(Santa Clara County Super. Ct. No. CV031946). The Honorable Kevin J. Murphy.
The opinion of the court was delivered by: Rushing, P.J.
CERTIFIED FOR PUBLICATION
Plaintiff Silvaco Data Systems (Silvaco) brought this action against defendant Intel Corporation (Intel) alleging that the latter had misappropriated certain trade secrets used by Silvaco in its software products. The primary gist of the claims was that Intel had used software acquired from another software concern with knowledge that Silvaco had accused that concern of incorporating source code, stolen from Silvaco, in its products. The chief question presented is whether Intel could be liable for such use if, as was effectively undisputed, it never possessed or had access to the source code but only had executable, machine-readable code compiled by its supplier from source code. We answer that question in the negative. One does not, by executing machine-readable software, "use" the underlying source code; nor does one acquire the requisite knowledge of any trade secrets embodied in that code. We also join the trial court in concluding that Silvaco failed to plead a viable claim not based upon a misappropriation of trade secrets. Accordingly, we will affirm the judgment.
Silvaco develops and markets computer applications for the electronic design automation (EDA) field, which covers the entire complex process of designing electronic circuits and systems. Among the various sub-categories of EDA software are circuit simulators, which permit the designer to create a virtual model of a proposed circuit in order to test its properties before incurring the expense and delay of manufacturing a working prototype. Defendant Intel, a major developer and manufacturer of integrated circuits, is a user and purchaser of EDA software, including circuit simulators. According to Silvaco's complaint, Intel has also developed some EDA software for its own use.
Among Silvaco's software products is SmartSpice, an analog circuit emulator. In December 2000, Silvaco filed a suit against Circuit Semantics, Inc. (CSI), a competing developer of EDA software, alleging that CSI, aided by two former Silvaco employees, had misappropriated trade secrets used in SmartSpice, and had incorporated them in its own product, DynaSpice. Silvaco eventually secured a judgment against CSI, including an injunction against the continued use of "technology" described in an exhibit attached to the judgment. It then brought actions against several purchasers of CSI software, including Intel.*fn1 It alleged that by using CSI's software, these end users had misappropriated the Silvaco trade secrets assertedly incorporated in that software. Silvaco charged Intel with misappropriation of trade secrets under the California Uniform Trade Secrets Act, Civil Code sections 3426 through 3426.11 (CUTSA), as well as with conversion, conspiracy, and violations of the unfair competition law, Business and Professions Code sections 17200 et sequitur (UCL).
Intel generally demurred to all causes of action. A series of challenges and amendments followed, culminating in the sustaining of demurrers to all of Silvaco's non-CUTSA causes of action on the ground, among others, that they were "preempted" by CUTSA.*fn2 Intel then moved for summary judgment on the CUTSA claim, arguing that (1) CUTSA defines "misappropriation" in a way requiring the plaintiff to show that the defendant possessed "knowledge of the trade secret" (Civ. Code, § 3426.1, subd. (b)),*fn3 and (2) there was no evidence that Intel ever possessed such knowledge.
In support of the motion Intel presented evidence that it never possessed the source code identified by Silvaco as constituting and containing its secrets, but only executable code supplied to it by CSI. The parties appear to agree, and we may accept for purposes of this opinion, that "source code" describes the text in which computer programs are originally written by their human authors using a high- level programming language.*fn4 (See Cypress Semiconductor Corp. v. Superior Court, supra, 163 Cal.App.4th 575, 580; Cadence Design Systems, Inc. v. Avant! Corp. (2002) 29 Cal.4th 215, 218, fn. 3.) One who possesses the source code for a program may readily ascertain its underlying design, and may directly incorporate it, or pieces of it, into another program. In order to yield a functioning computer application, however, source code must generally be translated or "compiled" into machine-readable (executable) code. After a program is compiled, it may still be represented as text, but the text is not readily intelligible to human beings, consisting of strings of binary (base 2) or hexadecimal (base 16) numbers.*fn5 For this reason, the source code for many if not most commercial software products is a secret, and may remain so despite widespread distribution of the executable program.
