Magnetic Memory Breakthrough

Racetrack memory mentioned in the text of "Spin torque dynamics with noise in magnetic nanosystems (Phys. Rev. B 82, 144404 (2010))": http://prb.aps.org/abstract/PRB/v82/i14/e144404

Nik_2213 mentions in http://www.physorg.com/news/2010-11-racetrack-memory.html : "Sounds like bubble memory again... Big problem with that was organising data to access it efficiently. FIFO does have a significant latency as block sizes rise..."

I'm betting existing SSD technology though.

Related: http://news.ycombinator.com/item?id=1907174