History (1997): Exclusive Interview With Mark Geenen, President, Trendfocus

Mark Geenen, 39, president of Trendfocus, was born in a HDD drive, or nearly so. He has dedicated 3 quarters of his life to the devices. For the past 13 years, as a consultant, and golf aficionado, he spins at some 10,000rpm in order to take the closest possible look at the market for HDDs and their components. He has said: “Three-quarters of my working life have been spent in the storage industry.” He conducts research in drives, heads and disks, and has done so since 1984 after working for a couple of storage companies. He is now president of Trendfocus, “for the first time, a research company actually connected with what’s happening in drives, and then down from there, heads and media, equipment and so on.” The firm was created in 1993, and numbers 6 employees in all, “with over a hundred customers worldwide.” He travels widely: “In the first 7 months of this year, we were been in Japan four times.” He was also formerly the executive director of IDEMA.

Computer Data Storage Newsletter: What’s the trend of the HOD market for 3Q and 4Q?
Mark Geenen: The third quarter, which we’re just finishing, will not be a very good quarter. Shipments will be relatively flat, which is really alarming in one sense. 1Q was about 31 million drives, 2Q was 31 million drives, and it’s possible that this quarter, 3Q, will also be 31 million drives, so that’s three straight quarters of flat shipments. Everyone’s hoping for and expecting a big 4Q, but it’s the same every year, and every year we are disappointed. Some people forecast 25 or 30% in one quarter, but there’s no way. 15 or 20% growth maybe. I think that the 4th quarter will be good for the business, but 3Q will be weak.

What happened to Seagate in 2Q was more a price war problem or a demand problem?
Well, to really understand that situation, we have to look back and remember that in the last 3 or 4 years, Seagate has been the leading manufacturer of the server-class drives, they’ve had 60 or 70% of the market It’s logical to expect that at some point, they will have competitors. And this year was the first year that its competitors really could ship competitive products in high volume. So you had IBM, Quantum, Fujitsu, WD, all start to enter the market. And Seagate expected to lose some market share. However, in the second quarter, and into the summer, the demand from the server business started to flatten out a little bit. So not only were there more competitors, but demand itself was weaker than expected, which is the worst possible scenario. Seagate, obviously, has been dramatically affected by this, and we don’t expect the demand for server-class drives will improve much until the end of this year or early next year. It will start to improve, but not regain its original strength for 3 to 6 months.

If you look at the history of the HDD industry, it’s mainly a US market. Japanese and other Asian firms have tried to enter, but with many difficulties. So what do you think of the new style of a Hyundai, or a Singapore Technologies?
In 1994, the non-US percentage of the drive business was at its lowest point ever, somewhere around 5 or 6%. But we could see, since we travel so frequently to these companies and their factories, the investments they were making, the technology choices, marketing strategies and manufacturing strategies. And we were saying 3 or 4 years ago that the US companies had not yet won the war, that the Korean and Japanese firms haven’t given up. 1994 was the lowest point, but in 1995 that percentage rose to 9%. In 1996, it was over 14%. We expect that both Korean and Japanese companies will continue to gather share. More importantly, key technologies are coming out of Japanese companies. Very, very powerful on the heads business, and that’s true with media business now too.

Are there still opportunities for a start-up in this sector?
To enter with a similar strategy or technology, it’s very difficult for a new company to come in. But if you look at Quinta or TeraStor, 3 or 4 months ago, came in essentially a start-up company. Quinta was purchased by Seagate for over $300 million and they never shipped anything! TeraStor has a lot of investment and a lot of support, so the way they were both able to manage is to have a new technology, a new approach, to answer a question that hasn’t yet been answered.

But if you look at JTS, they have difficulties.
Well, JTS’ biggest problem was a lack of money, and they were late in shipping a 3-inch drive. Their entry into the market was based mainly on 3-inch, but they were late in getting qualified, and most of their money came from 3.5-inch. But they don’t have a competitive edge in 3.5-inch.

