Intel Corp. v. Advanced Micro Devices, Inc., 756 F. Supp. 1292 (N.D. Cal. 1991)

U.S. District Court for the Northern District of California - 756 F. Supp. 1292 (N.D. Cal. 1991)
February 28, 1991

756 F. Supp. 1292 (1991)

And Related Counterclaims.

No. C-90-20571-WAI.

United States District Court, N.D. California.

February 28, 1991.

Lois W. Abraham, Brown & Bain, Palo Alto, Cal., and F. Thomas Dunlap, Jr., Intel Corp., Santa Clara, Cal., for plaintiff.

Douglas K. Derwin, Skjerven, Morrill, MacPherson, Franklin & Friel, San Jose, Cal., and Richard H. Lovgren, Advanced Micro Devices, Inc., Sunnyvale, Cal., for defendant.


INGRAM, District Judge.


This application for permanent injunction, and for finding of validity and infringement of plaintiff's asserted trademark is one of several phases of the trial of the above-entitled action. This first phase came duly on for trial before the court on January 7, 1991 and proceeded for sixteen trial days, and was submitted for decision on February 7, 1991. Following submission, both sides filed memoranda and briefing on discrete issues. I have carefully considered those submissions.

This Memorandum of Intended Decision constitutes Findings of Fact and Conclusions of Law consistent with the provisions of Fed.R.Civ.P. 52(a).

Inasmuch as this is a Memorandum of Intended Decision, and is decisive of only some of the issues included in the above-entitled action, I will entertain argument on objections to findings encompassed herein prior to the entry of any judgment herein, to be conducted at a time mutually convenient to counsel and to me.

This decision is dispositive of this first phase of the case. However, mindful that *1293 this disposition may not survive appellate scrutiny, and that an appellate court may find that a group composed of end users or some hybrid group constitutes the relevant public, I intend to file a subsequent memorandum disposition which will assume arguendo that the relevant market group deemed appropriate by plaintiff, i.e. business affiliated end users, constitutes the relevant public.

I do not at this time anticipate that the subsequent memorandum disposition will alter the end result of the instant decision, but it will contain an analysis of those surveys which deal with the end user group.

Counsel will please arrange with the Clerk of this Court an acceptable schedule for the preparation and trial of the second phase of this action.


Plaintiff is not entitled to the injunctive relief which it seeks by this application.

The combination 386 is generic under the evidence. Plaintiff has failed to prove by a preponderance of the evidence that the combination 386 is not generic.

The appropriate group by which to measure consumer perception of the combination 386 is one composed of Original Equipment Manufacturers (OEMs) who purchase microprocessors for inclusion as component parts in their manufactured products.


Plaintiff has the burden of proof on all issues presented by this phase of the above-entitled action, and the applicable burden is a preponderance of the evidence.


Each side advocates a different group of "purchasers" who constitute the appropriate group for purposes of measuring purchaser perception of the 386 combination. The group, if either, which I find to be appropriate will constitute the sine qua non for defining relevant evidence as to purchaser perception on the issues of the purported genericness of the 386 combination and/or whether that combination has achieved a secondary meaning or whether use of the accused mark will cause confusion or likelihood of confusion.

Plaintiff contends that the appropriate group should be comprised of an "end user" group. Plaintiff limits that group to persons who are users of personal computers which contain as a component part a 386 microprocessor. Plaintiff defines those users as persons who use personal computers in the course of their work in a business environment and for business purposes, and/or are MIS managers or data processing managers within businesses which utilize personal computers. This user/managerial group must also, in order to be a part of plaintiff's proffered group, have some part in the selection and purchase of personal computers for use within the employing business. One of the surveys offered by plaintiff required that the members of this group must also be employed at a site at which at least 100 persons are employed, and another specified a minimum number or more personal computers at the surveyed site.

Defendant contends that the appropriate group consists of Original Equipment Manufacturers (OEMs) who constitute the direct purchasers of microprocessors from the manufacturers thereof. This contention is made because, says defendant, microprocessors are rarely or never sold to end users of personal computers. They are, and have always been, sold to OEMs who then incorporate them into the construction of personal computers and other products which are in turn sold to end users. Testimonial support which reflects how these microprocessors pass in commerce is found at these references, among others: 4-100:21-101:3; 4-571:6-8; 6-1110:13-1111:13; 11-2069:10-2070:12; 2-194:16-18; 3-408:20-21.

Defendant contends also, with merit, that the facts presented in this case indicate that the identity of the microprocessor contained in a personal computer is in no way perceived by an end user purchaser, short of dismantling the computer. Normally, the microprocessor is neither seen, touched nor physically perceived in any manner. It *1294 is a part of an article sold in a different product, and is only purchased by the ultimate user in the sense that every part of the personal computer which makes up the whole may be said to be purchased.

