The history of Seagate by founder Al Shugart

Speech to Seagate Employees at Company Meeting
I see so many new, young faces in the audience today. I’m sure that many - if not most - of you weren’t even born when I began working on the industry’s first disc drive. So I thought this would be a good time to round out your education, and give you a little first-hand history.

I came into the computer industry quite by accident. I graduated from the University of Redlands in 1951 after four years and four different majors, and took a job with IBM in Santa Monica, California as a customer engineer (that’s field engineer nowadays) because I could start the day after graduation. The pay was good, too: $275 per month. I was 20 years old, married, one kid, and dead broke.

I learned after one week at IBM that field engineers were the lowest on the totem pole when a field engineer in the office was PROMOTED to Salesperson. I’ll never forget that early lesson, and I’ve held it against salespeople ever since.

After having fixed all the troubles one could have with punch card accounting machines, in 1955, about when Scott McNealy was born, I transferred to a small IBM R&D lab in downtown San Jose, California.

I won’t ever forget the day that Don Johnson, one of the pioneers in disc development, invited me over to see how he created his discs. He was rotating this giant 24-inch platter. Then he poured a solution of iron oxide on the disc from a Dixie Cup. No clean room, equipment so crude that we rotated the disc with a foot pedal. And that dixie cup didn’t look out of place at all.

I certainly had no idea I was walking into the beginning of a technology and product development program that would have such a profound impact upon the computer industry.

The First Disc Drive
Using the accepted “big blue” approach, we made our disc drive big. Using these crude techniques, we produced a disc that supported a recording density of 2000 bits per square inch with 100 bits per inch and 20 tracks per inch. We stacked fifty of the 24-inch discs on a vertical shaft, and had a disc file that would store five million characters. And it weighed only One Ton! Average access time was a screaming 1-and a half seconds, and we spun the disc at 1200 rpm. No fire code, no ECC, no address marks, no flags for spare tracks. It should have made the controller easy, but it didn’t. We didn’t even know how to clock data without a clock track. Oh, by the way, the electronics were all in vacuum tubes. No volumes of semiconductors to choose from.

Who Could Use This Much Storage
Now that we had this big drive, we had a bigger marketing concern. Who could ever use this much storage? Five million characters was a lot. Notice that I said characters. This was long before the days of the 8-bit byte. A character was six data bits and one parity bit. Dark Ages!

RAMAC
Our magnetic head technology was equally crude. Although we had learned a lot from drum recording technology, the disc presented its own set of problems. To create an air bearing to separate the head from the disc, we routed compressed air through tiny orifices in the head carrier. We used this same air supply to load the head. Unfortunately, this required a big supply of external air. In fact, the thing required so much air that we could use a total of only two heads to serve the one hundred disc surfaces.

To reach another disc, the two heads were unloaded, removed horizontally from the disc stack, then moved vertically to the desired disc, then horizontally again to the desired track, then loaded. This was a lot of mass, moving pretty fast, and the file really rocked in its shock mounts during accessing.

Actually, it worked pretty well. After all, users thought in terms of seconds per function, not milliseconds and nanoseconds. And we gave the user five million characters of storage for a rental cost of about $750 per month. Of course, this didn’t include the controller or the air compressor you needed to supply all that external air.

I believe that IBM built about 5000 of these files. Most of them were used in a system called RAMAC or Random Access Method of Accounting.

A Better Air Bearing
During these early production years of the RAMAC file, we were looking for a better air bearing. Then sometime during the late fifties, a set of articles in the IBM Journal of Research & Development discussed a very old principle -- the self-acting air bearing or slider. These articles became the bible for anyone interested in disc files.

The self-acting air bearing did not require an external air supply. This meant we could place a magnetic head on each disc surface.

IBM Advanced Disc File
What a breakthrough. IBM introduced another 24-inch diameter disc file called the ADF, or Advanced Disc File. Although it was about the same size as our first RAMAC, it stored ten times as much data -- 50 million bytes. And by eliminating head movement between discs, we cut the average access time to 165 milliseconds. Recording density was also improved by a factor of 20 -- 25,000 bits per square inch with 500 bits per inch at a track density of 50 tracks per inch.

Many people read the Journal articles, our bible, and it wasn’t long before IBM had some competition. Two early competitors that I remember were Telex and Bryant.

I do have to credit IBM management. They backed the disc as the random access technology of the future. In fact, they bet all their marbles on the new disc files. On the other hand, Univac chose the magnetic drum rather than the disc. It wasn’t the only bad decision by Univac, but it ranks right up there with the worst.

