“In requiring that Apple show that a large-vocabulary model—rather than a smaller abbreviated model—would have fit on the integrated circuit in the prior art, the Board committed legal error.” – CAFC The U.S. Court of Appeals for the Federal Circuit (CAFC) on Tuesday vacated and remanded a decision for a patent owner against Apple, concluding that the Patent Trial and Appeal Board (PTAB) made several errors in finding the claims of the speech recognition patent at issue were not shown to be unpatentable. Zentian Ltd. owns U.S. Patent No. 10,839,789, which is titled “Speech recognition circuit and method.” Apple petitioned for inter partes review (IPR) of multiple claims of the patent, arguing in relevant part that “claim 1 was obvious over prior art reference U.S. Patent No. 5,819,222 (“Smyth”) or the combination of Smyth and U.S. Patent No. 6,832,194 (“Mozer”) and that claim 29 was obvious over Smyth, Mozer, and several other references.” The dispute centered on the “acoustic model memory” and “integrated circuit” limitations of claims 1 and 29. Both claims 1 and 29 require “a single integrated circuit.” Claim 1 requires that the integrated circuit contain both the claimed acoustic model memory and the claimed “calculating apparatus” (i.e., the processor that calculates distances between feature vectors and the states of the acoustic model).” While Apple argued that Smyth or the combination of Smyth and Mozer “disclosed the acoustic model memory on an integrated circuit with a calculating apparatus,” the PTAB ultimately determined that Apple had failed to show the claims unpatentable, largely due to the fact that the Board implicitly construed the claimed acoustic model memory “to require holding a large-vocabulary model.” The Board found Apple had failed to show Smyth alone disclosed the claimed acoustic model memory on an integrated circuit because the claimed processor, (the Motorola