The purchase of the microprocessor as a discrete product is only accomplished as between the manufacturer of the microprocessor, such as plaintiff and defendant, and the manufacturer of the personal computer.

There is little dispute in this case as to the manner in which microprocessors are sold by their manufacturers. They are not generally, if at all, sold to other than OEMs who then utilize them for some purpose of their own.

OEMs are the "usual buyers" of microprocessors.

Plaintiff argues that the relevant market is comprised of end users of personal computers. There is testimonial agreement with that contention. 4-601:22-602:1; 5-726:9-18.

In support of its contention, plaintiff cites a number of cases which are "single product" cases,[1] meaning that the product as originally manufactured reached the end user in precisely the form in which it left the manufacturer, and not as a component part of some other product. Those cases are not very helpful in reaching a solution as to relevant market in this case.

Plaintiff has cited two cases which it characterizes as component cases, similar to this case. Thomas Pride Mills, Inc. v. Monsanto Co., 155 USPQ 205 (N.D.Ga. 1967), is a suit by a carpet maker against the maker of synthetic fibers used in carpets to cancel the latter's registered mark on the ground of its genericness, and to preliminarily enjoin continued enforcement of that registered mark. Plaintiff in that case offered in evidence a telephone survey. The court noted the survey, noted that its admissability and probative value would be determined at trial, and said nothing more about it. The case is no help.

The other cited case, E.I. Dupont de Nemours & Co. v. Yoshida International, Inc., 393 F. Supp. 502 (E.D.N.Y.1975) is so factually distinguishable that its precedent value in this case is unacceptable to me. Plaintiff in that case is a manufacturer of a TFE resin substance bearing the registered mark Teflon. The substance is a non-stick, lubricous component of cookware and other products not manufactured by plaintiff. Defendant is a foreign manufacturer of zippers called Eflon, which characteristically worked smoothly and well, but without the addition of the component additive Teflon. Plaintiff, as an aid to the marketing of the products of its manufacturer customers permitted the use of its registered mark, and in many cases issued to such customers a certificate of high quality, which was incorporated into the advertising of such customers and reference to which appeared on the face of the products in many cases. In bringing the action for trademark infringement plaintiff purported to act in behalf of its customers. The court recognized that paragraph 15 of the complaint alleges as much. The court recognized that for purposes of measuring proximity of products in an infringement action, the certification mark owner acts as a representative of certification mark users in the prosecution of an action for infringement of a trademark.

The case is further distinguishable because it is essentially a single product case. The product whose name was accused was a single product not enhanced by the use of the component Teflon. The pots and pans which formed the basis for questions in the "Teflon Survey" used in that case were single products whose utility was enhanced by the component Teflon, but whose essential identity as pots and pans was always the same, and whose relevant market was always the same.

Thus, the relevant market of the customer of pots and pans is the appropriate group for measurement purposes in that case, and in my view the case cannot stand for the proposition that the ultimate consumer *1295 constitutes the relevant market in a component case. This view is not altered in any way by the existence in the case of surveys, proffered by both sides, which surveyed groups of persons other than the immediate purchasers from plaintiff of the product Teflon. These surveys were conducted on the issue of the asserted genericness of the mark Teflon, and it does not appear from the opinion that there was any justiciable controversy over the consist of the relevant market.

In the course of the trial defendant has cited to me a number of cases[2] which it contends support its view that the relevant market in this instance is the OEM market, because the only direct purchasers or prospective purchasers of the component 386 microprocessors are OEMs, and not a market or group comprised of a nonpurchasing segment of the public. Cf. Blisscraft of Hollywood v. United Plastics Co., 294 F.2d 694 (2nd Cir. 1961).

Each of the cases cited by defendant defines as the relevant perception that of purchasers and prospective purchasers.

On the basis of the applicable authority, I find that the relevant market or appropriate measuring group in this case is the OEM group contended for by defendant. The reason for my finding is that there is no authority proffered supporting the contention that in a component part case such as this, the relevant market consists of purchasers of the product containing the component which bears the asserted trademark the validity of which is in issue.

Where there is a single product, complete in itself, it is easy to see why the ultimate consumer constitutes the relevant market. All other hands through which the product passes, wholesalers, distributors, retailers are commercial conduits which have a part in delivering the product to the user. That is the only role they play. They add nothing to the product; they do not constitute a market for the creation of a different product, they in no way consume the product.

In contrast, the OEM direct purchasers in this case are users of the microprocessors purchased by them. They have a part in the consumption[3] of the microprocessor, by using it as an ingredient of another product, whether it be a microprocessor, a factory automation product or a surgical instrument. Therefore, it seems to me that the relevant perceptions are those of the OEM buyers, who enter the microprocessor market as buyers to satisfy their own economic need for component parts for goods of their own production.