Intel cited discovery responses by Silvaco, and testimony by its agents and experts in this and related cases, to the effect that (1) the possession and use of DynaSpice in the form of executable object or binary code could not impart knowledge of any trade secrets embodied in the source code; and (2) Silvaco was unable to controvert Intel's evidence that it never possessed any of the source code from which DynaSpice was derived. In opposition Silvaco argued, first, that executable code incorporates the same "information" as the source code from which it is compiled, so that executing it on a computer constitutes "use" of any trade secrets reflected in the source code. Second it insisted that, quite apart from the informational content of the source code, the secrets claimed by it "include certain `methods, techniques and algorithms" that were "contain[ed]" and "use[d]" in the executable code. Third it denied the premise, which it inaccurately attributed to Intel, that liability under CUTSA requires proof "that the misappropriator comprehend the trade secret." It quoted the statement from the unpublished decision of a federal district court in Utah that "understanding of the underlying code and algorithm is unnecessary for misappropriation. There is no requirement of comprehension of the trade secret to state a claim... under the Utah Trade Secrets Act...." (ClearOne Communications, Inc. v. Chiang (D.Utah 2007) 2007 WL 4376125, *2 (ClearOne).) Finally, it suggested that by granting summary judgment the court would destroy the basis for its earlier conclusion that the non-CUTSA claims were preempted. Accordingly, it suggested, if the court was inclined to grant summary judgment it should, on its own motion, reconsider and vacate its orders sustaining Intel's demurrers.
The trial court granted Intel's motion for summary judgment, writing, "Intel has presented admissible evidence... that Silvaco has no evidence that Intel in fact acquired or possessed Silvaco's trade secrets (defined by Silvaco as source code), or that using CSI's software in executable code form somehow conveys knowledge of Silvaco trade secrets or equates to acquisition of trade secrets. [¶]... By acquiring the CSI software that `embodies' Silvaco's source code, Intel did not acquire, or gain knowledge of, the information that constitutes Silvaco's alleged trade secret.... It is not the functionality of the CSI software that constitutes Silvaco's trade alleged trade secret, but Silvaco's means of creating that functionality through the source code."
The court entered judgment for Intel. Silvaco filed this timely appeal.
I. Misappropriation of Trade Secrets
Silvaco persistently misstates the central issue in the case by insisting, here and in related appeals, that the question presented is whether a defendant charged with trade secret misappropriation "may escape liability" by establishing that it "does not comprehend the specific information comprising the trade secrets." This is not an issue, let alone the chief issue, in these matters. The posited question may be answered in the negative-as indeed it must-without resolving any aspect of this case. It is a smokescreen, a red herring, a straw man.
Intel has never suggested that it should avoid liability merely because it failed to "comprehend" Silvaco's trade secrets. Indeed it has never suggested that it would not have "comprehend[ed]" those secrets-had it ever seen them. Rather it asserts that it could not have misappropriated Silvaco's trade secrets-could not have acquired, disclosed, or used them-because it never had them. Since the evidence supporting this contention was effectively uncontroverted-there was no basis to find that Intel had access to either CSI's or Silvaco's source code-the contention presents a question of law: whether a defendant can be liable for misappropriation of a trade secret which is admittedly embodied in source code, based upon the act of executing, on his own computer, executable code allegedly tainted by the incorporation of design features wrongfully derived from the plaintiff's source code. It is undisputed that the object code executed by Intel could not disclose the underlying source code or permit the exploitation of its features and design. It could not, in short, impart knowledge of the trade secret. The question is whether, in such circumstances, Intel could be found to have misappropriated Silvaco's trade secrets. Silvaco has offered no coherent basis for an affirmative answer. To the extent such an answer is suggested by unpublished federal authorities, or authorities not decided under UTSA, we decline to follow them.
A cause of action for monetary relief under CUTSA may be said to consist of the following elements: (1) possession by the plaintiff of a trade secret; (2) the defendant's misappropriation of the trade secret, meaning its wrongful acquisition, disclosure, or use; and (3) resulting or threatened injury to the plaintiff. (3426.3; see id., §§ 3426.1, 3426.2.) The first of these elements is typically the most important, in the sense that until the content and nature of the claimed secret is ascertained, it will likely be impossible to intelligibly analyze the remaining issues.
CUTSA defines "[t]rade secret" to mean "information, including a formula, pattern, compilation, program, device, method, technique, or process, that: [¶] (1) Derives independent economic value, actual or potential, from not being generally known to the public or to other persons who can obtain economic value from its disclosure or use; and [¶] (2) Is the subject of efforts that are reasonable under the circumstances to maintain its secrecy." (§ 3426.1, subd. (d), italics added.) The sine qua non of a trade secret, then, is the plaintiff's possession of information of a type that can, at the possessor's option, be made known to others, or withheld from them, i.e., kept secret. This is the fundamental difference between a trade secret and a patent. A patent protects an idea, i.e., an invention, against appropriation by others. Trade secret law does not protect ideas as such. Indeed a trade secret may consist of something we would not ordinarily consider an idea (a conceptual datum) at all, but more a fact (an empirical datum), such as a customer's preferences, or the location of a mineral deposit. In either case, the trade secret is not the idea or fact itself, but information tending to communicate (disclose) the idea or fact to another. Trade secret law, in short, protects only the right to control the dissemination of information.
