Physicists have stored information for nearly two minutes inside the magnetic spins of atomic nuclei, producing the longest-lasting spintronic device yet and what could be the world's tiniest computer memory.
There's just one problem: the computer operates at -454 F (about 3.2 degrees K) and requires a magnetic field roughly 200,000 times more powerful than Earth's.
Still, it's a major advancement in spintronics, which involves storing data in the magnetic compasses of atomic particles. Usually, data is stored in the spin of electrons, but the memory lifetime is on the timescale of microseconds, according to the study, conducted by researchers in Australia and Utah and published today in the journal Science.
Instead, the researchers stored data in the longer-lived nucleus of an atom. D.R. McCamey and colleagues mapped the electronic spin of electrons orbiting phosphorus atoms that had been added to a silicon semiconductor. The team used near-terahertz electromagnetic waves to give the electrons a specific spin. Then they used FM-range radio waves to "write" the spin onto the phosphorus nuclei, as explained by the University of Utah.
After 112 seconds, the spin was mapped back onto an electron, to be electrically read out. The team used classical binary data (0 and 1) rather than quantum data. But the method could be used in conventional computers as well as quantum computers, according to Christopher Boehme, a University of Utah physics professor and senior author.
Two years ago, a different research group reported storing quantum data for two seconds within atomic nuclei, but they did not read it electronically, Boehme said. The longer time is more than adequate to create spintronic memory for computers, he said.
It's effective because nuclei are less susceptible to temperature changes and interference from other electrons, McCamey said: "Its spin isn't messed with by what's going on in the clouds of electrons around the nucleus. Nuclei experience nearly perfect solitude."
The next step would be figuring out how to accomplish the same task in higher temperatures and weaker magnetic fields, Boehme said.
"There’s just one problem: the computer operates at -454 F (about 3.2 degrees K) and requires a magnetic field roughly 200,000 times more powerful than Earth’s."
That sounds like two problems to me.
I don't know why should be so difficult store information in spins. I mean, physicians and physicists have worked together for at least 25 years optimizing MRI scanners that exploit water proton spins to build image of our body.
Solid state NMR tecniques are routinely used nowadays. In these methods long living nuclear spins constitute a problem that is overcame with sophisticated radiofrequency pulse sequences (this is not a critic, but the electron resonance frequency is far lower than a teraherz, it's about 100 GHz).
A NMR-active nucleus in a solid lattice without such tricks would remain polarized up to 24 hours, due to its slow vibrations.
Perhaps the real challenge will be reduce the huge magnetic fields used to achieve these working conditions and making these devices handy like a USB-pen.
The "author" of this piece, Rebecca Boyle, does not know that KELVIN is NOT measured in DEGREES.
There ARE NO "degrees Kelvin".
KELVINS -ARE- the unit of measurement themselves, equal to one DEGREE Celsius/Centigrade, starting from absolute zero.
So the temperature she mentions should read, "3.2 KELVINS".
@ breakfastman: The milk being poured into your bowl of cereal looks chunky. Kept too long, or too many kelvins?
On topic: So then once robotics and materials science have progressed far enough, we could turn Saturn into a supercomputer?
Here I have a good recommend for you: Tuneup360 and Norton system work, but the latter one is a little bit difficult. For the beginners, Tuneup360 is a good choice.
2 minutes today, a lifetime tomorrow.
Though interesting as hell! I don't see myself ever fully understanding how any of that works... Ridiculous stuff