This matter is before the court on IBM’s motion for a directed verdict. F.R.Civ.P. 50(a). In Chisholm Brothers Farm Equipment Co. v. International Harvester Co., 498 F.2d 1137, 1140 (9th Cir.), cert. denied, 419 U.S. 1023 , 95 S.Ct. 500 , 42 L.Ed.2d 298 (1974), the Ninth Circuit said:

With this standard in mind, the court will review the evidence that Memorex has presented. First, however, it is necessary to briefly discuss the substantive law that controls this case.

Memorex has alleged that IBM’s Madrid disk drive is an illegal tying arrangement. A tying arrangement involves a seller’s refusal to sell one product (the tying product) unless the buyer also purchases another (the tied product). Northern Pacific Railway Co. v. United States, 356 U.S. 1, 5-6 , 78 S.Ct. 514 , 2 L.Ed.2d 545 (1958). Memorex claims that IBM’s practice of selling the Madrid drive unit (the tying product) and the Madrid head/disk assembly (the tied product) for a single price violates Section 1 of the Sherman Act, 15 U.S.C. § 1 , and Section 3 of the Clayton Act. 15 U.S.C. § 14 .

In Moore v. Jas. H. Matthews & Co., 550 F.2d 1207, 1212 (9th Cir. 1977), the Ninth Circuit said:

The court indicated that in theory, Section 1 of the Sherman Act requires that both the second and third criteria must be found to establish the illegality of a tying arrangement, while either is sufficient under Section 3 of the Clayton Act. However, the court recognized that the practical difference between these two statutory prohibitions has steadily eroded. We need not concern ourselves with this distinction because the only issue here is whether the Madrid disk drive is “two separate and distinct products . . . tied into a single package.” Id. at 1214 .

“[Antitrust decisions and literature contain astonishingly little discussion of the criteria to be applied to distinguish between component parts of a single product and a multiplicity of products.” N. W. Controls, Inc. v. Outboard Marine Corp., 333 F.Supp. 493, 501 (D.Del.1971). The Ninth Circuit has come as close as any court to setting out a workable standard in this area. In Siegel v. Chicken Delight, Inc., 448 F.2d 43, 47 (9th Cir. 1971), cert. denied, 405 U.S. 955 , 92 S.Ct. 1172 , 31 L.Ed.2d 232 (1972), the Ninth Circuit said:

The court supplemented this general statement with a specific standard which it indicated would be applicable in a case such as this:

- Id. at 48.

Thus, the court enumerated three “criteria to be applied to distinguish between component parts of a single product and a multiplicity of products.” N. W. Controls, Inc. v. Outboard Marine Corp., supra, 333 F.Supp. at 501 .

In Moore v. Jas. H. Matthews & Co., supra, the Ninth Circuit had the opportunity to apply this standard. In that case plaintiffs were in the retail grave memorial business and also operated an installation service for grave markers. They sued a number of cemeteries, alleging that it was illegal to tie the purchase of a cemetery lot with the requirements that purchasers of markers buy the memorial from or through the cemetery and use the cemetery’s installation service. Reversing the District Court’s judgment for defendants, the court said:

The court concluded that “consideration of the ‘function of the aggregation’ leads inescapably to the conclusion that separate products and services are involved in each of the tying arrangements before us.” Id.

The court must, therefore, review Memorex’s evidence with an eye to the “function of the [Madrid] aggregation.” It must also consider whether integration of the head/disk assembly into the drive unit resulted in cost savings, and whether the head/disk assembly and the drive unit are normally sold or used as a unit with fixed proportions. However, before undertaking this evaluation, the court feels the need to recount some of the history leading up to the Madrid disk drive.

Storage of information is one of the vital functions of an electronic data processing system. In modern computers, this function is performed by devices peripheral to the central processing unit which are known either as tape drives or disk drives depending on the storage medium. For reasons which should become clear below, they are also described as serial access storage devices and direct access storage devices, respectively.

