UNITED STATES PATENT AND TRADEMARK OFFICE
`______________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`______________________
`MICROSOFT CORPORATION,
`Petitioner,
`v.
`SAINT REGIS MOHAWK TRIBE,
`Patent Owner.
`_______________________
`IPR2018-01607
`Patent No. 7,620,800
`__________________________
`
`PATENT OWNER SAINT REGIS MOHAWK TRIBE’S PRELIMINARY
`RESPONSE TO PETITION FOR INTER PARTES REVIEW OF U.S.
`PATENT 7,620,800
`
`
`
`i
`
`
`
`
`
`

`

`TABLE OF CONTENTS
`
`
`I. INTRODUCTION .................................................................................................. 1
`II. PERTINENT FACTS ............................................................................................ 3
`A. SRC Computers creates the first high performance reconfigurable
`computer. .................................................................................................... 3
`B. Related Proceedings. ................................................................................... 4
`III. THE BOARD SHOULD DENY INSTITUTION UNDER § 314(a). ................. 5
`A. The Board should deny institution because SRC is a sole-source
`supplier for the U.S. Army’s TRACER Program....................................... 7
`B. The Board should deny institution because there are two district
`court cases involving the same patent and overlapping prior art. ............ 10
`C. The Board should deny institution because of the Tribe’s status as a
`sovereign. .................................................................................................. 12
`IV. TECHNOLOGY BACKGROUND ................................................................... 13
`A. Conventional Computer Architecture. ...................................................... 13
`B. Field Programmable Gate Arrays. ............................................................ 14
`C. Reconfigurable Computing. ...................................................................... 15
`D. The ’800 patent: SRC invents methods for enhancing parallelism
`and performance in reconfigurable computing systems. .......................... 16
`V. PATENT OWNER’S CLAIM CONSTRUCTIONS .......................................... 18
`A. The Board should apply the Phillips standard. ......................................... 18
`B. Terms to be construed. .............................................................................. 20
`1. “functional unit” .............................................................................. 20
`2. “computational loop” ...................................................................... 21
`3. “form” .............................................................................................. 23
`
`
`
`ii
`
`

`

`4. “data driven” .................................................................................... 25
`5. “transforming an algorithm into a data driven calculation that
`is implemented by said reconfigurable computing system at
`the at least one reconfigurable processor” .................................... 26
`6. “pass computed data seamlessly between said computational
`loops” ............................................................................................ 27
`7. “stream communication” ................................................................. 29
`VI. Microsoft Failed To Demonstrate A Reasonable Likelihood Of
`Prevailing As To Any Challenged Claim. ..................................................... 31
`A. Microsoft has failed to demonstrate that Splash2 anticipates or
`renders claims 1, 18, 22, or 22 obvious because Splash2 does not
`teach the instantiation of any “computational loops” as required by
`those claims. ............................................................................................. 31
`B. Microsoft has failed to show that claims 1, 18, 21, and 22 are
`obvious under Splash2 in view of Gaudiot for the same reasons. ........... 38
`C. Microsoft has failed to show that claims 2-5, 22, and 23 are
`obvious over Splash2 in view of Roccatano, with or without
`Gaudiot. .................................................................................................... 38
`VII. CONCLUSION ................................................................................................ 43
`VIII. LIST OF EXHIBITS ....................................................................................... 44
`
`
`
`
`
`
`
`iii
`
`

