Avago Technologies International Sales PTE. Ltd. v. Netflix, Inc., No. 23-1817 (Fed. Cir. June 18, 2025)

PROST, Circuit Judge.

Avago Technologies International Sales Pte. Ltd. (“Avago”) appeals from a final written decision of the Patent Trial and Appeal Board (“PTAB”) in an inter partes review (“IPR”) determining that claims 1, 3–5, 7, 9–10, 12, 14–16, and 18–20 of U.S. Patent No. 8,646,014 (“the ’014 patent”) are unpatentable and dependent claims 6, 8, 11, and 17 are not unpatentable. J.A. 1–32. We affirm the Board as to the claims it determined are unpatentable. As to the claims the Board determined are not unpatentable, we vacate and remand for further proceedings consistent with this opinion.

BACKGROUND

The ’014 patent is titled “Multistream Video Communication With Staggered Access Points.” ’014 patent Title (capitalization normalized). The ’014 patent concerns “a system and method that provide reduced latency in a video signal processing system.” Id. at col. 1 ll. 63–65. According to the ’014 patent, latency is the “time between when a user makes a request for particular video information and when the system presents the requested video information to the user.” Id. at col. 1 ll. 36–39. Claim 1 recites:

[1a] A method in a video receiving system for receiving video information, the method comprising:

[1b] receiving, by a receiver, a request by a user for a unit of video information; [1c] receiving, by the receiver, a plurality of video information streams, each of which represents the requested unit of video information; [1d] identifying, by the receiver, which of the plurality of video information streams, when processed, is expected to result in a

3lower latency in presenting the unit of video information; and [1e] processing, by the receiver, the identified video information stream to present the unit of video information.

Id. at claim 1 (emphasis and bracketed labels added).

The main prior-art references relevant to this appeal are Baldwin¹ and Cooper.² Baldwin is titled “Fast Start-up for Digital Video Streams” and was filed on June 13, 2003. Baldwin teaches “reduc[ing] the effective start-up delay in the presentation of the first frames of the video content that occurs when a system tunes into a video stream.” Baldwin col. 4 ll. 60–64. “To further minimize the start-up delay, the exemplary fast start-up system may account for the randomness at which a user tunes into a video-stream channel by employing multiple alternative, [random access point (“RAP”)] phase-staggered video-stream transmissions.” Id. at col. 9 ll. 3–7. Cooper is titled “Robust Mode Staggercasting Fast Channel Change” and was filed on January 23, 2004. Cooper teaches “a method and apparatus for staggercasting a plurality of content representative signals.” Cooper col. 2 ll. 54–59. Content representative signals may include “a video image signal, audio sound image, program data, or any combination of these.” Id. at col. 3 ll. 52–54. “A composite signal is generated comprising the plurality of first and second . . . signals,” in which “each respective second . . . signal is delayed with respect to the corresponding first . . . signal.” Id. at Abstract.

Netflix, Inc. (“Netflix”) petitioned for IPR of claims 1, 3–12, and 14–20 of the ’014 patent. The Board found Baldwin alone would have rendered obvious claims 1, 3–5, 7, 9–4 10, 12, 14–16, and 18–20 of the ’014 patent. The Board also found that Cooper alone and the combination of Cooper and Baldwin would not have rendered obvious claims 6, 8, 11, and 17.

Avago timely appealed the Board’s final written decision as to claims 1, 3–5, 7, 9–10, 12, 14–16, and 18–20. Netflix timely cross-appealed the Board’s final written decision as to claims 6, 8, 11, and 17. We have jurisdiction under 28 U.S.C. § 1295(a)(4)(A).

DISCUSSION

Obviousness is a question of law based on underlying findings of fact. Novartis AG v. Torrent Pharms. Ltd., 853 F.3d 1316, 1327 (Fed. Cir. 2017). “What the prior art teaches [and] whether a person of ordinary skill in the art would have been motivated to combine references . . . are questions of fact.” Meiresonne v. Google, Inc., 849 F.3d 1379, 1382 (Fed. Cir. 2017). These “subsidiary factual findings are reviewed for substantial evidence.” Novartis, 853 F.3d at 1327. “Substantial[-]evidence review asks whether a reasonable fact finder could have arrived at the agency’s decision and requires examination of the record as a whole, taking into account evidence that both justifies and detracts from an agency’s decision.” Intelligent Bio-Sys., Inc. v. Illumina Cambridge Ltd., 821 F.3d 1359, 1366 (Fed. Cir. 2016) (cleaned up). “Where two different conclusions may be warranted based on the evidence of record, the Board’s decision to favor one conclusion over the other is the type of decision that must be sustained by this court as supported by substantial evidence.” In re Chudik, 851 F.3d 1365, 1371 (Fed. Cir. 2017) (quoting In re Bayer Aktiengesellschaft, 488 F.3d 960, 970 (Fed. Cir. 2007)).

