Microprocessor Enhancement Corporation (MEC) v. Texas Instruments (TI) and Intel (CAFC 2007-1249)

'593 claimed pipelined processing methods, accelerating conditional processing based upon instruction sequence dependencies.

Pipelined processors, however, operate like assembly lines, where the processor is subdivided into segments, each of which simultaneously completes its respective task on a different instruction. Encyclopedia of Computer Science and Engineering 1143 (Anthony Ralston ed., 2d ed. 1983); David A. Patterson & John L. Hennessy, Computer Architecture a Quantitative Approach 251 (1990). A pipelined processor is thus analogous to an assembly line designed to fetch a new instruction from memory before the previous instruction is completed and written to memory. For a linear set of instructions (i.e., a set of instructions that are neither branched nor conditional, discussed infra), a pipelined processor operates at maximum efficiency where one instruction is completed and one instruction is fetched on every clock cycle once the pipeline is full.

In order to operate in a useful fashion, programs often require the use of nonlinear instructions, i.e., instructions containing a branch or discontinuity in the instructional sequence, that result in "dependencies" among the individual instructions of an instruction set. Control dependencies occur, for example, when an instruction cannot be executed until the result of a prior conditional branch instruction is known. That is, a conditional instruction may specify that subsequent instructions are to be fetched and executed out of sequence, depending on whether a particular condition is satisfied. '593 patent col.2 ll.30-35.

MEC fended off district court confusion about mixing claim types, allowing the Judge Gajarsa on the CAFC panel to wax Shakespearean. Nice touch. A method claim specific to a particular structure does not mix claim types.

A single patent may include claims directed to one or more of the classes of patentable subject matter, but no single claim may cover more than one subject matter class. IPXL Holdings, 430 F.3d at 1384 (holding indefinite a claim covering both an apparatus and a method of using that apparatus). Applying this rule, the district court concluded that although claim 1 purported to claim a method of executing instructions in a pipelined processor, the structural limitations of the pipelined processor evidence an intent to claim the apparatus as well.

The drafting structure of claim 1 may be generally described as follows:

1. A method of executing instructions in a pipelined processor comprising:
[structural limitations of the pipelined processor];
the method further comprising:
[method steps implemented in the pipelined processor].

See '593 patent col.129 l.27 to col.130 l.32. Although this seeming preamble within a preamble structure is unconventional, its effect on the definiteness of claim 1 lacks the conclusiveness with which King Claudius's guilt is established by his reaction to Hamlet's play within a play. See William Shakespeare, Hamlet act 3, sc. 2. Method claim preambles often recite the physical structures of a system in which the claimed method is practiced, and claim 1 is no different. The conclusion of IPXL Holdings was based on the lack of clarity as to when the mixed subject matter claim would be infringed. 430 F.3d at 1384 ("[I]t is unclear whether infringement of claim 25 occurs when one creates a system that allows the user to [practice the claimed method step], or whether infringement occurs when the user actually [practices the method step]."). There is no similar ambiguity in claim 1 of the '593 patent. Direct infringement of claim 1 is clearly limited to practicing the claimed method in a pipelined processor possessing the requisite structure.

The district court applied the same interpretation to numerous instances of the term "condition code" in the claims, rendering the claim nonsensical to the court, leading to ruling the claims indefinite. The appeals court reminded that calling a claim gibberish by construction is an absolute last resort. The same claim term may be interpreted in context, not rigidly uniform.

Turning to claim 1 and claim 7 of the '593 patent, we note that "[a] claim that is amenable to construction is not invalid on the ground of indefiniteness" if the construction renders the claim definite. Energizer Holdings, 435 F.3d at 1371. Unlike the claim at issue in Process Control, "condition code" as used in claims 1 and 7 is not surrounded by uniform language that requires a single interpretation of the term. Cf. Epcon Gas Sys., 279 F.3d at 1031 (construing "substantially" as having two different meanings based on its use in "two contexts with a subtle but significant difference"). Rather, the appropriate meaning of "condition code" is readily apparent from each occurrence in context, and TI's expert, Dr. Patt, indicated that the '593 patent used condition code to refer to a value or a storage location based on its context within the claims. Indeed, the claims' apparent nonsensical reading under a uniform construction of "condition code" is indicative of the ease of determining the appropriate meaning of each use of the term from its context. For these reasons, the use of "condition code" in claim 1 and claim 7 does not render these claims indefinite.

Construing "pipeline stage," whether it meant position or timing, sent the CAFC traipsing through the Phillips v. AWH priority stack. MEC was inconsistent in its construction, arguing "pipeline stage" meant both position and timing, depending upon usage: positional when unmodified, temporal when modified.

Claim terms must be given "the meaning that the term would have to a person of ordinary skill in the art in question at the time of the invention." Phillips, 415 F.3d at 1313. This court ascertains the meaning of a disputed term by looking to "'those sources available to the public that show what a person of skill in the art would have understood disputed claim language to mean.'" Id. at 1314 (quoting Innova/Pure Water, Inc. v. Safari Water Filtration Sys., Inc., 381 F.3d 1111, 1116 (Fed. Cir. 2004)). "'Those sources include the words of the claims themselves, the remainder of the specification, the prosecution history, and extrinsic evidence . . . .'" Id. Phillips teaches that these sources should be accorded relative weights in the order listed, with the words of the claims themselves being the most relevant. Id. at 1314-19. Accordingly, we discuss each source of meaning of the claim term "the pipeline stage" in this order.

The specification was unclear, the prosecution history a mixed bag. A textbook relied upon by all concerned during litigation was dispositive that "pipeline stage" was temporal.

[H]aving thoroughly examined "'the indisputable public records consisting of the claims, the specification and the prosecution history,'" Phillips, 415 F.3d at 1319 (quoting Southwall Techs., Inc. v. Cardinal IG Co., 54 F.3d 1570, 1578 (Fed. Cir. 1995)), we find that the extrinsic evidence supports the conclusion that the inventor used "the pipeline stage" in its temporal sense, consistent with the term's ordinary meaning in the computer arts. For example, David A. Patterson & John L. Hennessy, Computer Architecture a Quantitative Approach 251 (1990), relied on by all three parties on appeal, discusses the concept of a pipe stage or pipe segment, and the discussion is framed by references to time and clock cycles rather than positions.

With that, noninfringement.