Tape drives were developed first and their operation is similar to conventional audio tape recorders. Information is recorded sequentially on the tape and may only be accessed in the same manner. Thus, the access time for data located at opposite ends of a reel of tape is relatively slow. Disk drives were developed in response to this shortcoming of serial access storage devices. The disks in a disk drive resemble a stack of phonograph records. Information is recorded randomly on the disks and may be accessed directly in much the same manner as the cuing mechanism on a stereo turntable makes it possible to set the needle down anywhere on a record. Access times for direct access storage devices are independent of the location of the data and are very fast.

The disk drive was invented by IBM in 1956. The first disk drive contained a number of large disks which could not be removed without destroying the information that was recorded on them. This device was a major breakthrough, but its limited storage capacity minimized its impact.

To accommodate the growing storage needs of users with the disk technology of the early 1960’s, IBM introduced the customer-removable disk pack. A disk pack contains several smaller disks which can be removed from the drive unit and stored in a plastic case. By increasing the number of disk packs per drive, users are able to increase the effective storage capacity of their drives. Disk packs are completely interchangeable between compatible drive units, and since their introduction, storage capacity has increased from 7.5 megabytes per pack to 200 megabytes per pack.

IBM introduced another new concept with its Winchester disk drive. The disk pack was replaced by a more elaborate mechanism called a data module. The data module was still customer removable, but some of the equipment that had originally been part of the drive unit was combined with the disks and the whole arrangement was enclosed in a protective plastic shell. One of the parts moved into the data module was the heads so that the same heads that recorded information would read it. Thus, it was no longer necessary for the heads on all compatible drive units to be aligned periodically to the same tolerance. The larger of the two versions of the data module had a storage capacity of 70 megabytes. While this capacity is considerably smaller than the larger disk packs, it is not unusual for technological advances to be tested initially on a limited basis.

The Madrid disk drive followed after the Winchester. It did not have a customer-removable disk pack or data module. In their place was something called a head/disk assembly (HDA). The HDA can be changed by a trained field engineer, but to run the proper diagnostic tests takes approximately one hour. The HDA includes even more of the equipment that was originally part of the drive unit than the data module. The HDA contains a greater number of disks than the data module and is considerably heavier, but its storage capacity of 317.5 megabytes is significantly larger than anything that came before it.

With these preliminary matters out of the way, the court is now in a position to review Memorex’s evidence on the Madrid claim in the light of both the directed verdict standard set forth in Chisholm Brothers Farm Equipment Co. v. International Harvester Co., supra, and the tie-in standard set forth in Siegel v. Chicken Delight, Inc., supra.

Under the Siegel standard, the court must first look to the “function of the [Madrid] aggregation.” As indicated above, the initial disk drive with its integrated disks had a very limited storage capacity. Customer-removable disk packs enabled users to increase the effective storage capacity of their disk drives, but with disk packs they lost the ability to keep all their data on line. As the storage capacity of disk packs increased, fewer packs were being used per drive so that a number of customers were paying for the removability feature and not taking advantage of it. The HDA is not customer removable. However, the Madrid “aggregation” offers users a significantly larger on-line storage capacity than has previously been available. It satisfied a recognized customer need, and regardless of how it is marketed, the HDA and the Madrid disk drive were designed to be and will be used as a unit.

The “function of the [Madrid] aggregation” is, therefore, to provide users with a very large storage capacity that is permanently on line. The fact that it is possible for a field engineer to remove an HDA from one Madrid disk drive and install it in another without destroying the information that is recorded on it does not undermine this conclusion. Such a transfer would only be made in an emergency situation and goes to the maintenance and reliability of the “aggregation” not to its “function.” It is also not important that users are able to upgrade their operations by having a fixed-head HDA installed. This change requires the assistance of a field engineer as well.

The second factor the court must consider is whether the integration of the HDA into the Madrid disk drive resulted in cost savings apart from those reductions in sales expenses and the like normally attendant upon any tie-in. The data module which was used on the Winchester disk drive had an automatic loading mechanism so that all the user had to do was set the module in the drive, push a button, and wait a short time for a ready light to come on. This mechanism was not necessary on the HDA because the HDA is not customer removable. Elimination of these parts resulted in cost savings to IBM some of which were passed on to customers in the form of a lower cost per megabyte of storage.

