Introduction by Jim Koch -
Our guest today is David McKendrick from
the storage industry center at UCSD. David’s interests
are in the effects of geographic dispersion, the advantage of
location in terms of competitive advantage as well as other aspects
of this industry. David received his doctorate from the University
of California, Berkeley where he’s taught as well as at the
University of Texas. David it is a pleasure to have you with us
here today.
DAVID MCKENDRICK
 |
The only thing worse then following
Al Shugart is being number 6 to talk about hard disk drives and
the hard disk drive industry. I have only been studying the storage industry for about 2 years but I might be able to bring to it a slightly different perspective on the evolution of the industry than what you’ve heard today. The story I want to tell is really three-fold. First is the rapid change in the industry, which I will go over very briefly. Second, the disk drive industry is characterized by intensive competition. IBM is the dominant factor in the history of the disk drive industry but some of the great benefits to consumers were brought by new firms entering the business that thought that they could compete against IBM and gain some market share. And the third aspect of the evolution of the disk drive industry, which perhaps anticipates the movement of other industries in the United States, has been the extensive globalization of assembly. |
 |
We’ve heard about this but I thought I would list on this SLIDE some of IBM’s technological first’s that been so influential and dramatic in the history of this industry. |
 |
But there are other “firsts” in the industry that are not necessarily technological but reflect the nature of this industry and that is taking existing technology and being first-to-market to take advantage of certain market opportunities. Memorex introduced the first voice coil actuator in 1968. NCR came out with the first thin film disk in a disk drive. Control Data bought NCR’s business in 1972 and the thin film disk effort died. Control Data never made any effective use of that thin film technology. Seagate Technology in 1980 came out with the first 5.25-inch disk drive for the desktop, followed by the first 3.5-inch inch disk drive by Rodime, a Scottish firm, and one of the few European successes in the history of this industry. PrairieTek introduced the first 2.5-inch disk drive. Conner Peripherals, again as a market opportunity rather than a fundamental innovation, introduced the first 1-inch high 3.5-inch disk drive. This feat was in no small measure related to their other first; Connor Peripherals was the fastest company in history to reach one billion dollars in sales. And then, finally, Integral Peripherals was the first company to introduce sub-2-inch disk drive. |
 |
Now, entries in competition. IBM, of course, but there were Bryant Computer products and Data Products, which are the first 3 companies to ship disk drives. A second generation of entrants were associated with the plug compatible industry. Once IBM established dominance in the industry with the technology, after a two-year lag, you found a lot of firms coming in by reverse engineering IBM products, improving upon different functionalities of those products and undercutting IBM in pricing. And then the third expansion, the take-off period with disk drives for the desktop computing market. |
|
This is going to be geographic
story in a sense. This picture shows what the disk drive industry
looked like in 1962. There were 3 companies: IBM, which
shipped in 1956; Bryant Computer Products which first shipped
in 1960, and then a company called Telex whose drive activities
were later spun off through a management buy-out that became
Data Products. Telex was originally in Minnesota but moved
all the activity to Culver City, California, and by 1962, two
out of three companies in the disk drive industry were located
on the West Coast. 8 years later there was considerable dispersion of economic and technical activity in the disk drive industry. Again, not a lot of innovativeness, but a lot of effort to capitalize on the technology created by IBM. This was really the birth of the plug compatible industry to make disk pack drives, but also some effort by other computer systems manufacturers to make their own drives, so-called “captive” drives. In 1970 you still had IBM and very successful companies such as Memorex, Information Storage Systems, Burroughs, Century Data Systems and Control Data, which was just beginning to be very successful as an OEM and plug compatible maker in Minnesota. There was a smattering of activity in New York with Potter Instruments, formerly a tape drive manufacturer. So there were some new entries that really generated strong competition in this industry. 10 years later, however, not as much activity on the East Coast as a decade earlier. Really just Digital Equipment Corporation in Massachusetts and Vermont Research, a small disk drive company. Most of the activity continued to be on the West Coast. |
|
Now, West Coast firms were disproportionately
successful, if measured by revenue and market share. Although
there was a lot of dispersion in economic activity, most of these
new companies did not capture much business. What you had were
the same 3 firms as in 1962 (this is kind of a no brainier);
in 1970, 4 out of 5 were on the West Coast -- Burroughs, Memorex,
ISS and, of course, IBM. Control Data, possibly the most successful
OEM company in the disk drive industry for a good 10 to 15 years,
was in Minnesota. Until 1972, DEC bought its disk drives from a small company in Arizona and Diablo Systems of Hayward, California, for the DEC minicomputer. After it began making its own drives, DEC quickly became one of the five largest disk drive companies in the United States. So for the first 25 years, basically, roughly 30 companies entered the industry. IBM still dominated with probably close to 75 percent of industry revenues in disk drives. But something happened, a sort of democratization of the disk drive industry. That is, the beginning of the decoupling between the computer system and the disk drive to the extent that computer systems manufacturers like IBM and DEC began to see their lead in disk drives erode. Then new start-ups entered with small form factor disk drives to supply new PC companies like Compaq, Dell and so forth. |
 |
These next SLIDEs show these changes.
I’ve neglected the Japanese here today as you see that most
of the company activity and new entries into the industry really
were generated by United States firms. Western Europe had a smattering
of computers systems manufacturers that made their own desk drives,
such as Phillips and Honeywell-Bull. There were a few East European
companies that made disk drive to supply computer system manufacturers
in the Soviet Block. With Brazilian Informatics policies
you began to see almost a dozen disk drive companies emerge there.
