SGI v. AMD: Chief Judge Rader on Claim Construction

Silicon Graphics, Inc. (SGI) v. ATI Technologies, Inc. and Advanced Micro Devices (AMD) (Fed. Cir. 2010)

In his first opinion as Chief, Judge Rader provided the parties with a lesson in claim construction. This decision is a case-in-point for the high value of professional patent drafting because the decision turns on the esoterics of claim construction. This decision also highlights the primacy of claim construction in the infringement analysis. Namely, the question of whether ATI chips infringe was boiled-down into an interpretation of the claim language.

SGI’s patent is directed to graphics processing hardware that uses floating point calculations.

Meaning of “A”: The asserted patent claims include “a rasterization circuit . . . that rasterizes the primitive according to a rasterization process which operates on a floating point format.” Although the accused ATI chips do use floating point numbers, the ATI rasterization process also uses fixed-point numbers. The district court held that the ATI chips could not infringe because their rasterization process does not operate in floating point format “as a whole.”

On appeal, the Federal Circuit re-parsed the claim language and rejected the district court’s construction as erroneous. Rather, the fact that a circuit rasterizes using fixed point format does not preclude the circuit from also using “a rasterization process which operates on a floating point format” as required by the specification.

The use of the indefinite article “a” in the claim, when coupled with the list of processes provided in the specification, makes it clear that the claims’ references to “a rasterization process” means “one or more rasterization processes.”

The limitation “a rasterization process which operates on a floating point format” therefore means that “one or more of the rasterization processes (e.g., scan conversion, color, texture, fog, shading) operate on a floating point format.” This construction is also in line with the rest of the specification. Nowhere does the specification teach that all rasterization processes must operate on a floating point format.

Summary of the Invention: The claim also requires that the rasterization circuit perform “scan conversion.” The district court required that the scan conversion be accomplished entirely using floating point numbers. On appeal, the Federal Circuit agreed with that interpretation. In making its decision to limit the breadth of the term, the court looked to a statement from the Summary of the Invention that: “the scan conversion process is now handled entirely on a floating point basis” and another statement from the specification that “this rasterization process is performed exclusively in a floating point format.” In several places, the specification indicated that various operations could be performed using fixed point numbers. However, the court held those statements insufficient because they were never specifically directed to the scan conversion process. “Thus general language in the specification permitting some operations to be done in fixed point does not work to contradict the specific language that requires scan conversion in floating point.”