Hey Chris,
As always, thanks for your input.
Switching gears, you will recall that the Maxtor deal required Ampex to co-finance the design of a new preamplifier to account for the 'more complicated noise' of KM. Below is an article that gives added context to that requirement and gives us an idea of the signal processing challenges that AXC has to meet to commercialize KM.
Improved electronics are but one of the stumbling blocks in this path. The pre-amp and read-channel chips that move data coming off the drive's head are already a limiting factor. Though ICs meet the requirements of today's desktop drives, designers of high-end SCSI-compatible drives have to lower disk density to keep from overwhelming the read channel.
Drive designers rely on new technology mix -
Capacity growth harder to come by
techweb.com
Let me just note again that AXC's current DCRsi 240 already contains 240 Mbits/in2 channel electronics and it is already working on 480 Mbits/in2 channel electronics for its 60 Mbyte/sec DCRsi recorders while the disk drive industry is still wrestling with 150-250 Mbits/in2 channel electronics. For sure, the signal processing issues are different for ruggedized datatape recording and disk drives, but I think it is reasonable to assume that a successful KM program creates opportunities for Ampex to adapts its core competence in signal processing to meet the needs of the disk drive industry.
By the way, I think IBM's yields on its MR and MRX heads are much better than the yields for its GMR heads so we may see IBM continue to make MR/MRX heads all the way up to 5.0 Gbits/in2 while it perfects the processes for its GMR heads, which are going to start at 2.74 Gbits/in2 and should take IBM all the way up to 10 Gbits/in2.
I think Chris makes a great point about how IBM's GMR heads don't use KM. I do think that the annoucement is favorable to the adoption of KM because it sets a more competitive tone for the rest of the industry. One of the reasons why the PC makers were proactively moving the disk drive industry to accelerate the transition to MR was because they knew that by inciting the drive makers to compete more vigorously by leapfrogging each other in density and other innovations, they were setting into motion a process that would ultimately result in cheaper disk drives. Check out this quote from RDRT:
"To set the bar for capacity, we have to get off the 60-percent-per-year rate and get to 70 or 80 percent," said Alan Lowe, president of Read-Rite. "The key to the future is not to develop new products, but to introduce products at high yields."
Assuming 30-40% post-channel density gain, KM allows a density leader like IBM or a density lagger like RDRT, SEG, QNTM, Fujitsu or APM to achieve 78-84% per year density growth instead of the industry average of 60% density growth per year.
The prudent assumption to make is that KM is not the only technology out there that can help do this, but it IS out there and I think, sooner or later, we are all going to benefit from the fact that Ampex has laid a solid academic and research foundation for this technology.
Gus |