Plaintiff is attempting to convert the target market of an advertising campaign into a judicially cognizable relevant market for trademark purposes. Plaintiff might be successful in doing that if plaintiff could prove that by reason of its "pull through" advertising campaign or other market forces the demand for personal computers and/or other products containing 386 microprocessors is so great that the end users of personal computers are the de facto direct purchasers of OEM purchasing. Plaintiff has not made such a showing here, and I am satisfied that only purchasers constitute the relevant market.


Plaintiff has the burden of proof in an unregistered mark case on the issue of genericness of the asserted mark in issue. A.J. Canfield Co. v. Honickman, 808 F.2d 291 (3rd Cir.1986), 1 McCarthy, Trademarks and Unfair Competition, Second Ed.1984 § 12:2.

The presumption of non-genericness of a product name where the name was not in common usage prior to its adoption by the holder, discussed in Murphy Door Bed Co. v. Interior Sleep Systems, Inc., 874 F.2d *1296 95, 100-101 (2d Cir.1989) does not apply here. The 386 designation is the most recent in a series of related numerical designations used for years by Intel and the industry generally. These designations are commonly descriptive of the 86 family microprocessor family to which they are attached. Intel adopted the unregistered 386 designation over 5 years after it began marketing the 80386 microprocessor, and over 2½ years after it began selling the product. The seriatim numbers of the 86 family were in common currency among the initiated public prior to the claiming of the unregistered mark by Intel in 1988, and the 386 designation was not "expropriated" by the public in the sense illustrated in Murphy.

In order to prove trademark significance or non-genericness, plaintiff must prove that the primary significance of the term in issue in the minds of the relevant consuming public is not the product but the producer. Kellogg Co. v. National Biscuit Co., 305 U.S. 111, 59 S. Ct. 109, 83 L. Ed. 73 (1938); Bayer Co. v. United Drug Co., 272 F. 505 (S.D.N.Y.1921).

Within the context of this case the question boils down to whether 386 is a brand name or a generic name.

Plaintiff has never treated any number combination designating a microprocessor as a trademark except for the 386, thus the microprocessors numbered 8086, 8088, 80186, 80188 and 80286 have not been treated as trademarks by plaintiff and have been widely used by third parties. 1-108:10-15; 5-891:19-892:2, 893:22-24, 895:5-7, 896:8-9. The term 386 was already in use as a nickname for the 80386 and was used extensively in the trade press before it was selected by plaintiff as a trademark. 5-902:15-22.

The Walker Survey, Ex. CQ, was conducted by Walker Research and Analysis commencing on December 17, 1990. The survey was designed and administered by Walker to demonstrate whether 386 is a brand name or a generic description or common term. The method of conducting the survey consisted of computer assisted telephone interviews. The sample list used in the survey was obtained from Dun and Bradstreet and consisted of a sample of the Dun and Bradstreet database of OEMs. The sample was of representative persons in charge of purchasing microprocessor components for inclusion in equipment manufactured by their OEM companies who on the basis of exemplary questions demonstrated an ability to distinguish between a brand name and what the survey questionnaire refers to as a generic description. The questionnaire is essentially modelled on the one used and approved in Dupont, supra, and described in that case as the "Survey B". The Walker survey is the only one of the several in evidence which surveys the proper group, i.e. OEMs.

Plaintiff has requested that I make certain findings to the effect that the Walker Survey is misleading because:

1. It surveyed an irrelevant target market.
2. Mr. Gordon, who supervised the survey abdicated his responsibilities and operated under the direction of defendant's attorneys.
3. The term "generic description" is ambiguous.
4. The examples used to distinguish between brand names and generic description suggest that a combination of numbers alone is not a brand name.
5. Respondents were led to believe that numbers are "descriptions" and not brand names.

On these topics, I find that the OEM group is appropriate because OEMs are virtually the only purchasers of microprocessors. The evidence in the case has established that the market for microprocessors, including 386 microprocessors is broader than personal computer use, and that the market for 386 microprocessors spreads broadly with the passage of time and with the advent of newer technology such as the 486 and RISK technology. The target market of the Walker Survey was not irrelevant.

I find it difficult to criticize survey managers on either side of this case who do not *1297 use independent judgment in defining the purpose and target group of a litigation survey. Unlike a marketing survey, usually in the hands of advertising agencies or marketing specialists, the context of litigation surveys to a large extent must necessarily be directed by lawyers who are familiar with the litigation at hand. There has been extensive lawyer participation on both sides in this case.

The term "generic description" may be ambiguous as Dr. Sorensen said, but Dr. Sorensen did not disagree with the manner in which the respondents distinguished between brand names and generic descriptions. 14-2660:24-2663:11.

