Has HD innovation stagnated?

[aka1nas]

Originally posted by: Pariah
A few years back, Conner designed a drive with 2 complete head assemblies which I believe was called a Chinook, Seagate later bought them out an may have been the one to release it. Either way, it obviously didn't catch on as the design was incredibly complex and expensive. This multihead design is not in any way related to how the optical True-X technology worked. The True-X method would be pretty worthless for hard drives since simple disc fragmentation would ruin any potential benefits, something that never occurs on optical discs which are ideally suited for the True-X tech because of the almost exclusive straight streamed reads for things like installations and media files.

The bar of read heads would not work since you would either need a head for every track which is obviously not practical, or the bar would have to be able to shift which would be a calibration nightmare for multiple heads on a fixed bar. Another problem is how close the read head has to be the surface of the platter to work, literally floating on a few molecules of air preventing it from contacting the platter surface. It would be quite the engineering feat to try and build a static bar that close to the surface of the drive. It would also mean that whenever the drive is moved even when not in use it would be extremely susceptible to the platters getting scratched by the read bar which isn't possible with the current tech as the read/write head is parked off the surface of the platter when powered off. Also, even the slightest wobble in the platter while on and spinning would potentially contact the platter at some point.

Maybe you could not have the platters spin, but keep them mounted securely and instead spin an assembly with multiple read/write heads, say maybe 4, over the platters. With the right command logic, your access times should on average be 1/4th of a single head design. Also, it might be easier to have a motor only spinning the read/write assembly than the entire platters. Does this seem logical at all? Has it been tried before?

 

[Gannon]

I think the problem with spinning heads is vibration which would make it very prone to read errors, I am sure that is why actuator arms are mounted firmly in place, but I am just guessing here.