CHAPTER 3

MAGNETIC DISK TECHNOLOGIES

David A. Thompson

INTRODUCTION

Magnetic hard disk technology is the undisputed leader for online storage applications, i.e., for storing data inexpensively, but in such a way as to be available within a time period less than human response times. Although the time delay for the first data bit to become available is thousands of times longer than for semiconductor memory, the data (when it does arrive) streams from the disk at tens of megabytes per second. The cost of hard disk storage is a few cents per megabyte, which is several orders of magnitude less than for semiconductor memory.

Tape storage is cheaper than disk storage, but its access time is much longer. It is possible to configure optical storage for online applications, but such a system is inferior to hard disk storage in every respect: cost, data rate, access time, etc. Only by moving to a library configuration or a data interchange application, which both utilize the removability of optical media at the expense of access time and mechanical complexity, does optical storage find a successful niche in which it can compete. There is at this time no technology that appears to have a chance of displacing hard disk storage in the next 10 years.

One major difference between hard disk and interchange media like tape and disk is the effect that standards and the need for backward compatibility have on the rate of technological progress. Hard disk standards involve external attributes of the drive only: physical size, power, cooling, and the data bus protocol and connector. There is no need for the customer to know what technology is being used to store and retrieve data within the box.

The situation with interchange media is different. The customer is vitally interested in his ability to read old formats and share media with other customers. There is also a benefit from utilizing standards from the audio and video entertainment industries. As a result, almost all of the tape media progress is spill-over from the much larger entertainment market, and many of the optical storage standards are based on compatibility with audio or video players.

This difference is also reflected in the nationality of the companies that manufacture disk drives. As recently as 1997, U.S.-based companies produced about 80% of the revenues and an even greater fraction of the hard disk drive (HDD) units sold. This is just the opposite of the situation in tape and optical storage, where Japanese and other Asian companies dominate the business for both entertainment and for computer storage devices (see Fig. 3.1 and Tables 3.1-3.3).

Fig. 3.1. U.S. and non-U.S. HDD market share per year (ASET/SRC, Miura, March 1998).

From this data, one might expect that Japan is not a center of HDD technology. This impression is completely false! Consider the following observations:

• The most profitable and fastest growing segment of the HDD business is the 2.5-inch laptop drive. It is also the portion of the business where areal density leadership is most important, since size and power consumption are more important in this application than is price per MB. All of the current 2.5-inch drives are designed in Japan. (This includes IBM's Fujisawa Laboratory.)
• According to Mark Geenen of TrendFocus (in a lecture on June 23, 1998), the fastest growing companies in the HDD business and in the components business (heads, disks) are Japanese. Hitachi and Fujitsu are especially strong in their base technologies, and TDK, Alps, and Yamaha are greatly increasing their market share in the head business. A shakeout due to overcapacity in the disk business will favor well-financed companies, which include some of the Japanese companies. Overall, it appears that Japan's share of the HDD business is poised to greatly increase in the next few years.
• The approach of the superparamagnetic limit may put an extra emphasis on technological means to change the evolutionary approach. The panel saw in this study that the larger Japanese companies have a strong advanced technology position in devices, materials, and recording physics.

Other examples of a rising Japanese presence can be seen by a few months of press clippings (Table 3.4), by the order in which companies have brought magneto resistive (MR) heads to the marketplace (Table 3.5), and in the areal density being shipped as of May 1997 (Table 3.6).

The connection between the HDD industry and the Japanese universities has been weak for many years, but the recent establishment of the Storage Research Consortium (SRC) has begun the process of building the sort of ties in Japan that exist in the United States and Singapore.