`

`TABLE OF AUTHORITIES
`
`CASES:
`Altiris, Inc. v. Symantec Corp.,
`318 F.3d 1363 (Fed. Cir. 2003) ............................................................................... 22
`
`
`Cuozzo Speed Techs. v. Lee,
`136 S. Ct. 2131........................................................................................................... 6
`
`
`Michigan v. Bay Mills Indian Cmty.,
`134 S. Ct. 2024 (2014) ............................................................................................. 13
`
`
`Phillips v. AWH Corp.,
`415 F.3d 1303 (Fed. Cir. 2005) ......................................................................... 21, 22
`
`
`Saint Regis Mohawk Tribe v. Mylan Pharm. Inc.,
`896 F.3d 1322 (Fed. Cir. 2018) ............................................................................... 13
`
`
`Seachange Intern’l, Inc. v. C-COR, Inc.,
`413 F.3d 1361 (Fed. Cir. 2005) ............................................................................... 27
`
`
`ADMINISTRATIVE ORDERS:
`Apple Inc. v. Uniloc Luxembourg, S.A.,
`IPR2017-02202, slip op. at 9-10 (PTAB Dec. 13, 2018) .................................. 37, 39
`
`
`Butamax Advanced Biofuels LLC v. Gevo, Inc.,
` Case IPR2014-00581, slip op. at 12–13 (PTAB Oct. 14, 2014) ............................. 12
`
`General Plastic Indus. Co., Ltd. v. Canon Kabushiki Kaisha,
` Case IPR2016–01357, Paper 19 at 9-10 (PTAB Sept. 6, 2017) .................... 6, 7, 11
`
`Infobionic, Inc. v. Braemer Manufacturing, LLC,
`
`IPR2015-01704, slip op. at 14-15 (PTAB Feb. 16, 2016) ................................ 37, 39
`
`Kinetic Techs. v. Skyworks Solutions, Inc.,
` No. IPR2014-00529, slip op. at 15-16 (PTAB Sept. 23, 2014) ........................ 37, 40
`
`Masabi Ltd. v. Bytemark, Inc.,
`
`IPR2017-01449, slip op. at 43-44 (PTAB Dec. 3, 2018) .................................. 37, 39
`iv
`
`
`
`

`

`
`Mylan Pharms., Inc. v. Bayer Intellectual Property GMBH,
` Case IPR2018-01143, slip op. at 12-14 (PTAB Dec. 3, 2018) ............................... 11
`
`NHK Spring Co., Ltd v. Intri-Plex Techs., Inc.,
` Case IPR2018-00752, slip op. at 19-20 (PTAB Sept. 12, 2018) ............................ 11
`
`TRW Automotive US LLC v. Magna Elecs., Inc.,
` No. IPR2014-00258, slip op. at 10-11 (PTAB Aug. 27, 2014) ........................ 38, 40
`STATUTES:
`35 U.S.C. § 312 ............................................................................................................ 31
`
`35 U.S.C. § 314 .................................................................................... 5, 6, 9, 11, 12, 43
`
`35 U.S.C. § 316 .............................................................................................................. 7
`REGULATIONS:
`37 C.F.R. § 42.104 ................................................................................................. 32, 40
`
`37 C.F.R. § 42.22 ......................................................................................................... 31
`
`37 C.F.R. § 42.65 ................................................................................................... 37, 39
`
`
`
`
`
`
`v
`
`

`

`I. INTRODUCTION
`Microsoft has shamelessly copied and claimed credit for inventing the paradigm
`
`shifting, high-performance reconfigurable technology invented by Seymour R.
`
`Cray’s last companySRC Computers. Microsoft has now launched an all-out
`
`assault on the patents covering this ground-breaking technology by filing ten
`
`petitions for inter partes review (“IPR”) against six patents that are the subject of a
`
`patent infringement lawsuit filed by SRC Labs, LLC (a successor to SRC
`
`Computers) and the Saint Regis Mohawk Tribe.
`
`The Board should exercise its discretion to deny institution for three reasons.
`
`First, IPRs were intended to be an efficient, cost-effective alternative to
`
`litigation, not a burdensome and inefficient duplicate of litigation. Microsoft’s
`
`Petition mirrors invalidity arguments it has asserted in the co-pending district court
`
`case.
`
`Second, Microsoft is one of the worst abusers of the IPR process. A recent
`
`empirical study by Robins Kaplan LLP shows that Microsoft is a top five filer of
`
`IPR petitions and that 59% of its petitions are duplicative, meaning that they
`
`challenge at least one claim that is the subject of another Microsoft IPR petition.
`
`EX. 2039 at 2-3. True to form, Microsoft has filed three duplicative petitions
`
`against claim 1 of U.S. Patent No. 7,620,800 (“’800 patent”). IPR2018-1605;
`
`IPR2018-01606; IPR2018-01607. Each of these duplicative petitions relies on the
`
`
`
`1
`
`

