International Business Machines Corporationv.Intellectual Ventures I LLC

Patent Trial and Appeal Board

May 27, 2016

10044786 (P.T.A.B. May. 27, 2016)

Trials@uspto.gov Paper No. 43
571.272.7822 Filed: May 27, 2016

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

INTERNATIONAL BUSINESS MACHINES CORPORATION,
Petitioner,

v.

INTELLECTUAL VENTURES I LLC,
Patent Owner.
____________

Case IPR2015-00302
Patent 6,819,271 B2
____________

Before BRIAN J. McNAMARA, PATRICK R. SCANLON, and
FRANCES L. IPPOLITO, Administrative Patent Judges.

IPPOLITO, Administrative Patent Judge.

FINAL WRITTEN DECISION
35 U.S.C. § 318(a) and 37 C.F.R. § 42.73
IPR2015-00302
Patent 6,819,271 B2
2

I. INTRODUCTION
Petitioner, International Business Machines Corporation, filed a
Petition on November 25, 2014, requesting an inter partes review of claims
1–7, 10, 35–37, 40–42, 45–47, and 57–61 of U.S. Patent No. 6,819,271 B2
(Ex. 1004, “the ’271 patent”). (Paper 2, “Pet.”). Patent Owner, Intellectual
Ventures I LLC, filed a Preliminary Response to the Petition on March 4,
2015 (Paper 6, “Prelim. Resp.”).
Based on these submissions, we instituted trial as to claims 1–7, 10,
35–37, 40–42, 45–47, and 57–61 of the ’271 patent on the following
grounds:
1. Claims 1–7, 10, and 57–61 as unpatentable under 35 U.S.C. § 103
over Glover;1 and
2. Claims 35–37, 40–42, and 45–47 as unpatentable under
35 U.S.C. § 103 over Glover and Cheng.2
Paper 8, 20 (“Dec. to Inst.”). After institution, Patent Owner filed a Patent
Owner’s Response (Paper 15, “PO Resp.”), and Petitioner filed a Reply
(Paper 26, “Reply”).
Additionally, Patent Owner filed a Motion to Exclude (Paper 30,
“PO’s Mot. Exclude”) to which Petitioner filed an Opposition (Paper 36,
“Pet. Opp. Mot. Exclude.”). Patent Owner filed a Reply (Paper 38, “PO
Exclude Reply”) to Petitioner’s Opposition.
An oral hearing was conducted on February 16, 2016. A transcript of
the oral hearing is included in the record. Paper 41 (“Tr.”).
We have jurisdiction under 35 U.S.C. § 6(b). This decision is a Final
Written Decision under 35 U.S.C. § 318(a) and 37 C.F.R. § 42.73 as to the

1 U.S. Patent No. 5,412,429, issued May 2, 1995 (Ex. 1010, “Glover”).
2 U.S. Patent No. 5,608,396, issued Mar. 4, 1997 (Ex. 1023, “Cheng”).
IPR2015-00302
Patent 6,819,271 B2
3

patentability of claims 1–7, 10, 35–37, 40–42, 45–47, and 57–61. For the
reasons discussed below, Petitioner has demonstrated by a preponderance of
the evidence that claims 1–7, 10, 35–37, 40–42, 45–47, and 57–61 are
unpatentable.
A. Related Proceedings
Petitioner indicates that the ’271 patent is involved in a United States
District Court proceeding in the Southern District of New York captioned
Intellectual Ventures I LLC v. Citigroup, Inc., C.A. No. 1:14-cv-04638
(S.D.N.Y.). Pet. 1. Petitioner represents that it is not a party to that District
Court proceeding. Id. The ’271 patent is also involved in IPR2015-00303,
IPR2015-00304, and IPR2015-00305.
B. The ’271 Patent
The ’271 patent is directed generally to systems and methods for
performing data compression and decompression using a plurality of data
compression/decompression engines in parallel. Ex. 1004, Abstract. Figure
1 of the ’271 patent is reproduced below.
IPR2015-00302
Patent 6,819,271 B2
4