Disc Pack/Standard 14-inch disc
In the early sixties, as IBM’s first disc file with slider bearing heads went into production, we came up with another breakthrough. This was the removable disc pack. And for the first time, we see the disc size that was to become the standard -- a 14-inch diameter disc.
It began with a pack holding about 3 million bytes, then the real standard, a disc pack that carried 7.5 million bytes.

Capacities Grew
Capacities continued to grow. The next pack held 29 million bytes. The recording density was now 220,000 bits per square inch or almost 100 times greater density than the first RAMAC file. This same general technology carried through to IBM’s first track following servo, with an aerial recording density of 1.5 million bits per square inch.

Technology Limits/The Winchester
It’s at this point that removable discs began hitting some technology limits. The height of the air bearing and the possibility of contamination were tough barriers for removable disc packs to overcome. The infamous head crash became an ever-present danger. This led to the development of the first hermetically sealed disc drive -- IBM’s Winchester.

Winchester technology featured low mass, lightly loaded heads, starting and stopping in contact with the disc.

More Improvements Needed/Fixed Discs
But improvements were still necessary. The disc packs, now sealed in a data module, were still removable and still expensive, so discs became fixed to further increase the recording density. Linear bit density was increased to 6425 bits per inch, and an incredible aerial density of over 3 million bits per square inch was achieved. This was followed by an aerial recording density of 7.8 million bits per square inch.

Compare this with the 2000 bits per square inch of the first RAMAC file. Density had improved by 3900 times.

And Then There Was The Floppy
As we reminisce about old times, we shouldn’t forget about a critical parallel development in data storage -- the humble but ubiquitous floppy.

The floppy disc was actually the result of advances in semiconductor technology. Here’s why:
In the early sixties with the introduction of the IBM 360, control memory was employed to a great extent in both CPU’s and peripheral controllers. This control storage was implemented in read only memory, because magnetic core and semiconductor memory were much too expensive.

However, by the time the IBM 370 was developed, semiconductor technology had advanced. Now, control storage could be implemented in semiconductor memory. Since this memory was volatile, a loading device was necessary. Magnetic tape was considered but the need for loading diagnostics as well as the control program seriously detracted from the desirability of this approach. Why not a cheap disc that would provide the random access speed needed for diagnostic loading?

With such a low cost disc, you would have an economical, random access, program-loading device. And once such a device was available, why not add a write capability for logging?

So semiconductor technology and the big IBM 370 set the stage for the floppy, the data storage that in turn, helped to launch the small systems revolution.

Many of us saw the great potential of this little disc. That’s why I formed my first company, Shugart Associates, in 1973.

On To Seagate
I was lucky to have played a role in the early days of floppies at Shugart Associates, and it led to a much more long-lived role at Seagate starting several years later.

The start of Seagate is sort of interesting, so I’d like to tell you about it.
In late September 1979, the desktop computer market was going bananas. Millions of units were being shipped annually and most of them had a small auxiliary memory device called a minifloppy disc drive. These minifloppies were a reduced size version of the original floppy disc drive introduced in volume about 5 years early by Shugart Associates.

I had been working around computers and disc drive memories for over 25 years, and had discovered one fundamental that transcends computer systems of any size; and that is: A computer system’s appetite for memory is insatiable.

And that was and is true for even very small computers. As more and more applications were put on these systems, the memory requirements grew. And in late 1979, these additional memory requirements were being met by adding a second and third and fourth minifloppy disc drive.

And Then Came Finis
Finis Conner, who joined me as a founder of Shugart Associates, came to me in late September of 1979 with the idea to build a fixed, rigid disc drive the same physical size as the minifloppy, with higher performance and higher reliability, and with 15 times the storage capacity at 3 times the cost. He said that if this were possible, he could sell to every desktop computer manufacturer that was shipping systems with more than one minifloppy; that is, our device would fill the memory need for more than one minifloppy.

I thought this was possible so we decided to go into business. On October 1, 1979, Finis and I hacked out an 8-page business plan that predicated our nearly taking over the world, and very quickly -- it was a very, very aggressive plan. It had to be. Finis and I had both run out of money and our personal habits needed recapitalizing.

The Search For Venture Capital
Each of us kicked in $10,000 and hit the road with our plan. We found a mechanical engineer, an electrical engineer, and an operations manager very quickly. It seems get-rich-quick schemes are easy to sell to poor people. We decided to let my daughter, Terry, who was in college, kept the books until we could afford a financial officer, since she worked cheap.

Finding the money to finance the venture wasn’t quite that easy. We reasoned that our idea was worth $2 million dollars, and that we would sell 25% of our plan for $500,000.