`

`same primary reference, Splash 2: FPGAs in a Custom Compute Machine. But
`
`Duncan A. Buell, a primary editor of this reference and the project manager for the
`
`Splash 2 system, described SRC Computer’s high-performance reconfigurable
`
`computing technology in 2004 as “a watershed event in the history of computing.”
`
`EX. 2035 at 5.
`
`Third, for more than a decade, SRC1 has been the sole source provider to
`
`Lockheed Martin on behalf of the U.S. Southern Command of high-performance
`
`reconfigurable processors for the Tactical Reconnaissance and Counter-
`
`Concealment Radar (“TRACER”) program, which requires extremely high-
`
`performance signal processing in a very limited size, weight, and power (“SWAP”)
`
`environment. EX. 2033 ¶¶ 2-23. No other vendor can match the capabilities of
`
`SRC’s systems. Id. ¶ 21. The cost of defending the ten duplicative IPR petitions
`
`filed by Microsoft may put SRC out of business. EX. 2036 ¶¶ 9-10. That result
`
`would be extremely detrimental to the United States national security interests. EX.
`
`2033 ¶¶ 22-23. Small innovative companies like SRC are key to developing the
`
`advanced cutting-edge technology needed for the Department of Defense’s most
`
`critical missions. Id. ¶¶ 4-8.
`
`In addition, Microsoft has failed to establish a reasonable likelihood of
`
`
`1 In this brief, SRC will be used as shorthand to refer to both SRC Computers and
`its successor DirectStream, LLC.
`
`
`
`2
`
`

`

`prevailing against any claim of the ’800 patent because the cited prior art fails to
`
`disclose multiple claim limitations. Therefore, the Board must deny institution.
`
`II. PERTINENT FACTS
`A. SRC Computers creates the first high performance reconfigurable
`computer.
`The inventors of the ’800 patent and their colleagues pioneered the use of Field
`
`Programmable Gate Arrays (FPGAs) as general-purpose processors to create
`
`small, energy efficient, supercomputers. These new supercomputers outperform
`
`conventional computers by a factor of 100x (or more) while using 99% less power.
`
`These innovations were the result of private research and development done by
`
`SRC Computers, which was founded in 1996 by Jim Guzy, Jon Huppenthal, and
`
`Seymour Rodger Cray (hence SRC), who is widely considered to be the father of
`
`supercomputing. EX. 2031. Notably, SRC Computers’ first customers were the
`
`National Security Agency (“NSA”), the Naval Postgraduate School, and George
`
`Washington University. EX. 2036 ¶ 7. SRC Computers spent over $100 million in
`
`research and development for its patented reconfigurable supercomputers. Id. ¶ 3.
`
`SRC Computers has restructured into three entities: a corporate parent FG-SRC,
`
`LLC, an operating company DirectStream, LLC, and a licensing entity called SRC
`
`Labs, LLC. Id. ¶ 1. DirectStream and SRC Labs operate in tandem and FG-SRC is
`
`responsible for the management and funding of both entities. Id. This proceeding,
`
`coupled with the other nine Microsoft IPR petitions, may force DirectStream out of
`
`
`
`3
`
`

`

`business as it is counting on licensing revenue from SRC Labs to operate and the
`
`costs of defending these IPRs may exceed $700,000. Id. ¶¶ 9-11.
`
`In an effort to diversify its economy and foster jobs, the Saint Regis Mohawk
`
`Tribe (“Tribe”) recently created an Office of Technology, Research and Patents
`
`(“OTRAP”). OTRAP’s purpose is to strengthen the Tribal economy by
`
`encouraging the development of emerging science and technology initiatives and
`
`projects, and promoting the modernization of Tribal and other businesses. EX.
`
`2053 at 1. The objective is to create revenue, jobs, and new economic development
`
`opportunities for the Tribe and its members. Id. OTRAP will also promote the
`
`education of Mohawks in the fields of science, technology, engineering, and math.
`
`Id. at 2.
`
`All revenue generated by OTRAP will go into the Tribal General Fund and be
`
`used to address the chronically unmet needs of the Tribal community, such as
`
`housing, employment, education, healthcare, cultural and language preservation.
`
`Id.
`
`The ’800 patent has been assigned to the Tribe and the Tribe granted SRC Labs
`
`an exclusive license. EX. 2036 ¶ 2.
`
`B. Related Proceedings.
`On October 18, 2017, the Tribe and its exclusive licensee, SRC Labs, LLC
`
`(“SRC”), sued Microsoft for infringement of six different patents that were
`
`
`
`4
`
`