Figure 1 shows splitting logic 240, parallel compression engines 570 that
each may implement a parallel lossless data compression algorithm, and
merging logic 260. Ex. 1004, 7:54–58, 8:6–10. As shown, uncompressed
data 230A may be provided to an input of splitting logic 240. Id. at 8:31–32.
Splitting logic 240 may divide uncompressed data 230A into a plurality of
parts/portions with one portion for each of the compression engines 570
coupled to splitting logic 240. Id. at 8:32–36. Splitting logic 240 may
provide the plurality of portions of uncompressed data 230A to compression
engines 570. Id. at 8:36–38. Compression engines 570A–D may compress
the received data concurrently in parallel. Id. at 8:43–52. The ’271 patent
further discloses that the parallel compression engines 570A–D may each
have its own history buffer or share a common history buffer. Id. at 8:62–
65. The compressed data portions then pass from compression engines
570A–D to merging logic 260 wherein the portions are combined into
compressed data 270A. Id. at 8:66–9:1.
IPR2015-00302
Patent 6,819,271 B2
5

The ’271 patent also describes systems and methods for performing
data decompression. Figure 2 is reproduced below.

Figure 2 shows splitting logic 280, parallel decompression engines 550A–D,
and merging logic 290. Ex. 1004, Fig. 2. Parallel decompression engines
550A–D receive a portion of compressed data 270B divided by splitting
logic 280. Id. at 9:66–10:14. Each of decompression engines 550A–D may
then decompress the particular portion of the compressed data that was
provided to it by splitting logic 280 to produce an uncompressed portion of
the original compressed data. Id. at 10:15–22. Merging logic 290 may
receive the decompressed portions and merge these into uncompressed data
230B. Id. at 10:41–54.

C. Illustrative Claim
Of the challenged claims, claims 1, 35, 40, 45, and 57 are
independent. Claims 1 and 35, reproduced below, are illustrative of the
subject matter of the ’271 patent:
1. A data compression system comprising:
IPR2015-00302
Patent 6,819,271 B2
6

a plurality of parallel compression engines, wherein each
of the plurality of parallel compression engines operates
independently and implements a parallel data compression
algorithm;

wherein each of the plurality of parallel compression
engines is operable to:

receive a different respective portion of uncompressed
data; and

compress the different respective portion of the
uncompressed data using the parallel data compression algorithm
to produce a respective compressed portion of the uncompressed
data; and

output the respective compressed portion;

wherein the plurality of parallel compression engines are
configured to perform said compression in a parallel fashion to
produce a plurality of respective compressed portions of the
uncompressed data.

35. A memory controller, comprising:

memory control logic for controlling a memory; and

a plurality of parallel compression engines, wherein each
of the plurality of parallel compression engines operates
independently and implements a lossless parallel data
compression algorithm;

wherein each of the plurality of parallel compression
engines is operable to:

receive a different respective portion of uncompressed
data; and

IPR2015-00302
Patent 6,819,271 B2
7

compress the different respective portion of the
uncompressed data using the parallel data compression algorithm
to produce a respective compressed portion of the uncompressed
data; and

output the respective compressed portion;

wherein the plurality of parallel compression engines are
configured to perform said compression in a parallel fashion to
produce a plurality of respective compressed portions of the
uncompressed data;

wherein the respective compressed portions output from the
plurality of parallel compression engines are combinable to form
compressed data corresponding to the uncompressed data.

II. ANALYSIS
A. Claim Construction
In an inter partes review, “[a] claim in an unexpired patent shall be
given its broadest reasonable construction in light of the specification of the
patent in which it appears.” 37 C.F.R. § 42.100(b); see also In re Cuozzo
Speed Techs., LLC, 793 F.3d 1268, 1277–79 (Fed. Cir. 2015) (“We conclude
that Congress implicitly approved the broadest reasonable interpretation
standard in enacting the AIA” and “the standard was properly adopted by
PTO regulation.”), cert. granted sub nom. Cuozzo Speed Techs., LLC v. Lee,
136 S. Ct. 890 (mem.) (2016). There is a presumption that a claim term
carries its ordinary and customary meaning. CCS Fitness, Inc. v. Brunswick
Corp., 288 F.3d 1359, 1366 (Fed. Cir. 2002); In re Translogic Tech., Inc.,
504 F.3d 1249, 1257 (Fed. Cir. 2007).
1. “independently” (claims 1, 35, 40, 45, and 57)
In the Petition, Petitioner proposes that “independent” means “without
depending on others.” Pet. 4–5. For the Decision to Institute, we agreed
IPR2015-00302
Patent 6,819,271 B2
8