What’s your opinion on Quinta and TeraStor technologies?
Both of those companies are putting together different technologies to provide extremely high storage and recording densities. Historically, when companies try that kind of approach, it takes a little bit longer to go into mass production. We don’t expect there will be a huge impact from Terastora/Quinta for at least another year, possibly another 2 or 3. It’s not clear, but those companies’ potential is fascinating.

And do you see their technology replacing fixed HDDs?
I don’t know, that’s a very challenging question. Over the years that I’ve been watching this business, I’ve seen a lot of pretenders, from bubble memory to optical. Optical is probably the cheapest example of that. All the billions and billions of dollars that have been spent on optical storage, R&D, and yet has never seriously challenged HDD technology.

What are your predictions on 3-inch?
Our view all along has been that 3-inch technology has good potential. Which means that if companies can actually make this product, it can find a portion of the market that’s currently occupied only by 2.5-inch. However in the last year, the suppliers of the 3-inch drive have been extremely slow to enter into production. And that’s logical, any time you go into production of a new size, it takes more time, there are challenges. Secondly, from the systems side, the current design of a notebook computer, one year ago there were not many notebooks that could accept 3-inch, and 6 to 8 months ago, suddenly there were many that could accept 3-inch. Today, well over half of notebooks shipped can accept 3-inch. So the market is there. Unfortunately, still, the suppliers of the 3-inch disk drive aren’t in mass production, and so the opportunity is there, but 3-inch suppliers aren’t taking advantage of it quite yet. But we feel confident that they will start to do that by the end of this year.

Do you see other companies than Integral, JTS and Western Digital entering in this market?
It seems there are 2 areas for the 3-inch, one is the portable, the other is the high rotational speed server-client-class application. For portable storage, there’s a possibility that at least one other company can enter the market. In terms of the high rotational speed enterprise or server-class applications, there are several companies looking at that right now, and it’s very intriguing.

The 5.25-inch?
5.25-inch has found its niche, and that is in the consumer PC market for the user that needs high storage capacity at a reasonable cost. The performance of the 5.25-inch isn’t as good as the 3.5-inch. There’s really only one supplier of any consequence, and that’s Quantum.

Is the high-end 5.25-inch form factor viable?
Again, there’s only one supplier there, and that’s Seagate, and yes, we think Seagate will remain active in that market in the future. It’s not a huge market in terms of units, but in terms of revenue and profitability, it’s quite attractive for Seagate.

Why do 1.8-inch HDDs remain a niche market? Is it going to change?
The biggest challenge for the 1.8-inch business has been finding the right market that needs that kind of storage at the price. They have found that in the pen-based market, the ultra sub-notebook market, to some extent the PDA market, and other small segments of demand. Until the 1.8-inch drive can store much larger capacities more cost effectively, it will remain only in some of the smaller application areas. So it’s possible, however the exciting thing about 1.8-inch or another form factor is that some of the application areas, for example digital cameras, some of the Japanese companies have digital cameras with a 260MB drive in them. A 260MB 1.8-inch drive can hold 3,000 still pictures, but only take about 20 minutes of video. So if they could get 1GB, so they could take over an hour of video, and literally over 10,000 still pictures. 10,000 still pictures, that’s enough still pictures probably for my lifetime – more than my wife will ever take. But for video applications, that could turn out to be a very huge market for 1.8-inch drives.

How do you see the current and future evolution of magnetic disk heads with MIG, thin-film and MR technology?
I think that for the most part, the game is over, and it’s an MR world. Yes, inductive heads will still be a significant part of the market for the next 1 or 2 years. But by the end of 1998, there will be very few drives shipped with inductive heads. Maybe some of the more mature cartridge drives, like Nomaï. SyQuest, I don’t know. But among fixed disk drives, almost all of the inductive heads will be out of production, so it will be MR. And then GMR starts next year.

What’s the current percentage of HDDs with MR, thin-film and MIG heads?
Well, right now, MR accounts for nearly 60% of the products that are shipped. Inductive heads are about 35 to 40%, and then the rest is MIG, a small percentage.

Your projections?
Well, we estimate that by 1999, MR will account for over 80%, and GMR will be most of the remainder. Inductive and MIG will be just phasing out at that time.