It is critical to any CUTSA cause of action-and any defense-that the information claimed to have been misappropriated be clearly identified. Accordingly, a California trade secrets plaintiff must, prior to commencing discovery, "identify the trade secret with reasonable particularity." (Code Civ. Proc., § 2019.210 (§ 2019.210); see M. Lemley, The Surprising Virtues of Treating Trade Secrets as IP Rights (2008) 61 Stan.L.Rev. 311, 344 [plaintiff should be required to "clearly define what it claims to own, rather than (as happens all too often in practice) falling back on vague hand waving"].) Pursuant to this requirement, Silvaco filed a document, under seal, designating the trade secrets claimed by it. The designation identified trade secrets in six categories, as further detailed in two attached exhibits. The first five categories, and the first exhibit, referred only to source code. The designation did not set forth the code itself, which presumably was voluminous, but referred to files that were contemporaneously produced on optical disks.
The sixth category of claimed trade secrets was described as "the... trade secrets identified in Exhibit B... and the source code implementing such trade secrets." (Italics added.) Exhibit B consists of 22 pages of technical verbiage most of which may be readily intelligible only to those within the EDA field. This much, however, seems reasonably clear: the exhibit does not designate information as such but rather describes various features, functions, and characteristics of the design and operation of Silvaco's software products. Thus the first of the 24 listed sub-categories is a "proprietary method" of carrying out a function apparently found in competing programs as well. (Italics added.) This asserted secret is also described as "a methodology for" implementing that function, apparently in an unusual way, which "contributes [to] performance and accuracy improvements." This "trade secret methodology" is "implement[ed]" by two named "modules," also described as "functions," which "represent part" of the critical "algorithm." Three "unique features" of this method are listed: The "[i]ntegration" of two other operations; a "[m]ethod" of "changing and controlling" a variable, which "affects the performance of the simulation"; and a mode of "implementation" that produces "[e]fficiency."
Silvaco's sixth category thus appears to attempt to characterize various aspects of the underlying design as trade secrets. This of course contravenes the principles discussed above. The design may constitute the basis for a trade secret, such that information concerning it could be actionably misappropriated; but it is the information-not the design itself-that must form the basis for the cause of action. And while the finished (compiled) product might have distinctive characteristics resulting from that design-such as improved performance-they cannot constitute trade secrets because they are not secret, but are evident to anyone running the finished program. Indeed, to the extent they tend to disclose the underlying design, it ceases to be a protectable secret for that same reason. The sixth category therefore fails to describe a trade secret other than source code. Since none of the other categories even purport to do so, Intel is quite correct to premise its argument on the proposition that the only trade secrets at issue are found in Silvaco's source code.
1. Disclosure; Acquisition
Under CUTSA, misappropriation of a trade secret may be achieved through three types of conduct: "[a]cquisition," "[d]isclosure," or "[u]se."*fn6 (§ 3426.1, subd. (b).) The act does not define these terms, but leaves their delineation to be adjudicated in light of the purposes and other provisions of the act. The question is whether Silvaco raised a triable issue of fact concerning Intel's commission of any of these three forms of conduct.
There is no suggestion that Intel ever disclosed Silvaco's source code to anyone, and it is difficult to see how it might have done so since there is no evidence that it ever had the source code to disclose. Silvaco emphasizes that wrongful acquisition of a trade secret may be actionable in itself. (See § 3426.1, subd. (b); San Jose Construction Inc. v. S.B.C.C., Inc. (2007) 155 Cal.App.4th 1528, 1544, fn. omitted [" `misappropriation' can occur through improper acquisition of a trade secret, not only through use"].) But there is no basis to suppose that Intel ever "acquired" the source code constituting the trade secrets. To acquire a thing is, most broadly, to "receive" or "come into possession of" it. (1 Oxford English Dict. (2d ed. 1989), p. 115.) But the term implies more than passive reception; it implies pointed conduct intended to secure dominion over the thing, i.e., "[t]o gain, obtain, or get as one's own, to gain the ownership of (by one's own exertions or qualities)." (Ibid.; see id. at p. 115 ["acquisition" as "[t]he action of obtaining or getting for oneself, or by one's own exertion"].)
One does not ordinarily "acquire" a thing inadvertently; the term implies conduct directed to that objective. The choice of that term over "receive" suggests that inadvertently coming into possession of a trade secret will not constitute acquisition. Thus one who passively receives a trade secret, but neither discloses nor uses it, would not be guilty of misappropriation. We need not decide the outer limits of acquisition as contemplated by CUTSA, however, for there is no suggestion here of acquisition even in the broadest sense, i.e., that Intel ever came into possession of the source code constituting the claimed trade secrets. Indeed Silvaco does not directly argue that Intel acquired the trade secrets at issue but only that, under the terms of the statute, it could have done so without itself having "knowledge" of them. We doubt the soundness of this suggestion, but ...