Page Mill Group
Our first stop was the Page Mill Group, a venture capital firm made up of very successful people from the electronics industry. They would surely see the wisdom in what we were doing. Bob Noyce, Lester Hogan, John Young, Ken Oshman, and several other equally famous and successful people.

After my presentation, John Young, who, if you don’t know, was the president of Hewlett -Packard, said: “Al, why should we pay half a million dollars for only 25% of a company that’s only an idea in the minds of you and Finis?”

I said: “John, perhaps you shouldn’t.” And they didn’t! (Finis said I needed to brush up on my marketing technique).

Exxon et. al.
But Finis and I decided that perhaps they didn’t have enough money. So we set our sights on bigger bucks. We knew that the Exxon Corporation made venture investments, and Exxon seemed to have a good balance sheet and a lot of cash. So we made an appointment with the Exxon guy in New York who handled that sort of thing and we flew off to New York.

We arrived early in the evening the day before the meeting, and went out for a really nice dinner. We decided to celebrate this big deal we were going to close in the morning, so we got a bottle of really fine (and expensive) wine. When we returned to the hotel, there was a message from the Exxon guy that said he had to leave town, the meeting was canceled, and he would call us in a few weeks. That was an expensive call.

But we weren’t discouraged. Following that, we got turned down by the Mayfield Fund, and Idanta Partners, and several funds didn’t even return our calls.

But we still weren’t discouraged. And money wasn’t our only problem. We needed a disc, and let me explain how important THAT was.

We Needed A Disc - 3M
In a rigid disc drive in those days, the data was magnetically recorded on an oxide coated aluminum disc. There was a great deal of technology, and a lot of tooling money involved in producing magnetic discs. We needed a commitment from a magnetic disc manufacturer to develop and manufacture a disc that was a different physical size from any in the industry. It would require a manufacturer to not only spend a lot of money on developing the disc, but an even greater amount in tooling for production. The total dollar requirement for this made our monetary needs seem small.

So first we flew to Minneapolis to see the 3M Corporation. The 3M people were very interested in the project, but they couldn’t do anything because our schedule was inconsistent with their view of reality.

But they were really nice people and agreed to help our effort to get the company off the ground by cutting down some larger discs to the required 5.25-inch size we needed. And even though the center hole of the disc was larger than we could tolerate in actual use, the disc should serve as a good visual aid.

While we were waiting for the 3M sample discs, we called on Memorex, but they never called us back. Within several days, 3M hand-delivered 6 disc samples to me in California, just to help us get going.

Dysan
With the disc samples in hand, we called on Norm Dion, president of Dysan Corporation in Santa Clara, a magnetic disc manufacturer. I handed Norm one of the sample 5.25-inch discs and he just held it and stared at it for what seemed like hours (probably 15 seconds). Dysan was just getting into production of an 8-inch disc, having manufactured 14-inch discs for several years. Finally, Norm said: “You know, 8 inches was the wrong size.” I figured we had him at that point. He saw the tremendous future in what we wanted to do, and agreed to develop and manufacture the 5.25-inch disc.

Then he asked us how we were doing on getting financed. Not wanting to show any weakness, I told him we expected to close something soon, trying to keep my voice from cracking. He said that was a shame since he thought it would make a good package for Dysan to fund our development effort as well as commit to the disc.

We quickly saw the wisdom in this and, on November 14, 1979, six weeks after we put our plan together; Norm gave me a check for $10,000, as a show of good faith for his agreement to invest $500,000. We shook hands, and we had a deal.

We had always planned to get the lawyers to document the deal, but we never got around to it, and it was never really necessary. The following June, we raised another million dollars in capital through venture capitalists - they DID need to document the deal. So the total venture capital put into Seagate was only $1.5 million. An unbelievably small amount of money by today’s standards.

We Needed Parts - Turn Left At Leo’s Liquors
It was tough to find more believers at that time. We had 8 people in the loft of a suite in Scotts Valley, and we needed to place orders for parts; some big orders.

We knew we had to get magnetic heads on order quickly so we called the manufacturer’s representative. I told him we wanted to buy 100,000 magnetic heads (at the time this was about a $2 million deal). He said he’s come to see us. Where were we located, he asked. He had never heard of Scotts Valley.

I gave him directions . . . come down out of the mountains, turn onto Santa’s Village road, go a quarter of a mile, turn left at Leo’s Liquors, cross a little bridge and go into suite C in the only building there.

He repeated the directions and then said: “And you want to buy 100,000 magnetic heads?” I expected him to say, “come on now, who is this?”

Product Development
We completed our product development in 5 1/2 months and showed our product in a hotel suite at the National Computer Conference in Anaheim in May of 1980. We got orders during that show including a $200,000 prepayment, and began shipments 6 weeks later.