Avago raises two main issues on appeal: (1) that the Board’s findings that Baldwin teaches limitations [1c] and [1d] are not supported by substantial evidence; and (2) that the Board erred by refusing to decide whether claim 1 requires limitation [1c] to be performed before limitation5 [1d]. Netflix’s cross-appeal challenges the Board’s findings that claims 6, 8, 11, and 17 were not shown to be unpatentable and that Netflix failed to demonstrate a motivation to combine Cooper and Baldwin. We address each argument in turn.

I

Avago’s main argument is that the Board’s findings that Baldwin teaches limitations [1c] and [1d] are not supported by substantial evidence. We disagree.

Limitations [1c] and [1d] are reproduced below:

[1c] receiving, by the receiver, a plurality of video information streams, each of which represents the requested unit of video information; [1d] identifying, by the receiver, which of the plurality of video information streams, when processed, is expected to result in a lower latency in presenting the unit of video information.

’014 patent claim 1. The Board found that Baldwin teaches “that its receiver can concurrently receive a plurality of streams and select one of those streams to reduce latency of presentation.” J.A. 18 (citing Baldwin col. 6 ll. 25–33).

On appeal, Avago primarily disputes the Board’s factual findings that Baldwin discloses that the receiver receives a plurality of streams and identifies which of the plurality of streams is expected to result in lower latency, i.e., limitations [1c] and [1d]. For example, Avago argues that “[i]n Baldwin’s multicast system, the server transmits multiple streams, but a specific receiver only receives a single stream.” Appellant’s Br. 46. Avago also argues that Baldwin teaches that the server, not the receiver, selects the lower-latency stream. See Appellant’s Br. 55–56; see6 also Appellant’s Reply Br. 23–25. The Board’s reading of Baldwin is supported by substantial evidence.

Relevant to Avago’s argument that Baldwin does not disclose the receiver receiving a plurality of streams, Baldwin states: “By sending multiple different streams, tuning time is improved because the receiver may select one of the lead-in streams to play. The one selected will typically be the one which will be ready to be presented the quickest after the time at which the user tunes.” Baldwin col. 9 ll. 24–28 (emphasis added). Thus, according to the Board, “Baldwin provides a straightforward, express disclosure of ‘sending multiple streams’ to a receiver and allowing the receiver to select from those streams.” J.A. 19; see also J.A. 18–19 (citing Baldwin col. 9 ll. 24–28). The Board also found that Baldwin’s disclosure that the “alternative transmissions ‘need not be sent concurrently to a particular receiver’” further supports its finding that Baldwin teaches sending multiple streams to the receiver. J.A. 19 (quoting Baldwin col. 9 ll. 29–33 (emphasis added)). The Board reasoned that “by stating that the streams do not have to be sent concurrently, Baldwin further teaches or reasonably suggests that alternative streams can be sent concurrently to a receiver.” J.A. 19 (emphasis in original). Avago has not demonstrated that these findings lack substantial evidence.

As to Avago’s argument that Baldwin does not disclose the receiver identifying which of the plurality of streams is expected to result in lower latency, Avago maintains that Baldwin’s Figures 5 and 6 “establish that Baldwin’s server identifies the lower latency stream, not the receiver.” Appellant’s Br. 55. The Board also rejected this argument. It found that “the fact[] that Baldwin might also teach that a server can select a stream and transmit only one stream does not negate the plain teachings of Baldwin.” J.A. 19. The Board also found that “Baldwin expressly teaches sending multiple streams of the same content with staggered RAPs so that the receiver can receive the multiple7 streams and identify the stream that is expected to result in reduced latency.” J.A. 19. Thus, Avago has not shown that the Board’s findings are not supported by substantial evidence.

II

Next, we address Avago’s argument that the Board erred by refusing to decide whether claim 1 requires limitation [1c] to be performed before limitation [1d].

The Board found that there was no need to reach Avago’s proposed sequential-order construction for two main reasons. First, the Board determined that Avago did “not sufficiently explain why the claim should be interpreted in this way,” i.e., that claim 1 requires limitation [1c] to be performed and completed before limitation [1d]. Second, the Board found that “we need not decide this [sequential-order construction] issue because we determine Baldwin teaches or reasonably suggests performing the steps in [Avago’s] proposed order.” J.A. 19 n.6.

Because we determine that substantial evidence supports the Board’s finding that Baldwin teaches limitations [1c] and [1d] performed in Avago’s proposed order, we see no error in the Board’s decision to not reach Avago’s sequential-order construction.

