Ok, first post, here is an idea, might be tottaly stupid, but whatever.
I dont know how much a head (not the actuatror/arm) costs or how small it can be made, and I dont know how many tracks a 500/750gb hdd has, BUT what about heads as many as the tracks “mounted” in the “floor” and the “ceiling” of the hdd?
In a few words, a head for each track of the platter.
here is a qiuck pick
Lets just say that the platter has 28 tracks, and each head has a radius that “shadows” 3 thracks. Of course the center of each head (red) can read one track). So we place the first line of heads (the horrizontal) to read the tracks 1,4,7,10 etc. To cover the missing tracks shift right by a track a duplicate line of heads (remove last head) and rotate them from the center of the platter as little as possible to gain more space. Here the second line of heads are 15 deg rotated. We do the same for the 3rd line.
Its a very simple idea and a very simple paradigm. It shows each heads possition in the floor/ceiling, but I dont know if its possible that the whole floor/ceiling is made from common head parts in order to reduse the spacing of the readers/writters.
Hope you understand the idea of mine. As you can see now there is a head for each track. With this we might be able to know about the usage of any track, like how much free sectors it has,
Any thoughts for this are welcome!
Edit: If its possible, that means no moving parts= no wear= no heat, parallel read/write = top speed, smaller in height hdds= more disks in RAID.
The idea isn’t stupid at all. I also came to the same idea a few months ago. Then I conducted a patent search - many guys through the last 25 years saw the distinct advantages of SHD - static head disk (as opposed to SSD and HDD abbreviations). You counted the advantages right, while there are even more: increased MTBF, less noise, reliability, shock resistance etc.
To accommodate such large amount of static heads, the heads should be small - in order of few microns. Thus, VLSI (very large scale integration) techniques should be employed for manufacturing. And here comes the biggest challenge - how to produce viable (capable of read and write) head in such dimensions.
I personally have a vast experience in VLSI and academic research - and I tell you this isn’t simple at all. I think (again it’s my personal belief) that it’s possible by combining advanced VLSI with advanced signal processing (and this will be the real patentable know-how). Consequently, to provide a meaningful answer re applicability of this approach to the modern drive capacities, a serious research effort (probably 2-3 years) to be furnished. Such research would concise multiphysics simulation, numerous prototypes and extensive experimenting. A tight connection to a big HDD manufacturer, like WD is also mandatory for both technical and budget issues.
Summarizing, your question (IMHO) should be actually addressed to WD, not to forum participants:
“Hey, WD guys, did you investigate seriously the possibility to design SHD (static head disk)? If yes, please, post your results (we understand that since SHD isn’t on the market, the decision was negative. But why? - technical? cost? other?). If not, but you’d like to - contact me by email and we’ll discuss the aspects of high quality research of this topic.”
Thanx for your support in this! I posted here, cause I thought that wd staff would check it out. Pm me your email.
As for the small heads, each head consist of many smaller parts, so if one or some of those parts can be placed in bigger size (perhaps the whole area that has all the stationary heads) in order to be common for all the heads, then perhaps it would solve the problem.
I have seen some drives that had a mix of fixed-head and moving-head architectures. Moving-head drives are cheaper per byte but are slower. Fixed-head drives are very good for swapfiles in multiuser OS’s. The ones I remember had a large moving head that did not cover the outside inch or so of the disk. (This platter was about a foot across.) At the edge there were some fixed heads that the OS would use for swap files. It gave a fairly inexpensive way to provide some fast access disk along with the regular moving-head storage.
This goes back to the adage about disk drives: Big, Fast, Cheap: you get to pick two.