The objections predicated on a suggestion that numbers alone cannot be a brand name and that respondents were led to believe that they were descriptions and not brand names are quibbles. There is no evidence other than that of Dr. Sorensen and his testimony is not persuasive to me on those issues.

Question 8 of the Questionnaire is the key question and asked each respondent "Would you say ______ is the brand name or a generic description?" The question was asked with respect to each of five separate terms, one of which was 386. The order of the terms was rotated, so that no term kept its original position in the order of questioning at all times.

The result of the survey was that of 400 total interviews 72% of respondents stated that 386 is a generic description, 21% of respondents stated that 386 is a brand name, and 7% of respondents didn't know whether 386 is a generic description or a brand name.

If a properly conducted survey shows that 72% of the relevant public regards a combination as generic, then that is its principal significance, and a court is justified in according substantial weight to such a survey. King-Seeley Thermos Co. v. Aladdin Industries, Inc., 321 F.2d 577 (2nd Cir.1963); Anheuser-Busch, Inc. v. Stroh Brewery Co., 750 F.2d 631, 639 (8th Cir.1984); Dupont, supra.

The Walker Survey is an important piece of evidence because it was concerned with OEMs, the group that I have found to constitute the relevant market. It is the only survey in evidence which is directed at that group. Plaintiff has not undertaken a survey of the OEM market.

I find that the Walker Survey was competently conducted, taken of a sample of the relevant market, is projectable and is uncontradicted.

The evidence shows that within plaintiff corporation's organization in 1988, and thereafter, there was a consciousness that the combination 386 was being perceived by others as a product description and not as a brand name.

Ms. Tsuyemura, plaintiff's Trademark Program Manager informed people both inside and outside plaintiff's organization structure that the designation of the part number 80386 was a generic term. 12-2172:16-21; Ex. BA; Ex. BB. Prior to 1988 the chief executive officer of plaintiff corporation was concerned over the fact that the 386 combination was in use in the advertising of computer manufacturers without attribution to plaintiff in any way. Ex. DE, 5-955:7-956:3. There were references within plaintiff's corporate organization to a "take back the 386 name campaign." 6-1158:3-1159:7; Ex. BK; Ex. BL. Although many of plaintiff's witnesses denied the existence of such a campaign, or referred to it as "hallway jargon" I do not find these denials wholly credible in the circumstances.

While these evidentiary items do not have great significance in themselves, they do illustrate a conscious apprehension within plaintiff's organization that the combination 386 and the part number 80386 were by 1988, and perhaps before, generic names, at least among users of personal computers. The Benchmark Survey, Ex. 38, and its results have primary reference to a business end user group, as distinguished from an OEM group, and will therefore be discussed in my subsequent opinion.

Industry practice for a number of years had been to refer to microprocessors in the family lineage of the 386 (Ex. FR) by a *1298 contraction or diminutive for the microprocessor part numbers; i.e. "286" for 80286, "386" for 80386. 8-1414:10-21. The practice arose out of the custom of "second source" manufacturers in the naming of their products. 8-1405:4-22. The combination 386 and the part number 80386 are historically used and presently interchangeable among OEMs and do not connote a particular vendor but are used generically among OEMs. 8-1419:10-17.

Plaintiff has not proven by a preponderance of the evidence that the combination 386 is not generic. To the contrary the evidence, primarily the Walker Survey, shows that it is generic in the relevant OEM market.


The combination 386 is not entitled to trademark protection, and in consequence there can be no infringement thereof.

McCarthy, supra, § 12.1
Surgicenters of America, Inc. v. Medical Dental Surgeries Co., 601 F.2d 1011 (9th Cir.1979).

Plaintiff's application for a permanent injunction is denied.


[1] Surgicenters of America, Inc. v. Medical Dental Surgeries Co., 601 F.2d 1011 (9th Cir.1979); Stix Props, Inc. v. United Merchants Mfrs., Inc., 295 F. Supp. 479 (S.D.N.Y.1968); Big "O" Tire & Rubber Co., 561 F.2d 1365 (10th Cir.1977).

[2] Bayer Co. v. United Drug Co., 272 F. 505 (S.D. N.Y.1921); Blisscraft of Hollywood v. United Plastics Co., 294 F.2d 694 (2nd Cir.1961); Application of Abcor Development Corp., 588 F.2d 811 (C.C.P.A.1978); Fuddruckers, Inc. v. Doc's B.R. Others, Inc., 826 F.2d 837 (9th Cir.1987); Charles Schwab & Co., Inc. v. Hibernia Bank, 665 F. Supp. 800 (N.D.Cal.1987); Calvin Klein Cosmetics Corp v. Lenox Laboratories, 815 F.2d 500 (8th Cir.1987).

[3] Consume means to utilize (an economic good) in the satisfaction of wants or the process of production. Webster's Third New International Dictionary, 1967.