Slashdot Mirror
'Racetrack' Memory Could Replace Hard Drives?
Galactic_grub writes "An experimental new type of memory that uses nanosecond pulses of electric current to push magnetic regions along a wire could dramatically boost the capacity, speed and reliability of storage devices. Magnetic domains are moved along a wire by pulses of polarized current, and their location is read by fixed sensors arranged along the wire. Previous experiments have been disappointing, but now researchers have found that super-fast pulses of electricity prevent the domains from being obstructed by imperfections in the crystal."
Actually, it's more like Bubble Memory
http://en.wikipedia.org/wiki/Bubble_memory
echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
stores 1 bit per "core." The article is about a form of memory which continually cycles multiple bits stored as magnetic regions through a single physical ring. The OP is correct in that this is similar to cycling photons through an optical ring.
Looking back, this is all very similar to shift register memory, one of the earliest forms of solid state memory.
"National Security is the chief cause of national insecurity." - Celine's First Law
The basic technique is even older than that. Google "Mercury Delay Line" for early examples: they'd make a long thin tube of mercury with transponders at each ender. It was around 5 ft per K, IIRC.
- cheap
- reliable - OK, hard drive errors do exist but I wish my car, for example, was as reliable
- standardized - OK, there are a number of standards but not that many
Yes, in the long term I don't see the hard drive as the best method of data storage but the altenatives have a long way to go before they replace it.init 11 - for when you need that edge.
Mercury delay lines were the cause of a bizarre
computer architecture. The normal form of instructions
had an "address of next instruction" field.
After getting the program to "work", i.e get the correct
answer, the "optimization" stage consisted of working out how
long each instruction would take, and then positioning the "logically next"
instruction at the location just about to appear out of the delay line.
There was no advantage to inner loops that were faster than the
delay round the mercury loop. Unless you could unroll and fit two
repetitions into one trip round.
Of course, all of this was done by hand.
Generally known as n + 1 addressing, where n was how many operands had addresses in the instruction. Also used with drum memory, which was in the physical shape of a cylinder ion the one drum machine I used, but was mainly a head per track disk, so no seeking required. Some drums had multiple heads per track for some tracks to reduce latency further.
The optimization was great fun, my favorite part. You could make programs scream if you paid attention.
Infuriate left and right