The third factor the court must consider under the Siegel standard is whether the HDA and the Madrid disk drive are normally sold or used as a unit with fixed proportions. It is already clear that these two items were designed to be and normally will be used as a unit, and IBM concedes that they are always sold as a unit. This situation contrasts sharply with the customer-removable disk packs and data modules which were neither sold nor used as a unit with the disk drives on which they operated.

In addition, other equipment manufacturers who developed disk subassemblies that were not customer removable, including Memorex, sold the subassembly and the drive on which it operated as a unit for a single price. Whether industry practice is considered to be subsumed under the third factor above, Siegel v. Chicken Delight, Inc., supra, 448 F.2d at 48, n.4 , or an independent factor, N. W. Controls, Inc. v. Outboard Marine Corp., supra, 333 F.Supp. at 501 , it is clear that the practice in this industry is to sell integrated disks and the drives on which they operate for a single price.

Upon reviewing the evidence Memorex has presented in the light of the Siegel tie-in standard, the court concludes that the HDA and the Madrid drive unit are a single product, and that no reasonable jury could find otherwise. Indeed, Memorex itself has as much as conceded this fact in some of its product literature. In its 601 OEM Disk Storage Drive Introduction it said:

Memorex undertook to develop the 601 or Maverick product prior to IBM’s announcement of the Madrid program. However, it was obviously aware of the Madrid program when this brochure was written. On page 10 of the brochure, the Maverick disks are described as fixed (non-removable) and it is emphasized that because the disks are included in the price of the drive, there are no packs or cartridges to buy. Finally, on page 11, Memorex stresses that the sealed deck plate assembly makes for easy maintenance and allows for field upgradeability.

The Maverick product was sold to original equipment manufacturers. However, Memorex’s Madrid-type product, which competes in the end-user market against IBM’s Madrid, is also sold for a single price. Memorex indicated to customers that separate prices would be quoted if requested, but it would have been more consistent with its position in this case to have offered the disk subassembly and the disk drive at separate prices as a rule rather than as the exception.

It appears to the court that what Memorex is really attempting to do here is to hold the line against progress in disk storage technology. The limited storage capacity of the first disk drive was responsible for the growth of separate markets for customer-removable disk products and drive units, but there was no guarantee that these dual markets would not merge again as storage capacities increased. If IBM had simply bolted a disk pack or data module into a drive and sold the two items as a unit for a single price, the “aggregation” would clearly have been an illegal tying arrangement. Instead, IBM introduced a new product with the largest permanent on-line storage capacity ever offered. Just as buggy whip manufacturers had to face the fact that the invention of the automobile meant the end of their market as they had known it, this new disk storage technology will have a serious impact on the market for customer-removable disk products. However, the antitrust laws were not designed to insure the maintenance of the status quo for any competitor.

The court concludes that the HDA is a component part in the Madrid disk drive. While it would be possible for IBM to sell the HDA for a separate price from the rest of the drive unit, just as it would be possible to sell many of the other components separately, IBM is not required to do so by Section 1 of the Sherman Act or Section 3 of the Clayton Act.

After reviewing Memorex’s evidence, the court is convinced that the Madrid disk drive is a single product and that no reasonable jury could come to any other conclusion. The evidence shows that the Madrid head/disk assembly was not customer removable. As a consequence, IBM was able to use new technology to dramatically increase the storage capacity of disk storage devices. Integration of the disks into the drive resulted in lower manufacturing costs and lower costs per megabyte of storage to users. Customers wanted a large permanent on-line storage capacity, and there was a practice in the industry to offer such devices as a unit for a single price.

Because the court has concluded that the Madrid disk drive is a single product, IBM’s intent is irrelevant. Good intentions will not change two products into one, and likewise, a single product does not become separate and distinct products because of a malevolent intent. For the reasons indicated above, the court hereby grants IBM’s motion for a directed verdict on the tie-in aspect of Memorex’s Madrid claim.