h a l f b a k e r yMy hatstand runneth over
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Given that heads are thin film technology, and that control electronics gets cheaper every year; at some point it will become cheaper to make a hard drive with no head movement servo at all; but one stationary head per track instead. WOuld increase read and write speeds too in proportion to th3e number
of heads. Same idea works for cd-rom and dvd of course.
Fixed-head mainframe disk, circa 1975
http://www.gizmodo....hnically-018574.php
"Each track therefore had its own head, so the seek time was zero - you just electronically switched heads." [krelnik, Oct 17 2004]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
Annotation:
|
| |
Damn! I was hoping stationary had been misspelled and this was about a paper headband. |
|
| |
Actually, this is the way disk drives were made originally, back in the days when the disk platters were half a meter across and the drive cost $100,000. See link for a picture. |
|
| |
I think track spacing is too tight. The
other problem would be data rate - a
single head can read up to 100MB/s
with 15K Atlas drives. That would be
approximate 50GB/s (given that there
are about 1000 tracks on a 3.5" platter). |
|
| |
Far bigger savings would be in seak
times, which would be reduced to
rotational latency. Unfortunately, I think
high capacity flash drives will reign
supreme as there's no latency at all. |
|
| |
I imagine the heads distributed over the surface in a Fibbonaci spiral like the seeds in a sunflower. |
|
| |
Read and write speeds would not increase in _direct_ proportion to the number of heads; there are overheads in parallel data processing and transmission meaning the speed increase is less than you might expect. |
|
| |