Thanks for the perspective, Steve.
It would be great if AXC could forge another OEM deal again with a global player like IBM because that would allow them to piggyback on a well-established global distribution and support network while it is still building out its own network, a modest one considering the modest resources available. Petco, for example, spent 1 1/2 years looking for the right tape drive until it discovered DST. One can only wonder how many Petcos are out there kicking the tires of the numerous tape drive solutions without being aware of DST.
Anyway, I know you used to be an engineer and I was wondering if you could give us your insights into the IDEMA test results.
Remember the David Davies presentation at that 7/8/97 IDEMA symposium in Colorado? Well, true to form, Ampex used that technical forum to disclose the progress of their work with KM+MR up to that point.
Here are the test results that Ampex disclosed at that forum:
1) MR Head Overlayer Keepered Media Test Results
a) SAL MR heads can saturate keeper layers with reasonable bias currents
b) Effective keepered media track width is narrower than with unkeepered media
c) Keeper Layer shields pick-up of signals from adjacent tracks
d) Signal from keepered media decreases with keeper thickness
e)Reduced signal levels make it impractical
2) MR Head Underlayer Keepered Media Test Results
a) Underlayer keepered media has flatter frequency response than unkeepered media and shows high density gains
b) Underlayer keeper layer shields pick-up signals from adjacent tracks
c) Underlayer keepered media has better pulse shape for PRML channel
d) Noise is somewhat reduced with an underlayer keeper
There are also slides that show the thermal decay protection that KM provides.
As I understand it, the typical components of a disk drive platter are as follows, with the overcoat being the platter component closest to the read/write head:
1) substrate (glass or aluminum)
2) magnetic layer (typically multiple thin magnetic films)
3) overcoat (typically carbon)
Overlayer keepered media refers to:
1) substrate
2) magnetic layer
3) keeper layer (soft magnetic alloy)
4) overcoat
Underlayer keepered media refers to:
1) substrate
2) keeper layer
3) magnetic layer
4) overcoat
I think that AXC has made significant progress since that 7/8/97 presentation. Using the KM+TFI road map as a guide, we know that when Ampex starts to work on the channel electronics (PRML chip, preamplifier, etc), they are close to completing their work for a particular head design For example, the deal with Maxtor called for the co-funding of the design of the preamplifier required to work with the PRML chip that Maxtor was contemplating on using with KM.
Again, what makes this most interesting is the addition of Roger Wood of IBM, a signal processing heavyweight who also happens to be the co-inventor of Patent No. 5,041,922 - the KM patent - when he was at Ampex. These are the groups that we know which have worked on KM recently:
1) Stanford (TRMC 97) - nonlinear equalization
2) IBM (PRMC 97) - Roger Wood specializes in signal processing (PRML, EPRML, etc)
3) University of Plymouth, UK (PRMC 97) - dual-stripe MR heads/permanent magnets
4) University of Minnesota - Thermal decay study continues. Incidentally, J. Chen, whose thesis is "Thermal Decay and Keepered Media," completes his Masters degree this month. He would be a valuable addition to the KM crew, which looks like it is also looking to add a recording channel engineer.
One can argue that IBM may only be interested in helping out the first major OEM MR head customer in using KM. But, if KM works with IBM head designs and it also works, as Samsung required, with MR heads from specified manufacturers, there is no reason for IBM not to use KM especially since they are hellbent on toppling SEG from the top of the disk drive pile. IBM currently has the lead in the density race with its 2.65 Gbits/in2 MR heads (Travelstar). KM turns that from something like 3.1 Gbytes/3.5" platter to something 4.0-4.3 Gbytes/3.5" platter, assuming that AXC is able to generate 30-40% post-channel density gains.
KM+MR is still an advanced work-in-progress, but, I think it is telling that at this stage of the game, Ampex is already starting to work on the channel electronics and on dual-stripe MR heads. I think it is reasonable to assume that their work with a variety of single stripe MR heads (SAL MR heads, merged SAL MR heads, etc) is close to completion with no fatal hitch so far.
Thanks for your thoughts.
Regards,
Gus
P.S. By the way, am I correct in assuming that there are 3 basic ways to introduce the bias current required to create the aperture in the keeper layer during the read and write processes?
1) By adding another wire (KM+TFI uses 3 wires)
2) By using the biasing scheme inherent in the MR design (SAL MR)
3) By using permanent magnets (dual stripe heads??) |