We have considered Avago’s remaining arguments and find them unpersuasive. For the foregoing reasons, we affirm the Board’s findings that claims 1, 3–5, 7, 9–10, 12, 14–16, and 18–20 are unpatentable.³

8III

Last, we address Netflix’s cross-appeal. Netflix argues that the Board’s finding as to dependent claims 6, 8, 11, and 17 should be vacated and remanded for the Board to address Netflix’s motivation-to-combine arguments concerning the combination of Cooper and Baldwin. We agree.

In its petition, Netflix argued that a person of ordinary skill in the art would have been motivated to combine Cooper and Baldwin, J.A. 70–75, and relied on Cooper to teach the additional limitations found in dependent claims 6, 8, 11, and 17, see J.A. 106–16, 121–22. The Board found that Cooper alone teaches all the limitations of claim 1, except for limitation [1d]. J.A. 13–16. Specifically, it found that “Cooper does not teach that its receiver identifies which of two streams is expected to result in a lower latency, as required by limitation [1d].” J.A. 13. The Board also found that it was “not persuaded” by Netflix’s “argument that a person of ordinary skill in the art would have been motivated to combine Cooper and Baldwin to yield the subject matter recited by the challenged claims.” J.A. 9. The Board explained that “it is not clear how or where [Netflix] relies on the proposed combination to satisfy any specific claim limitations.” J.A. 9. As a result, the Board found that because Netflix “has not shown that [the] independent claims . . . are unpatentable in view of Cooper alone or based on the combined teachings of Cooper and Baldwin, [Netflix] has not shown that [the] dependent claims [6, 8, 11, and 17] are unpatentable in view of Cooper alone or based on the combined teachings of Cooper and Baldwin.” See J.A. 25; see also J.A. 27, 30.

The problem with the Board’s conclusion is that it fails to reach or provide any substantive analysis of Netflix’s motivation-to-combine arguments. In its petition, Netflix provided a detailed explanation of why a person of ordinary skill in the art would have been motivated to combine Cooper’s receiver teachings with Baldwin’s teachings of9 providing alternative streams with staggered access points and selecting the stream with the next available access point. See J.A. 70–75. Netflix explained that a person of ordinary skill in the art would have combined the two prior-art references according to known methods to yield predictable results “because both teach video receivers that receive alternative streams representing content selected by a viewer.” J.A. 72; see also J.A. 71. Netflix also argued that combining these two prior-art references is a “‘simple substitution of one known element’ (e.g., Baldwin’s teaching of identifying which stream has the next available access point) for another (e.g., Cooper’s teaching of identifying which stream is more readily available) to obtain predictable results.” J.A. 73. Last, Netflix explained that “[i]mplementing Baldwin’s teaching of identifying which stream has the next available access point represents a choice from a finite number of identified predictable solutions with a reasonable expectation of success.” J.A. 74. Despite Netflix raising these motivation-to-combine arguments, the Board did not substantively address Netflix’s arguments or provide any explanation as to why a person of ordinary skill in the art would not have been motivated to combine Cooper and Baldwin. See J.A. 8–10.

We also disagree with the Board that “it is not clear how or where [Netflix] relies on the proposed combination to satisfy any specific claim limitations.” J.A. 9. Netflix’s petition clearly argued that the combination of Cooper and Baldwin is necessary to meet limitation [1d] should the Board find that Cooper does not teach limitation [1d]. Netflix’s petition stated:

To the extent that [Avago] argues that the foregoing teachings of Cooper relating to a receiver that identifies which of the multiplexed packet streams is not time-delayed does not render limitation [1d] obvious, then a [person of ordinary skill in the art] would have found it obvious for Cooper and

10Baldwin to teach this limitation based on Baldwin’s teachings.

J.A. 89. Accordingly, we vacate the Board’s determination as to claims 6, 8, 11, and 17 and remand for the Board to address (1) Netflix’s motivation-to-combine arguments for the combination of Cooper and Baldwin in view of the Board’s finding that Cooper does not teach limitation [1d]; and (2) whether Cooper teaches the additional limitations found in dependent claims 6, 8, 11, and 17. See Cross-Appellant’s Br. 52, 63.

CONCLUSION

For the foregoing reasons, we affirm the Board’s determination as to claims 1, 3–5, 7, 9–10, 12, 14–16, and 18–20 of the ’014 patent. We vacate and remand the Board’s determination as to claims 6, 8, 11, and 17 of the ’014 patent for further proceedings consistent with this opinion. AFFIRMED AS TO THE MAIN APPEAL; VACATED

AND REMANDED AS TO THE CROSS-APPEAL

COSTS

Costs to Netflix.