`

`originally assigned to SRC Computers: The ’800 patent and U.S. Patent Nos.
`
`6,076,152, 6,247,110, 6,434,687, 7,225,324 (“’324 patent”), and 7,421,524. This
`
`case was originally filed in the Eastern District of Virginia (SRC Labs, LLC et al v.
`
`Microsoft Corporation, No. 1:17-cv-01172-LO-JFA) but was transferred to the
`
`Western District of Washington for Microsoft’s convenience (SRC Labs, LLC et al
`
`v. Microsoft Corporation, No. 2:18-cv-00321-JLR).
`
`On October 18, 2017, the Tribe and SRC also filed a patent infringement action
`
`against Amazon Web Services, Inc., Amazon.com, Inc., and VADATA, Inc.
`
`(collectively “Amazon”) in the Eastern District of Virginia (SRC Labs, LLC et al v.
`
`Amazon Web Services, Inc. et al, No. 1:17-cv-01227-LO-JFA), which was
`
`similarly transferred to the Western District of Washington (SRC Labs, LLC et al v.
`
`Amazon Web Services, Inc et al, 2:18-cv-00317-JLR). The Amazon case also
`
`involves the ’324 and ’800 patents.
`
`Between August 24, 2018 and September 11, 2018, Microsoft filed ten IPR
`
`petitions. IPR2018-01594, IPR2018-01599, IPR2018-01600, IPR2018-01602,
`
`IPR2018-01603, IPR2018-01604, IPR2018-01605, IPR2018-01606, and IPR2018-
`
`01607. Three of the petitions, including this one, challenge claims of the ’800
`
`patent and three challenge the claims of the ’324 patent.
`
`III. THE BOARD SHOULD DENY INSTITUTION UNDER § 314(a).
`The Board should exercise its discretion under 35 U.S.C. § 314(a) to deny
`
`
`
`5
`
`

`

`Microsoft’s Petition for three reasons. First, Microsoft’s petitions are an attack on a
`
`small, innovative, U.S. company that is the sole-source supplier to the Department
`
`of Defense. Second, there is a district court case and two other IPR petitions
`
`involving the same patent. And third, the Patent Owner is a sovereign.
`
`Section 314(a) provides the Director with discretion to deny a petition. See 35
`
`U.S.C. § 314(a); Cuozzo Speed Techs. v. Lee, 136 S. Ct. 2131, 2140 (“[T]he
`
`agency’s decision to deny a petition is a matter committed to the Patent Office’s
`
`discretion.”).
`
`To assist the Board’s assessment of the potential impacts on both the efficiency
`
`of the IPR process and the fundamental fairness of the process for all parties, the
`
`General Plastic decision was recently designated a Precedential decision because it
`
`enumerated the following seven non-exclusive factors that the Board will consider
`
`in exercising discretion on instituting IPR:
`
`1. whether the same petitioner previously filed a petition
`directed to the same claims of the same patent;
`2. whether at the time of filing the first petition the
`petitioner knew of the prior art asserted in the second
`petition or should have known of it;
`3. whether at the time of filing of the second petition the
`petitioner already received the patent owner's preliminary
`response to the first petition or received the Board's
`decision on whether to institute review in the first
`petition;
`
`
`
`6
`
`