We shipped 50 units our first month, and by October we were shipping 10 units a day out of a 1000 square foot lab.

The ST506 disc drive stored 6.38 million bytes of data and sold for $1,500 in single quantities: down to $775 in quantities of 5,000.

A Huge Market
The market for this size disc drive was quite large. We projected that the worldwide market would grow from 1100 units in 1980 - which was our total production output - to one million units in 1983. And although our projected shipment grew at an astounding rate, we told people that we didn’t predict being able to maintain the 100% market share we enjoyed in 1980. In addition to Texas Instruments and CII-Honeywell Bull whom we had licensed to manufacture and market the product as a second source, we did expect to see several competitors in the marketplace later in 1981.

In our first full operating year we did about $12 million in revenues and made about $1.8 million net profit.

Things moved so fast that we had an initial public offering of our stock only 22 months from when we started.

What Makes The Great Opportunities Possible?
So what really makes these great opportunities? The availability of capital? Certainly. But I really think it has more to do with changes in our society. Let me talk a little about that.

When I was working at IBM, the corporation organized a science advisory board made up of a group of very distinguished scientists. This group met periodically with IBM management and senior technical people to give us the benefit of their wisdom and learning.

I was quite fortunate when in the 1960’s I was invited to a luncheon with the science advisory board in San Jose, along with other senior technical people from the lab I worked at.

Following lunch, the IBM host asked the members of this advisory board if they would each comment on the terrible unrest that was going on in our universities, and the apparent change in behavior of all our younger people.

If you don’t remember or weren’t around at the time, the 1960’s found a lot of our young people in jail for acts against public policy. I recall one columnist writing that while he was driving down the road he saw a sign that said “free firewood,” and his immediate thought was: Who is this guy Firewood, and why is he in jail?

Anyway, each of the 6 or so distinguished scientists addressed the subject:
Norbert Weiner, the famous nuclear physicist and Nobel Prize winner, began and expressed great disappointment in our youth with their erratic behavior, and concluded that we were going downhill. The following speakers expressed the same disappointment, and offered theories on the behavior, and proposals for fixing it.

The opinion was generally unanimous until the last distinguished scientist spoke. I can’t recall his name but I can picture him clearly. He was a world-renowned mathematician, long since retired from his position at Columbia. This quite elderly, gray-haired gentleman said that what was going on with our young people was the result of a change in society that was underway -- and that he was both pleased and excited about it. He mentioned individualism, opportunity, creativity, and a true thinking and nourishing of society. He said we could close our eyes and hide from the change, or open our eyes and participate in it -- because it was changing anyway. And he had no fear of the future.

<strong>(I thought at the time -- “Easy for him to say; the old bugger is in his 80’s and won’t be around when these crazies are running the country”).
But you know, the man was absolutely right. We were seeing an expression of individualism as a result of a change in society that had begun, and is still in process, and will continue as long as people have ideas.

The Information Society
We’ve moved from a mass industrial society to an INFORMATION Society, with a much more profound impact than the 19th century shift from an agricultural society to an industrial society.

We’ve moved to an age of the power of the individual. Where the strategic resource in the industrial society was capital; the strategic resource in the postindustrial society is Information and Knowledge. And that’s not only renewable, but it’s self-generating as well.

I believe that this provides for tremendous entrepreneurial activity in the world today. Because the strategic resource is now what we have in our heads. Access to the system is much easier. We have seen an impressive increase in the creation of small businesses over the last 20 years, and large institutions have restructured to encourage entrepreneurial activity within decentralized organizations.

In 1950, 65% of the people working in the country were working in the industrial sector, and only 17% in the information sector. Today, we’ve flipped that.

The age of the individual has brought decentralization. We have seen large airlines collapse while new local and regional airlines have been established.

Large circulation, general-purpose magazines have folded while thousands of special interest magazines are being published.

Great umbrella organizations like the American Medical Association continue to weaken as the groups within it -- pediatricians, plastic surgeons, and cardiologists -- specialize and get stronger, along with county and local medical groups.

And it’s happening all over the world. This great new age of individualism and its subsequent decentralization has led to the great number of opportunities for new leaders today.

Unfortunately, the age of individualism and special interest groups has also found a lot of jobs for a lot of lawyers, who are misusing our legal system. But that’s a story for another time.

Seagate is a leader in the new Information society. Seagate is in the Information business. We have not only survived, but thrived by preparing for, and embracing what we know is inevitable - CHANGE. <strong>We are leaders. So we will go out there again this year and do what we do best: “Find a parade, and get in front of it!