March 21, 2023
On March 13, in Intel Corp. v. PACT XPP Schweiz AG, the Federal Circuit concluded that the “known techniques” rationale may support a motivation to combine two references so long as the combination is a “suitable option” to address a known problem in the art, even if the combination does not result in an obvious improvement to the proposed system. Based on this, the Federal Circuit reversed the Patent Trial and Appeal Board (PTAB) because it had misapplied the “known techniques” rationale in finding that there was no motivation to combine two references. The court’s decision may be viewed as lowering the bar for showing a motivation to combine references in the absence of an explicit teaching or suggestion to combine.
PACT’s patent at issue, U.S. Patent No. 9,250,908, relates to multiprocessor systems and how processors in those systems access data. These systems require a mechanism to monitor and maintain cache coherency. One way to maintain such cache coherency is by “snooping” along a shared bus. Kabemoto, one of the references in the obviousness combination, teaches this method. Kabemoto also separately teaches a secondary cache. Bauman, the other reference in the combination, teaches using a global, segmented secondary cache to maintain cache coherency. The relevant claim language at issue requires “an interconnect system interconnecting each of the separated cache segments with each of the processors . . . and each of the separated cache segments with neighboring separated cache segments.”
Intel argued that a person of ordinary skill would replace Kabemoto’s secondary caches with Bauman’s segmented global secondary cache by connecting the global secondary cache to Kabemoto’s snoop bus on the processor. This modification would result in a system with the claimed separated cache and interconnecting system. During the PTAB trial, PACT did not dispute that this combination taught all limitations of the challenged claim. Rather, PACT argued that there was no motivation to combine Kabemoto with Bauman. The PTAB agreed because Kabemoto purportedly already addressed the problem of cache coherency with its snoop bus. The PTAB therefore failed to see why one skilled in the art would use the technique of Bauman to address the issue of cache coherency where it was not an obvious improvement over Kabemoto’s own solution for cache coherency.
Revisiting the motivation to combine issue on appeal, and in particular, the “known techniques” rational for a motivation to combine, the Federal Circuit reversed. The Federal Circuit first emphasized that “universal” motivations known in a particular field provide a motivation to combine even without any teaching or suggestion to combine in any reference itself. Likewise, if a technique has been used to improve a device, and a person of ordinary skill would recognize that this technique could improve similar devices, the technique is obvious if it is within the skill of a person of ordinary skill. Importantly, the Federal Circuit emphasized that so long as a known problem in the field is being addressed by combination of references, it is not necessary that the combination be the best option, but only that it be a suitable option.
Under this legal framework, the Federal Circuit determined that the PTAB had erred in rejecting Intel’s known-technique rationale to combine the references. The PTAB had reasoned that because the Kabemoto reference already addresses the issue of cache coherency, there would be no reason why a person of ordinary skill would implement Bauman’s cache coherency solution in Kabemoto, because Bauman’s technique was not an obvious improvement. The Federal Circuit rejected this reasoning on the basis that under the “known techniques” rationale, a combination merely needs to be suitable to address a known problem in the field—it does not need to be an obvious improvement. Because Bauman’s global secondary cache was used to solve cache coherency in Bauman itself, and because a person of ordinary skill would recognize that a global secondary cache would be a suitable solution to the cache coherency problem in similar systems (such as Kabemoto), there is a motivation to combine the implementation of Bauman’s cache with Kabemoto.
This decision provides additional guidance on the application of the “known techniques” rationale beyond what was articulated earlier by the U.S. Supreme Court in KSR v. Teleflex. In particular, where two references describe devices in the same field, and there is a known technique that has been used to improve the functionality of one device, then a motivation to combine the references may exist if a person of ordinary skill would recognize the technique as a way to improve similar devices. It is not necessary that either reference express a suggestion to use the technique on other devices, or that the combination would result in an obvious improvement. All that is required is a known problem in the art, a solution to the problem found in a reference, and a finding that the combination would not be beyond the capabilities of the ordinary artisan.
For more information on this holding, please contact Fitch Even associate Brian P. Herrmann, the author of this alert.
Fitch Even IP Alert®