`

`4. the length of time that elapsed between the time the
`petitioner learned of the prior art asserted in the second
`petition and the filing of the second petition;
`5. whether the petitioner provides adequate explanation
`for the time elapsed between the filings of multiple
`petitions directed to the same claims of the same patent;
`6. the finite resources of the Board; and
`7. the requirement under 35 U.S.C. § 316(a)(11) to issue
`a final determination not later than 1 year after the date
`on which the Director notices institution of review.
`General Plastic Indus. Co., Ltd. v. Canon Kabushiki Kaisha, Case IPR2016–
`
`01357, Paper 19 at 9-10 (PTAB Sept. 6, 2017) (precedential).
`
`But these factors are neither dispositive nor exclusive and are “not intended to
`
`represent all situations where it may be appropriate to deny a petition.” See Trial
`
`Practice Guide Update (August 2018) at 10. There may be other reasons where the
`
`“effect … on the economy, the integrity of the patent system, the efficient
`
`administration of the Office, and the ability of the Office to timely complete
`
`proceedings … favor denying a petition even though some claims meet the
`
`threshold standards for institution.” Id. Some examples listed in the Trial Practice
`
`Guide Update include other proceedings “related to the same patent, either at the
`
`Office, in the district courts, or the ITC.” Id.
`
`A. The Board should deny institution because SRC is a sole-source supplier
`for the U.S. Army’s TRACER Program.
`The Department of Defense’s most critical missions rely on cutting-edge
`
`
`
`7
`
`

`

`technology developed and manufactured in the United States by American
`
`companies like SRC. EX. 2033 ¶¶ 4-7. Because SRC’s patented, reconfigurable
`
`supercomputers are much smaller and more energy efficient than traditional
`
`computers, the technology is a perfect fit for applications where space and power
`
`are at a premium. EX. 2032. As a result, Lockheed Martin chose SRC to be the
`
`sole source vendor of processors for the U.S. Army’s TRACER program. EX.
`
`2033 ¶ 9.
`
`SRC TRACER Signal Data Processor (SDP)
`
`SRC Series H MAP processor
`
`
`
`The TRACER program addresses a critical need to identify hidden targets,
`
`
`
`facilities, and enemy equipment such as small roadside targets and buried weapons
`
`caches. EX. 2034. TRACER has been operational outside the continental United
`
`States for ten years performing operations for the U.S. Southern Command. EX.
`
`2033 ¶ 14. SRC/DirectStream’s processors have allowed these surveillance
`
`operations to produce images of targets on the ground in real-time, providing
`
`immediately actionable information. Id. ¶¶ 15-16. The TRACER program’s unique
`
`
`
`8
`
`

`

`real-time radar capabilities have allowed the U.S. Southern Command to detect
`
`rebel forces, drug cartel activities, and terrorism that occurs in the dense jungles of
`
`that region. Id. ¶ 18.
`
`According to Lockheed Martin’s Engineering Program Manager in charge of
`
`the TRACER program, it is not in the national security interests of the United
`
`States to require SRC to spend time or money defending IPRs. Id. ¶ 23. Instead, it
`
`is in the best interests of the United States as a whole, and Lockheed Martin in
`
`particular, to keep companies like SRC/DirectStream healthy and unencumbered
`
`so they focus on new technology development. Id. ¶ 22. In fact, Lockheed Martin’s
`
`procurement process has shown that no other vendor can match the capabilities of
`
`SRC’s TRACER processors. Id. ¶¶ 19-21.
`
`Microsoft’s petitions are an attack on a small, innovative, U.S. company that is
`
`a sole-source supplier to the Department of Defense. Small innovative companies
`
`are the lifeblood of our economy and the intended beneficiaries of the U.S. patent
`
`system. Microsoft should not be allowed to use the IPR process to avoid paying
`
`royalties when it steals the intellectual property of small companies like
`
`SRC/Directstream.
`
`Accordingly, the Tribe asks the Board to exercise its discretion under § 314(a)
`
`to deny this Petition.
`
`
`
`9
`
`

