New IBM storage method - "millipede"
<a href="http://apnews.excite.com/article/20020611/D7K2NC281.html"; target="_blank">http://apnews.excite.com/article/20020611/D7K2NC281.html</a>;
Cool. Sort of. IBM doesn't look like it's going to do it though - here's why.
You need a whole bunch of devices that can read and manipulate those tiny bumps. Say each arm is a couple hundred microns long, then each can access a circle 400 microns in diameter. There's a whole lot of wasted space on the polymer, even if the cirlces touch edges.
The wasted plastic is no problem - but how can you make millions of tiny arms all work correctly? And make them cheaply? As far as I know, STMs are rather expensive.
I think I see how the idea works though. You have a sheet of flexible polymer and a reading arm. The arm can navigate around to many different places on the sheet. the arm has three settings when it's over a "dot" on the sheet - press lightly, press hard, and press hard then pull up slowly. 0 = flat with the rest of the sheet. 1 = pressed down. The press lightly setting reads the dot without changing its state from 0 to 1 orr otherrwise. It pulls back fast so that the polymer can't stick to it. The press hard changes the dot from a 0 to a 1. It pulls back fast so the polymer can't stick to it. The press hard-pull back turns a 1 to a 0.
I have no idea how fast these arms can move.
I don't know how the densities of this sheet can compare to, say, an SDRAM chip.
I also think it would generate a good deal of heat.
And it's still a 2-D storage medium! Grr. I want holography!
Cool. Sort of. IBM doesn't look like it's going to do it though - here's why.
You need a whole bunch of devices that can read and manipulate those tiny bumps. Say each arm is a couple hundred microns long, then each can access a circle 400 microns in diameter. There's a whole lot of wasted space on the polymer, even if the cirlces touch edges.
The wasted plastic is no problem - but how can you make millions of tiny arms all work correctly? And make them cheaply? As far as I know, STMs are rather expensive.
I think I see how the idea works though. You have a sheet of flexible polymer and a reading arm. The arm can navigate around to many different places on the sheet. the arm has three settings when it's over a "dot" on the sheet - press lightly, press hard, and press hard then pull up slowly. 0 = flat with the rest of the sheet. 1 = pressed down. The press lightly setting reads the dot without changing its state from 0 to 1 orr otherrwise. It pulls back fast so that the polymer can't stick to it. The press hard changes the dot from a 0 to a 1. It pulls back fast so the polymer can't stick to it. The press hard-pull back turns a 1 to a 0.
I have no idea how fast these arms can move.
I don't know how the densities of this sheet can compare to, say, an SDRAM chip.
I also think it would generate a good deal of heat.
And it's still a 2-D storage medium! Grr. I want holography!
Comments
I got the idea wrong. Still pretty cool.
<dang! can't ge the thing to link properly without needing to register!>
<a href="http://www.nytimes.com/"; target="_blank">http://www.nytimes.com/</a>;
<img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />
<strong>
And it's still a 2-D storage medium! Grr. I want holography!</strong><hr></blockquote>
If you could change over to wanting something hallucinatory you're more likely to get what you want sometime soon.
Here's a batter apxlantion.