Paul,
I agree that Ampex's current core DST business is technically very competitive and the fact that Ampex is ramping up its marketing efforts should translate into increased penetration of its markets. It is worth repeating, however, that these products are new, less than 4 years old (even less if you factor in the constant refinements that AXC is scaling in), and that it was only after Ampex completed its restructuring in early 1996 AND only after Ampex improved its working capital position with the sale of its campus, was it able to devote more resources toward increasing sales. From 1995 to 1996, DST storage products doubled!
My understanding is that Ampex's tape libraries are scalable. As a matter of fact, part of its 1992 development deal with E-Mass, a unit of E-Systems - a secretive electronics systems defense contractor that is now part of Raytheon, involved an OEM agreement that ended in the first half of 1994. One of the things that E-Mass did was to use the ER-90 (E-Mass branded first generation DST) tape drive as the building block for a massive 300 terabyte robotic library for the DOD in 1994. IBM and StorageTek dominate this area, though. For example, the National Security Agency's (big brother) storage requirements are met, in part, by massive tape libraries from StorageTek, IBM, and Convex(HWP). Still, Ampex got its robotics expertise from Odetics (ATL) and some firmware technology from E-Mass so it is conceivable that Ampex, with some major advancement in tape drive technology and with a larger cash flow, could match its leading-edge tape drives with ever-improving automation technology (hardware and software) to eventually compete in these markets. It wasn't that too long ago that a brash startup named StorageTek wrestled dominance of this market away from a giant named IBM.
When Ampex shifted to a direct sales marketing model in 1994, it probably decided that it simply did not have the distribution and service network to support huge installations like those so they went for the niche markets like the oil/gas exploration folks who were increasingly taking their computers and storage systems out to deeper and deeper waters to process the increasingly complex seismic images used in drilling and exploration. Universities and other supercomputing environments also proved to be fertile ground for its early generation DST libraries. Now they are targeting the Fortune 100 companies and the broadcasters.
The thing about these niche markets, however, is that these are low-volume high margin type of markets with only so many sales units available so expansion necessarily has to come from entering more and more of these niches. Fortunately for Ampex the macro trends are creating new markets that play into their strenghts. Ampex, because it shares a high degree of common components across its tape-based product line, is also now in a very good position to react quickly to the way these trends are shaping potential markets.
Take the telecommunications products for one . As I understand it now, Ampex's latest line of instrumentation recorders/libraries are being used by customers like Sprint, MCI and ATT which are intimately involved in the construction of the fiber-optic backbone of the data and voice networks. Below is a link to a 2/3/97 article in the LA Times that provides a PDF map and an overview of the ongoing expansion of the net backbone. Below is an excerpt that probably describes the area where AXC's unique DIS products fill a need:
"...Because the backbone's highest-capacity fiber lines, or "pipes," can handle more traffic than the fastest router, the routers - electronic post offices that sort e-mail and other data flowing through the Internet--can become bottlenecks. But the smaller pipes can also get clogged when too much information is pumped through them. The backbone companies are continually monitoring their systems, trying to figure ways to shift traffic away from pressure points and avoid having information "packets" needlessly bounced like pinballs from router to router, causing delays and lost data...."
latimes.com
I was actually pleasantly surprised that, without much fanfare, AXC was able to quickly reconfigure its tape drives to handle the high data rates of fiber optic technologies and put together a robotic library as well.
Telemedicine is another potential market for DST products. Transmission and Compression standards are still being formulated, but we alreadly have some idea of the shape and form of this market, definitely a huge potential market. From the Georgetown Medical Center, DOD, Lockheed, and StorageTek trials going on we have some indications of the datasets involved:
Typical 400-bed hospital has an existing library of 30-50 terabytes of radiology and mammography images alone.
Typical requirements of x-ray images only = 5 terabyte annually
Typical requirements of mammography only = 10 terabyte annually
Typical size of 1 x-ray image = 8-10 MB
Typical size of 1 set of 4 mammography images = 32-40 MB
These are emerging image-intensive applications for which Ampex's DST products are ideally suited for. Ampex has made trememdous progress in improving its ability to participate in new markets created by the trend towards multimedia databases, satellite communications (120 commercial satellites launched in the last 10 years, 1200 commerical satellites to be launched in the next 10 years!) and others that involve the storage of huge datasets, particularly images. It is useful to remember that the continued increase in computing power necessarily leads to increases in our ability to analyze even larger sets of data.
The traditional achilles heel of the helical scan format, however, is the relatively high cost of maintenance, primarily due to the heavy wear and tear of the heads and media. The hottest tape format now is DLT or Digital Linear Tape, a 1/2 inch mid-range serpentine recording format that involves the use of multiple stationary recording heads that have a MTBF (mean time between failure) of 30,000 hours. Ampex's helical scan recording heads typically have a MTBF of between 2,000-3,000 hours. Improvements are being made all the time, though. Also, Ampex has patent #5,585,977 by B.Gooch that synthesizes a large part of Ampex's recent head/media innovations into a new and revolutionary (? - you decide) advanced recording scheme that may remove this technical and marketing weak point. I do not know if this is the basis for a recording platform that AXC can use to participate in the higher volume compact tape market. Those of you who have watched Iomega struggle to support a SCSI product being sold to the consumer market are aware of the peculiar difficulties and costs inherent in that task.
Someone once wisely commented that a lot of people are going to go for the quick 200-300% pop once the KM contracts start to come in and are going to miss out on the truly dramatic upside that this company's technologies present IF it can execute its plan. Might be worth thinking about it while we wait for more contracts to come in.
Gus
P.S. At tek.com Tektronix has a white paper that describes the various datatape recording schemes (helical scan vs serpentine recording) and directly compares AXC's DST vs IBM's 3590. |
