There's a battle brewing over data storage. As computers and consumer electronics start requiring more storage capacity for applications like music and video, hard drives are adding to their capacity and decreasing in size.
But NAND-type flash memory chips are becoming an ever-more economical solution for certain types of storage. Take Apple's iPod for instance: The 40- and 60-GB models that support video require hard drives, while the slimmer iPod nano uses flash memory.
No wonder companies who make flash memory have big plans. One of them is to replace the hard drives in notebook PCs, and there's reason to believe it may not be an unrealistic goal. Flash chips consume less power than hard drives and are less likely than notebook hard drives to fail when dropped.
Naturally, William Watkins, chief executive officer of Seagate (STX
), the world's largest hard drive manufacturer, is unwilling to cede any ground to flash memory. BusinessWeek.com writer Arik Hesseldahl caught up with Watkins recently to talk about the changing storage business and why the hard-drive sector is plenty healthy.
Steve Appleton, CEO of Micron (MU
), a chip maker pushing into flash memory, says the day is coming soon when flash memory will, at least in notebooks, replace hard drives. What do you think?
Look at the three largest NAND flash memory producers in the world: Samsung. They're spending a fortune on their disk-drive business. The next is SanDisk (SNDK
) and Toshiba (TOSBY
), and Toshiba is spending a fortune on its hard-drive business. Then there's Hitachi (HIT
) [a participant in flash memory maker Renesas] and it's spending a fortune to stay in hard drives. And I've asked all three of them: If you believe in flash so much, then sell me your hard-drive business. None of them would.
You approached Samsung, Toshiba, and Hitachi about buying their hard-drive units?
Sure. Every time they talk about how great flash is going to be I've said, great, then sell me your hard-drive business. There's going to be a lot of flash storage sold, no doubt about it. But here's the other thing about flash. If you think about moving content between your home and your handheld and your car, and you have a device whether it has a flash drive or a hard drive, somewhere there's an enterprise server that has delivered the content to your home. And you've downloaded from a PC to your handheld device, and eventually you may want to move to your car or whatever. But if you think about it, that behavior of moving that music drove three storage purchases: One at the enterprise level, one in the home, and one in your hand. If the flash guys get the handheld one, that's great for them. I get the home and the enterprise one. And then because the content is so valuable you want to back it up, you buy a backup drive. Those are all big hard drives. We're much more interested in the behavior of moving content electronically. And that drives storage in four or five different application sets. Some of them will be flash, but a whole bunch of them are going to be hard drives.
Appleton says the average notebook drive is about 30 GB, and that once Micron and companies like it can produce flash that's price competitive, he can win that business away from companies like Seagate.
We shipped an average [hard drive] last quarter of about 70 GB.
So they're getting bigger. You're saying the flash vendors like Micron and Samsung and Hynix probably can't get to those capacities in a way that beats a hard drive on price and power consumption?
Power consumption won't be anything that people will pay anything for. What it's going to come down to is price per gigabyte. Here is the thing about flash. Whatever the flash chip is, whether they're doing it on a 2-GB or a 4-GB process, that becomes their lowest cost denominator. And then all they do is stack chips to get to the next higher capacity. So if they can make a 4-GB for $5, then it's $20 for 32 GB, and $40 for 64 GB. On the other hand, whatever drive we can make with a single platter becomes our lowest cost item. So when we do a single-platter, 100-GB notebook drive, well it doesn't cost me very much to go to 200 GB. All I have to do is add another platter. It's very cheap for me to go up. But it's very hard for me to go down. For example, right now we make an 80-GB, single-platter notebook drive—a single platter, two heads. That's my cheapest platform. It's very hard for me to go from 80 GB down to 20. But I can get to 160 GB really cheaply, and with flash it's just the opposite. In three years we would say the average notebook drive will be 250 GB and they'll be selling for $45. If they think notebooks are averaging 30 GB right now, then they already show they don't understand the market. We're averaging more than twice that much already, and it's growing every quarter.
So it will really come down to price per gigabyte?
It will be a price-per-gigabyte battle. They have better power, we handle video better. But we're working on power, and they're working on streaming video.
That makes hard drives look like a pretty good business, but you just reported quarterly earnings that disappointed some analysts.
We were kind of surprised (by the reaction). What did people think, we weren't going to have to pay for the Maxtor integration? We told everyone there would be restructuring costs. I think we were able to pull it in, and people were having a hard time understanding our June numbers. If you look at Seagate standalone we made $241 million, and if you look back we hit the middle of the guidance. We think things are going pretty good. We're pulling Maxtor in a little sooner than expected, and we think the pricing environment for hard drives is a little more aggressive than we thought, and I think some people did some stupid things on pricing. But that said, we feel pretty good. We said we're expecting aggressive pricing for the rest of the year. But we think with new products coming out we'll be able to deal with that and get our margins back up. We threw out a number. When you throw out all the one-time things, we're going to make $1.2 billion in operating profit next year. We made a billion this year. That's a 20% growth in operating profit and greater than 20% growth in revenue. That's a problem we'd all love to have. I just wish people wouldn't get so lost in purchase-cost accounting numbers.
What should we be looking for from Seagate for the rest of the year and into 2007?
There's some macro things that people have to get their heads around, which is what the Fed is doing and what is going on with gas prices and consumer spending. The next quarter is going to be consumed with what consumers are doing and whether or not they're going to be spending at Christmas. I think Microsoft's (MSFT
) Windows Vista will be a big driver in the January-February time frame. I think it will drive notebooks quite well. It will be interesting to see how fast it moves into businesses and homes. I think you've got some interesting things happening with gaming. Sony's (SNE
) PlayStation 3 is coming out, and we're a qualified supplier to that system as well as the Xbox 360. We have a whole new product refresh coming up that drops our cost structure quite significantly that will allow us to be highly competitive. And we have this new opportunity in 1.8-inch drives, which is starting to grow for handheld video. Apple's (AAPL
) iPod is one, but there are a lot of people putting out handheld video devices. We did not have a drive for that business that will come in capacities of 60 and 120 GB, and so we will have one in the December quarter and we think that will be a nice little driver for us. We feel a whole lot better about things than Wall Street does right now.