`

`B. The Board should deny institution because there are two district court
`cases involving the same patent and overlapping prior art.
`The Board should deny institution because Microsoft is relying on the same
`
`prior art (Splash2, Roccatano, and Gaudiot) and arguments in its district court
`
`invalidity contentions as asserted in this Petition. Microsoft’s Petition asserts that
`
`claims 1-5, 18, and 21-23 of the ’800 patent are unpatentable over the following
`
`prior art:
`
`1. Claims 1, 18, 21, and 22 are anticipated by Splash2 (EX. 1007).
`
`2. Claim 1, 18, 21, and 22 are obvious over Splash2.
`
`3. Claim 1, 18, 21, and 22 are obvious over Splash2 in view of Gaudiot
`
`(EX. 1010).
`
`4. Claims 2-5, 22 and 23 are obvious over Splash2 in view of Roccatano
`
`(EX1012), with or without Gaudiot.
`
`Petition at 5.
`
`In the district court case Microsoft has also asserted anticipation based on
`
`Splash2 and obviousness based on Splash2 in view of Roccatano and Gaudiot. EX.
`
`2052 at 11, 12, 22, 26-27; EX. 2055; EX. 2057; EX. 2060. But Microsoft waited
`
`ten months to file this Petition and then moved to stay the district court case
`
`pending the resolution of its ten IPR petitions. EX. 2019; EX. 2020. Prior to the
`
`stay, the district court was scheduled to have a Markman hearing on December 20,
`
`2018 and trial in November 2019. EX. 2018 at 2. Thus, the district court would
`10
`
`
`
`

`

`have analyzed the same issues and resolved them before any trial on this Petition
`
`could have concluded.
`
`Microsoft’s petitions were filed to delay the resolution of these issues by the
`
`district court. This is contrary to the overall goal of the AIA, which was to “make
`
`the patent system more efficient by the use of post-grant review proceedings.” See
`
`General Plastic Industrial Co., Ltd. v. Canon Kabushiki Kaisha, Case IPR2016-
`
`01357, slip op. at 16-17 (PTAB Sept. 6, 2017) (Paper 19) (precedential as to §
`
`II.B.4.i). IPRs were not intended to be a tool to enable efficient infringement by
`
`trillion-dollar corporations, such as Microsoft.
`
`The Board should deny this Petition based on these inefficiencies as the Board
`
`has recently done in other proceedings involving parallel district court cases with
`
`overlapping prior art. See NHK Spring Co., Ltd v. Intri-Plex Techs., Inc., Case
`
`IPR2018-00752, slip op. at 19-20 (PTAB Sept. 12, 2018) (Paper 8) (denying
`
`institution under § 314(a) because of co-pending litigation); Mylan Pharms., Inc. v.
`
`Bayer Intellectual Property GMBH, Case IPR2018-01143, slip op. at 12-14 (PTAB
`
`Dec. 3, 2018) (Paper 13) (denying institution under § 314(a) because of co-pending
`
`litigation involving overlapping prior art).
`
`Unlike those cases, the ’800 patent is also being asserted in a second district
`
`court case against Amazon. And Amazon has also asserted invalidity arguments
`
`based on the same primary reference at issue in this proceeding, Splash2. EX. 2050
`
`
`
`11
`
`

`

`at 3; EX. 2051. Amazon will not be estopped from asserting invalidity defenses
`
`based on Splash2 in district court if this trial is instituted and proceeds to a final
`
`decision in Patent Owner’s favor. Thus, the inefficiencies here are greater than in
`
`the NHK and Mylan cases cited above.
`
`Finally, Microsoft has filed three separate, but largely duplicative, petitions
`
`against the ’800 patent: IPR2018-01605, IPR-01606, and IPR-01607. All three
`
`petitions argue that claim 1 of the ’800 patent is unpatentable based on the Splash2
`
`reference. Microsoft’s duplicative challenges of the ’800 patent is both inefficient
`
`and impacts Patent Owner’s ability to defend its patent. See Butamax Advanced
`
`Biofuels LLC v. Gevo, Inc., Case IPR2014-00581, slip op. at 12–13 (PTAB Oct.
`
`14, 2014) (Paper 8) (“Allowing similar, serial challenges to the same patent, by the
`
`same petitioner, risks harassment of patent owners and frustration of Congress’s
`
`intent in enacting the Leahy-Smith America Invents Act.” (citing H.R. Rep. No.
`
`112-98, pt. 1, at 48 (2011)).
`
` Accordingly, the Tribe asks the Board to exercise its discretion under § 314(a)
`
`to deny this Petition because of these inefficiencies.
`
`C. The Board should deny institution because of the Tribe’s status as a
`sovereign.
`The Tribe is a federally recognized, American Indian Tribe and owner of the
`
`’800 patent that is the subject of this proceeding. The Tribe, as a sovereign
`
`government, is not amenable to suit unless it expressly consents or Congress
`12
`
`
`
`

