2001 meeting of the Materials Research Society meeting. It appears that about half of the talks will be about MTJs (the type of structures used in TMR MRAM) and another really big block about these dilute magnetic semiconductors. These are not new per se, but have been strongly revived by spintronics research. Especially important is the development of III-V and Group IV magnetic semis, primarily the result of dilute manganese doping. Really exciting stuff nowadays.
What i've always wondered about quantum computers is how the devil you would measure them.
For those that don't know, nothing is ever finite in quantum mechanics, only probable. (See heisenbergs uncertainty principle).
also by measuring the system by some method you are interfering with the system. And thus how do you know whether the result is the result or a product of your interference.
Now if you were using a lot of electrons on a surface (don't ask me how) all in the same starting spin state, then you performed an operation on them you could then measure the macroscopic properties of your sheet of electrons. IE the bulk magnetisation (see ESR spectroscopy), but single electrons would be impossible to discover with current theory.
The other complication is that the electron spin states are degenerate (have the same energy)in the free state and in the abscence of an external magnetic field.
Thus to measure a change in spin state you would have to stick the quantum device in a very large amgnetic field (5 Tesla or greater). This requires a large super conducting magnetic cooled to the temperature of liquid helium.
Therefore this whole quantum Computer thing is a fairly difficult to achieve and willnot happen over night and probably not until someone comes up with a different theory to current QM theory, which explains in more detail the actions of small particles.
Modern definition of an expert: Someone who comes from far away with a powerpoint presentation.