A harddisk development issue

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View

Hi there.

I have a pretty odd question about harddisks development and production.

First of all I apologize if this problem doesn't belong to this ng.
Then, pls, point me to the right ng.

Well. My friend argues that the development of a new HD is not
precalculated and projected. Therefore the result is not predictable.
There's produced "something". Then thoroughly tested and measured its
capacity and other properties. And even after this procedure the
technical data are declared on a new product. Moreover she told me that
this is a reason why there are invested billions of dollars to testing
machines and systems.

What do you mean about this? I can't believe in that!

Thanks to all for your judgements.
Pea C
NEVER LIE, but stretch the truth as far as possible!

Re: A harddisk development issue

Pea C wrote:
Quoted text here. Click to load it

I would say that she is more right than wrong.

What you have to understand is that given current hard disk densities and
performance you will always be able to achieve better performance on
carefully built prototypes in a lab environment than you will with
production samples in some guys bedroom.

So what I believe happens is that they have a pretty reasonable idea of the
performance and capacity from computer modelling and past experience. They
will then test pre production samples to find a reasonable window of
reliable performance, taking into account things such as failure
tolerances, percentage yield of units able to function to desired
specification, cost to improve production methods to achieve higher yield
at desired performance vs added value in product achieved from that
performance, etc..etc

Then they will perhaps form some sort of re-grading system (this is done on
Processors and other semiconductors) where a unit that fails QA at one
performance level may pass at a lower capacity/speed performance window.

Hard drives differ from semiconductors in that performance is pretty much
down to how closely you dare pack the bits together on the disk surface
before getting too many read errors, how fast you dare spin that surface
and still be sure you can read/write it in a reliable manner and how fast
you dare move the heads around when doing so. Pretty much all of these
things are set by the drive controller firmware, but that in turn was
programmed to perform in a optimal manner for the average performance of
mechanics it is attached to.

They probably could have set your drive to be a bit faster or higher
capacity, but in doing so they would have either reduced the chances of the
thing working at the end of the production line or reduced it's mean time
before failure beyond limits they thought acceptable.

Re: A harddisk development issue

Pea C wrote:
Quoted text here. Click to load it

It's just speculation on my part, but I would guess it's true that the
specifications for a new harddrive has been determined empirically from
tests of the hardware.  That does not mean, however, that the engineers
don't have a good idea what those numbers will be before the testing and
refinement stage.

For instance, the capacity of the drive is a consequence of the geometry
that it employs to store the data.  That would have to be designed in.
It is possible, though, that the design would have to be tweaked as a
consequence of problems uncovered by testing.

Re: A harddisk development issue

Pea C wrote:
Quoted text here. Click to load it

With few exceptions, hard drive technology has not advanced greatly
since 2000. Two hurdles are areal(bit) density and rotational speed. The
  SATA interface does not increase transfer rates; it allows for 50%
more bandwidth than ATA-100, when either one or both of the former are
increased. Engineers know all of the electrical and mechanical
properties before the prototype is built.

Re: A harddisk development issue

Quoted text here. Click to load it

They do know the exact structure of the platter, its composition is checked
with a 3D ATOM PROBE which measures both content and structure in ATOMS.

Technical director CKCCOMPUSCRIPT
Apple Computers, Intel, Roland audio, ATI, Microsoft, Sun Solaris, Cisco and
Silicone Graphics.
Wholesale distributor and specialist audio visual computers and servers

Re: A harddisk development issue

Chris wrote:
Quoted text here. Click to load it

Thanks to all for the contributions.

Though, I think the research of the  new materials MUST produce some
relevant results. The results on which the developers of HDs can build
their new products. If Marie Curie was able to crash atoms in primitive
conditions, contemporain scientists have to know how to stick them for
new materials.

Pea C
NEVER LIE, but stretch the truth as far as possible!

Site Timeline