Zibb

Paul RakoTechnical Editor Paul Rako looks at analog technology in power supplies, interface, the signal path, and life in general.



   Advertisement

Profile

RSS Feed

  • Add this blog to your RSS newsreader!

Recent Posts

Recent Comments

Most Commented On

Archives

By Category

Analog Design Articles

Blog

Friday, April 11, 2008

IBM makes new bubble memory

Apr 11 2008 12:14AM | Permalink |Comments (5) |


IBM has a new non-volatile memory called racetrack memory. To those of us old enough, it sounds just like bubble memory, which they used back in the 1980s. The term I remember about bubble memory was chevrons. That was the little angled electrodes that the “bubbles” of magnetic domains jumped across. This new racetrack memory seems to be the same thing only they use nanowires to transport and retain the bubbles of magnetic domains,. The domains are much smaller than they were with he chevron deal so I am sure the new racetrack memory is more power efficient. The good thing about both types of magnetic memory is that they are non-volatile. The bad thing about both types is that reading the memory erases it so you have to constantly re-write it, just like in the old ferrite core memory days.

I am sure the technical challenges of adapting bubble memory to nanowires was formidable so this really does qualify as a new invention. I just think it is neat to see that just like disk drives are a derivative of the phonograph, racetrack memory is a derivative of the old wire recorders from the 1930s. Plus ça change, plus c'est la même chose


Related entries in: Analog | 


Reader Comments



at 4/11/2008 2:46:26 PM, Larry M said:
At IBM in 1970, we had a 700 bit "Harmonic Memory" in the 5496 Data Recorder, a card punch with memory that punched 96-column cards. It was a stiff wire coiled into a 5-turn spiral with about 7 inch outside diameter. One transducer introduced bits by vibrating one end of the coil. Another collected each serial bit at the other end and changed or re-introduced it. Sort of an early combination with aspects of both of serial EEPROM and DRAM. ISTR it held around 700 bits, with a read time in milliseconds.



at 4/11/2008 10:55:57 PM, D Vanditmars said:
Most of the new FRAM and other magnetic IC memories also have a destructive read cycle. So not sure how 'bad' this would be...



at 4/12/2008 9:19:10 AM, Not Analogous said:
A disk drive is not a derivative of a phonograph, and the racetrack memory is not derivative of a recording wire.

The analogy is completely inaccurate in both cases. It might sound good, but it isn't correct at all.



at 4/22/2008 3:57:09 PM, Ken Folse said:
Who made wire recorders in the 30s? I remember a Westinghouse (I think) magnetic tape recorder from 1943 or so, but it used metal tape and had a capacity of one minute. The first consumer wire recorder in the U.S. was from Webster.




at 4/22/2008 7:09:36 PM, john Walter said:
I had a wire recorder made by the Majestic Radio company. One half hour of material could be recorded on the stainless steel wire which wrapped around the 78 rpm phono turntable. Material could be recorded from the am radio, phono, or crystal microphone.

Post a comment



Display Name

Change Image
Before submitting this form, please type the characters displayed above.
Note the letters are NOT case sensitive.


ADVERTISEMENT

©1997-2009 Reed Business Information, a division of Reed Elsevier Inc. All rights reserved.
Use of this Web site is subject to its Terms of Use | Privacy Policy

Please visit these other Reed Business sites