`

`abrogates its immunity. See, e.g., Michigan v. Bay Mills Indian Cmty., 134 S. Ct.
`
`2024, 2030 (2014).
`
`Last summer, in Saint Regis Mohawk Tribe v. Mylan Pharm. Inc., 896 F.3d
`
`1322 (Fed. Cir. 2018), the Federal Circuit held that sovereign immunity cannot be
`
`asserted in an IPR because an “IPR is more like an agency enforcement action than
`
`a civil suit brought by a private party.” Id. at 1327. The Tribe believes that case
`
`was wrongly decided and filed a petition for writ of certiorari on December 20,
`
`2018 that asks the Supreme Court to decide whether sovereign immunity may be
`
`asserted in IPRs before the Patent Trial and Appeal Board. EX. 2049.
`
`But the Federal Circuit also held that the USPTO “Director bears the political
`
`responsibility of determining which cases should proceed.” Saint Regis Mohawk
`
`Tribe, 896 F.3d at 1327 (emphasis added). And that the Director can deny a
`
`petition for IPR “based on a party’s status as a sovereign.” Id. Accordingly, the
`
`Tribe respectfully requests that the Director exercise his discretion to deny this
`
`Petition based on the Tribe’s status as a sovereign.
`
`IV. TECHNOLOGY BACKGROUND
`A. Conventional Computer Architecture.
`Conventional computers utilize general purpose processors from Intel or AMD
`
`and employ a Von Neumann architecture. In a conventional computer, “hardware
`
`is fixed and cannot be changed after manufacturing.” EX. 2046 ¶ 9. To execute a
`
`
`
`13
`
`

`

`software program, the processor “goes through a fixed routine of steps”:
`
`1. Instruction Fetch - read the instruction whose address
`is specified by the program counter into the designated
`processor internal register, and advance the program
`counter to point to the next instruction.
`2. Instruction Decode - Decipher the work needed by
`the instruction.
`3. Execute - carry out the work needed if data is
`available internally, if not then prepare the address of the
`data.
`4. Data Memory Access - read/write data from/to
`memory.
`5. Write back - write the results into an internal register.
`Id. This is referred to as the fetch-execute cycle. Because of their architecture,
`
`conventional computers must operate in a sequential manner. Id. ¶ 10.
`
`B. Field Programmable Gate Arrays.
`A field programmable gate array (“FPGA”) is a reprogrammable integrated
`
`circuit that contains an array of configurable logic blocks (functional units)
`
`connected by configurable interconnects. EX. 2046 ¶ 11. The user can configure
`
`the FPGA to perform a desired computation by configuring (or instantiating) the
`
`configurable logic blocks to perform the desired operations (arithmetic, logical,
`
`control, data movement, etc.) and then configuring the interconnects so that the
`
`configured logic are connected in the order needed to perform the desired
`
`computation. Id. An FPGA is configured by loading a file called a bitstream into
`
`
`
`14
`
`