Do you see any future for planar heads?
Yes, as a matter of fact, we think that planar technology has a lot of merit, and as you apply MR or GMR, you have a chance not only to yield at a higher level, but also to drive costs down. So we think that planar technology will have a role in this industry.

When do you see GMR heads integrated into disk drives?
I think that we will see some GMR heads shipped in drives in the second half of next year.

Which companies?
It’s difficult to say, but the Japanese have probably been the most aggressive about incoporoting GMR, so I would expect some Japanese firms among the first.

And after GMR?
Well, I don’t know.

What’s the limit of magnetic technology?
It’s beyond my own personal training to say, but if you listen to the leading IBM researchers, they feel confident that for at least the next 15 years, magnetic storage is just fine. No problem.

Who are the principal leaders and what is their market share for disk media?
If you look at the independent market, Komag is still the largest manufacturer. They have in the range of 20 to 25% of the market. After that, in the independent area, Fuji Electric is second, and then there are several companies, Showa Denko, Mitsubishi Chemical, HMT and then several more after that. One market leader and one close second, and 5 to 10 viable companies that hold their own.

What about non-independents?
Well, Seagate’s by far the largest manufacturer. For example, in the second quarter this year, they made about 22 million disks. That compares with Komag, which made roughly 14 or 15 million that quarter.

What about Western Digital, are they a significant player?
Western Digital makes about 5 or 6 million per quarter. And so they are a sizeable player, but Seagate is really the big powerhouse.

IBM?
IBM is also quite large, they’re making in the range of 8 or 9 million per quarter, approaching 10, but right now, IBM doesn’t have enough capacity to satisfy even their own requirements, so they’re buying disks from outside today.

What about heads?
The two big players at this time are Read-Rite and SAE/TDK. Both of those companies are the market leaders among independents. Yamaha, I think is the third, but there are a number of players in the head business that are coming up. If you include the captive market, then once again Seagate is the largest, they made over 70 million heads in 2Q, in comparison with the roughly 25 million that Read-Rite made. Three times Read-Rite’s amount.

IBM?
IBM is quite large, they made over 20 million heads last quarter. There are many companies emerging to try and take advantage of that market, for example Alps Electric. They used to be a big MIG head supplier several years ago; and then they were quiet, and now they’re a very sizable MR head’ manufacturer. On top of that, there are several Japanese companiess, like Sanyo, Hitachi Metals, and others that are now entering the MR head business, and it will be very interesting to watch next year as the big players mature their technologies, and the new ones enter in. It could be quite competitive in 1998.

Do you have an idea of the percentage of heads that IBM sells outside?
Oh, it’s well under 10% today. By the end of the year, it could be just over 10%.

What’s your opinion on the vertical integration strategy of HDD drive companies? In other words, do you think that it’s an obligation for an HDD manufacturer to have its own manufacturing facility for heads and media, in order to succeed?
This has been one of the age-old questions: what’s the formula for success? Here we are almost all the way through 1997, and Seagate the largest disk drive company. But there are a lot of others who don’t have as high a vertical integration, so I don’t believe that vertical integration is an absolute requirement for success. What is a requirement, though, is if a drive company is not vertically integrated, it must work together very closely with its key technology partners. I don’t want to make that sound like a c/iche, but it’s true, there needs tobe a very close partnership between drve company and the components partner. Some people say that Seagate’s business style is the only way. I think that’s not true. There are many head makers and media makers, so there’s a chance to work closely together with those companies. achieving maybe not the same low costs that Seagate enjoys, but to attain the same time-to-market and the same production yield. Seagate may have very cheap costs, but they have their own technical challenges that independent head and media suppliers are probably better at dealing with.

So you see some future for independent head and media makers?
I do. It will not be an easy market to participate in, it will get even more difficult than it is today, but we still think that there is an opportunity today to prosper and profit in that area.