And surprisingly, very few companies in Asia, outside of Japan,
entered the industry; unlike other parts of the value chain in
the computer industry, mother boards, floppy disk drives and
monitors, this industry has really been dominated by American
firms. The first wave of companies into the disk drive industry were the plug compatible makers. The second wave is associated with introduction of the PC and new entry by many Silicon Valley start-ups to service this growing market. By 1985, many of those firms that I previously listed began to enter the industry to make 5.25-inch disk drive, then 3.5-inch inch disk drives. Most of the spin-offs were from IBM in Northern California. Burroughs and Pertec in Southern California also made that area a very successful center for magnetic storage. By 1990 more winnowing out, with greater concentration of firm activity on the West Coast. In 1995, 80% of the companies in the industry are on the West Coast. |
 |
Now the West Coast firms again were more
successful and ultimately became more dominant as the pictures
show. So again by revenue (this one circle up here in Idaho
was Hewlett Packard’s Boise operation) 3 out of 5 companies
are really California-based. By 1990, 4 out of the top 5 companies are California firms. By 1995 Western Digital entered the top 5 and at this time all 5 of the top companies in hard disk drives in the United States are on the West Coast. |
 |
Why would that be? One is the cost
advantage associated with meeting rapid products life cycles
and ramping up to volume manufacturing. But there is also
an externality associated with being in a region of new start-ups
that generated strong competition for the incumbent firms.
Most of these start-ups did not survive, but they moved the industry
forward technologically. One of the great benefits that consumers
had from progress in the drive industry resulted from the risks
taken by new entrants. In the 5.25-form factor Seagate did survive
and make the transition. Tandon, International Memories,
Computer Memories, Rotating Memories, were among the leaders
in pioneering the 5.25-inch form factor, but they disappeared.
In the 3.5-inch inch form factor, Microcomputer Memories, Microscience
International, La Pine Technology, Tandon. And in 2.5-inch drives,
Connor Peripherals and Areal Technology all disappeared.
IBM is now the only company pioneering a new form factor (the
one inch drive). Now, in 5.25-inch disk drives about 7 of the first 10 companies where United States firms. Of those, about 80% were West Coast firms. So part of the success of the West Coast is the innovation generated by these start-ups. These West Coast firms put pressure on incumbent firms in these new technologies. For 3.5-inch drives, about 80% of the companies were West Coast firms and the same held for the 2.5-inch inch drive. |
|
The new wave of desktop computing introduced
a new competitive logic in hard disk drives. Firms not only had
to introduce new products, but they had to also outproduce their
competitors. The new desktop drives had to reach volumes that
were never needed to serve the mainframe systems. Back in 1970
volume output by a plug compatible manufacturer was about 500
disk drives per month. A year ago Quantum had a capacity
of about a million and a half drives per month. So we have enormous
economies of scale and greater attention to cost of assembly
because the large purchasers of disk drives are are seeking all
the cost advantages they can get. That triggered a search for locations with the lowest operating cost. Operating cost of labor, tax breaks, fast regulatory approvals, and infrastructure quality determined where manufacturing was done. The 5.25-inch disk drive, I think, peaked at about 8-9 million drives a year. Any form factor larger than that never shipped in excess of a million drives in any year, and these larger drives are barely visible on this chart. You can see from the 3.5-inch disk drive how fundamentally different the economics of that disk drive were. The small drives as you can see were really the future of the industry. So what happened? In 1981 where were companies who were making these desk top disk drives? Well 100% of their drives in 1981 were being assembled in the United States. In 1983 Seagate began to assemble drives in Singapore. Following the apparent success of Seagate, others copied that model. Computer Memories and Tandon followed. Singapore had this unusual combination of good infrastructure, positive regulatory environment, and decent wage rates, which dropped the direct labor share in the cost of a disk drive from about 25% to less than 5%. |
 |
By 1990 you began to see a dispersion from Singapore
to new areas in East Asia, places such as Malaysia, and Thailand. The United States continued to lose share of production. By 1990, more than 66% of all United States disk drive units were produced in South East Asia. So volume production in the United States was quite low, with most in Singapore by 1990. Today essentially all disk drive assembly is located in Southeast Asia. |
|
Now this movement to Southeast Asia, relative
to other competitors, primarily the Japanese, including Fujitsu,
Hitachi and NEC, gave US firms a cost advantage. There were stories
in the late 1980s that Seagate was exporting drives from Singapore into Japan. The Japanese were much slower to move their assembly to South East Asia, relative to the American firms. That is what this chart shows: the percent of all firms with assembly facilities in Southeast Asia. The United States firms went there sooner and that enabled them to attain an advantage relative to Japanese competition. Not just in technology, because if you look at any of the Disk/Trend Reports and measure areal density in any of these products there’s a leap-frogging affect. Technologically, Fujitsu, Hitachi and IBM were equals in terms of areal density, from one product generation to the next. It was not just technology that preserved the advantage of US firms; it was their ability to ramp up volume manufacturing in low cost locations, manage that international transfer process for product development, and develop an infrastructure that was a very powerful weapon in competition. By 1995 Japan came on very quickly to relocate assembly to the Philippines and match that low cost advantage that the American firms had. Today there is a much more level playing field regarding cost between the Japanese and United States firms. |
 |
Let me conclude. The evolution of the disk drive industry is usually summed up -- when I hear people talk about what characterizes this industry -- as companies introducing faster drives, smaller drives, cheaper drives, and always with more capacity. But the point of my presentation is to suggest there are other dimensions of the industry that have been critical to United States success. These are new entry, intensive competition to generate and grow consumer markets, and the globalization of assembly. |