`

`the FPGA. Id.
`
`C. Reconfigurable Computing.
`Reconfigurable computing systems are built from reconfigurable computing
`
`devices, such as FPGAs, that serve as coprocessors to microprocessors. EX. 2054
`
`at 1; EX. 2037 at 1. “In its simplest terms, reconfigurable computing, based on
`
`FPGA technology, could be defined as the capability of reprogramming hardware
`
`to execute logic that is designed and optimized for a specific user’s algorithms.”
`
`EX. 2037 at 2.
`
`With an FPGA one can implement only the hardware that is needed and can
`
`avoid many of the slowdowns that come with load-decode-fetch-execute sequences
`
`in traditional Von Neumann processors.
`
`Much of the speedup from FPGAs comes from the fact that intermediate results
`
`do not need to be stored back in memory. Instead they live on the wires
`
`(reconfigurable routing resources) inside the FPGA as they flow from one
`
`processing element (i.e., configurable logic blocks) to another. This is the key for
`
`processing performance gains on reconfigurable computing systems. FPGAs can
`
`instantiate many processing elements, which allows them to perform many
`
`computations before having to store the results back to external memory. If one has
`
`to compute, store externally, and then immediately fetch back, one loses.
`
`In January 2004, Duncan Buell, an editor and author of the Splash2 reference
`
`
`
`15
`
`

`

`and the project manager for the Splash 2 system (EX.1007), co-authored an article
`
`about implementing Defense Advanced Research Projects Agency (DARPA)
`
`benchmark code on SRC Computer’s SRC-6 reconfigurable computer. EX. 2035 at
`
`1. That article described the successful implementation of this benchmark code on
`
`the SRC-6 as “a watershed event in the history of computing” and proclaimed that
`
`the “era of effective programming of a reconfigurable computer has arrived.” Id. at
`
`5.
`
`In 2008, Dr. Buell co-authored another article about SRC Computer’s SRC-6
`
`entitled “The Promise of High-Performance Reconfigurable Computing.” EX.
`
`2054 at 2. That article found that high-performance reconfigurable computers
`
`(HPRCs) “can achieve up to four orders of magnitude improvement in
`
`performance, and up to three orders of magnitude reduction in power consumption,
`
`and two orders of magnitude savings in cost and size requirements compared with
`
`contemporary microprocessors when running compute-intensive applications based
`
`on integer arithmetic.” Id. at 8.
`
`This shows that SRC Computers was a pioneer and leader in the HPRC field.
`
`D. The ’800 patent: SRC invents methods for enhancing parallelism and
`performance in reconfigurable computing systems.
`The ’800 patent claims techniques for enhancing parallelism and performance
`
`in reconfigurable computing systems. EX. 1005 at col. 1:39-43. At the time of the
`
`invention, “most large software applications achieve[d] high performance
`16
`
`
`
`

`

`operation through the use of parallel processing” that required “multiple processors
`
`to work simultaneously on the same problem.” Id. at col. 1:44-52. The problem
`
`was that “as more and more performance is required, so is more parallelism,
`
`resulting in ever larger systems” to the point that “[c]lusters exist … that have tens
`
`of thousands of processors and can occupy football fields of space.” Id. at col.
`
`1:53-58. “Systems of such a large physical size present many obvious downsides,
`
`including, among other factors, facility requirements, power, heat generation and
`
`reliability.” Id. at col. 1:58-61.
`
`The inventors of the ’800 patent realized that this problem could be solved by
`
`“a processor technology … that offers orders of magnitude more parallelism per
`
`processor.” Id. at 1:65-67. And that this type of processor technology is “possible
`
`through the use of a reconfigurable processor” because reconfigurable processors
`
`can “instantiate as many functional units as may be required to solve the problem
`
`up to the total capacity of the integrated circuit chips they employ.” Id. at col. 2:1-
`
`7.
`
`The inventors of the ’800 patent also realized that additional, and less obvious,
`
`performance gains could “also be realized by reconfigurable processors due to the
`
`much tighter coupling of the parallel functional units within each chip than can be
`
`accomplished in a microprocessor based computing system.” Id. at col. 2:18-25.
`
`The ’800 patent explains that:
`
`
`
`17
`
`

`

`In a multi-processor, microprocessor-based system, each
`processor is allocated but a relatively small portion of the
`total problem called a cell. However, to solve the total
`problem, results of one processor are often required by
`many adjacent cells because their cells interact at the
`boundary and upwards of six or more cells, all having to
`interact to compute results, would not be uncommon.
`Consequently, intermediate results must be passed
`around the system in order to complete the computation
`of the total pro