What’s the trend in the number of platter media per disk drive?
In the past 3 or 4 years, the number of heads and disks per drive has actually gone up slightly. We think now that MR technology is coming on strong. And while the demand from the user, though still growing, is not growing as fast as we thought, we in fact believe that the number of heads and disks per drive will rotten out, and begin to decline very slowly in the next 3 or 4 years. Now it’s not a huge problem for the head and disk makers, but it is a challenge for cost and technology.

Today, what are the average head and media ratios?
Well, it’s about 3 disks per drive, plus or minus a little bit, and about 6 heads per drive.

What is the trend in coercivity of media?
Media coercivity is also changing. Today the mainstream of the market is 2,000Oe and 2,200Oe.

Which means what in terms of capacity?
There’s no correlation between coercivity and capacity, but a 2,000Oe generally speaking would be used for 850MB to 1GB per disk. and 2,200Oe is roughly 1.2GB. The next gen is 2,400Oe, and that’s 1.6GB and higher. There’s no direct correlation, but there’s a trend.

Do you follow also the sliders’ market?
Yes. Hutchinson is the biggest supplier, but there are 2 or 3 others that are serious. One is called NHK Spring, a Japanese firm that is very well positioned with some companies. Then there’s KR Precision, operating out of Thailand, and the third is Magnecomp, and they have a factory in Hong Kong and also in China.

The trend in the slider format?
Almost everything is 50% now. There are some companies that are using 30%. Samsung is using 30% heads, because that’s what Silmag makes. Almost all of Fujitsu’s drives use 30% heads, but the transition to 30%, the complete transition 30% is at least another year or two away.

Do you see more automation in HSAs and HGAs in the future?
We do. Right now, looking at the 30% transition, Fujitsu and Silmmag are the big players. It just so happens that both of those are very highly automated in their suspension or HGA assembly. As the slider becomes so small that it’s almost like the head of a pin, it’s difficult for a manual operator to work with that and guide the wires.

Will it be the same with drive assembly?
Actually, no. I think the drive companies fo the most part have found the approprite balance between automation and manual assembly. As you know, MKE is almost completely automated in the assembly of drives. On the other hand, companies as advanced as IBM, whether making drives in Hungary or Thailand, are very manual-oriented. So companies are looking for the appropriate balance, and I believe they’ve found thot Now there will be additional, not automation, but mechanization, which means that certain stations in the process will be automated, but when you go on to the next station, it will be manual. So there will always be that balance.

What is the percentage in value of each component of a HDD drive: heads, media, PCB?
If you analyze a mainstream or low-end disk drive, and say that the material cost is somewhere on order of a hundred dollars, and that’s using 2 disks and 4 heads, well the disks will cost around $10 to $11, so that’s $20 to $22 total disk cost. The head cost is somewhere around $25 or $30, so well over half the cost is heads and disks. I’m talking about the whole headstack assembly. The PCB is significant, the spindle motor can be 7 or 8% of the drive cost, but heads and disks are the major cost component, and then the PCB with the chips.

What percentage is the workforce?
It’s well under 10%.

And testing?
Testing is 1 or 2%. The cost of material is well over 80% of the cost af the drive, you have warranty and overhead and labor, but that’s under 20%.

What’s the difference between the cost of a drive out of factory and the OEM cost?
Actually, to sell to OEMs is typically slightly less profitable then selling into distribution. The reason is that the OEM customer promises continuous volume, or continuous orders, and in return for that security, wants a lower price. Distribution is much less predictable, demand can be very strong or very poor. Typically, though, the profit is 2 to 3% higher, on average, than with an OEM sale. If you look at the cost of drives now, we had a price war in the second quarter, in the high-end of the business. And this summer, even in the mainstream PC market saw pricing compress, forced downward. And if you look in the newspapers now, in the US, you can see 3GB drives for well under $200. It’s amazing how prices have moved downward, and it’s very dangerous for the profitability of the disk drive company.

What are your hobbies?
My family, my three children. After that, sports and particularly golf. I’m a keen golfer.

How many days of vacation a year?
Last year it was 2 weeks, this year it will only be one to one and a half.

This article is an abstract of news published on issue 117 on October 1997 from the former paper version of Computer Data Storage Newsletter.