with Petitioner’s proposal and adopted that construction for the purposes of
the Decision to Institute. Dec. to Inst. 7–8.
At this juncture, Patent Owner does not oppose our construction or
propose an alternative construction. PO Resp. 7, 44. Nonetheless, we
observe that Patent Owner argues that Glover does not teach parallel
compression engines that operate “independently” because one of ordinary
skill in the art would understand Glover to use a shared dictionary and
lookup table. PO Resp. 56 (“Glover does not disclose compression engines
that ‘operate independently’ because the compression engines in Glover
depend on each other to build shared look up tables, in the case of Huffman
coders, and shared dictionaries, in the case of Lempel-Ziv coders.”). To the
extent that Patent Owner relies on a construction of “independently” that
excludes the use of shared dictionaries, we determine that Patent Owner has
not presented any persuasive explanation or evidence that supports this
position, nor will we endeavor to scour the record to generate these
arguments for Patent Owner.3 Thus, based on the complete record before us,
we discern no reason to alter our construction.

3 We note the Petition refers to the prosecution history of the ’271 patent
and observes that
[a]s an example of how the engines of their claimed invention
operate independently, Applicants stated that ‘they do not use a
logically shared dictionary, but have their own independent
dictionaries which do not affect the other dictionaries.’ (Ex. 1018
at 6-7.) Applicants repeated these arguments when appealing the
rejection of the claims. (Ex. 1020 at 7-11.) In finally allowing the
claims, the Examiner stated that ‘it is now recognized that the
compression engine’s separate history buffers provide
reasonable support for the independent operation asserted in the
remarks.’ (Ex. 1021 at 2.)
IPR2015-00302
Patent 6,819,271 B2
9

Moreover, we observe that the Specification does not expressly define
“independently.” Under such circumstances, “in determining the ordinary
and customary meaning of the claim term as viewed by a person of ordinary
skill in the art, it is appropriate to consult a general dictionary definition of
the word for guidance.” Comaper Corp. v. Antec, Inc., 596 F.3d 1343, 1348
(Fed. Cir. 2010) (citing Phillips v. AWH Corp., 415 F.3d 1303, 1322–23
(Fed. Cir. 2005) (en banc)). One dictionary provided by Petitioner,
Webster’s II New College Dictionary, defines “independent” as “[f]ree from
the influence, guidance, or control of another or others.” Ex. 1013, 4.
Accordingly, we maintain that the plain and ordinary meaning of
“independently” means “without depending on others.”
2. “in parallel fashion” (claims 1, 35, 40, 45, and 57)
For the Decision to Institute, we construed the claim phrase “in a
parallel fashion” to mean “in a manner that is parallel.” Dec. to Inst. 8.
Patent Owner does not object to this construction. PO Resp. 7–8. Based on
the complete record before us, we maintain that the term “in a parallel
fashion” means “in a manner that is parallel” under the broadest reasonable
interpretation. This construction is consistent with the Specification, which
describes the parallel engines 550A–D, 570A–D shown in Figures 1 and 2 as
performing “in . . . parallel fashion.” Ex. 1004, 8:46–49, 10:22–25.
Accordingly, based on the entire record, we discern no reason to alter our
construction.

Pet. 5. Patent Owner has not proposed a construction based on
prosecution history disclaimer. PO Resp. 7, 44.
IPR2015-00302
Patent 6,819,271 B2
10

3. “concurrently” (claim 2, 36, 41, 46, and 57)
For the Decision to Institute, we determined that the claim phrase
“concurrently” should be construed as “occurring at the same time.” Dec. to
Inst. 8. Patent Owner does not object to this construction. PO Resp. 8–9.
Accordingly, based on the entire record, we discern no reason to alter our
construction of “concurrently” as “occurring at the same time.”
B. Claims 1–7, 10, and 57–61 — Asserted Obviousness over Glover (Ex.
1010)
1. Relevant Legal Principles
A claim is unpatentable under 35 U.S.C. § 103(a) if the differences
between the claimed subject matter and the prior art are such that the subject
matter, as a whole, would have been obvious at the time the invention was
made to a person having ordinary skill in the art to which said subject matter
pertains. KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 406 (2007). The
question of obviousness is resolved on the basis of underlying factual
determinations including (1) the scope and content of the prior art; (2) any
differences between the claimed subject matter and the prior art; (3) the level
of skill in the art; and, (4) where in evidence, so-called secondary
considerations, including commercial success, long-felt but unsolved needs,
failure of others, and unexpected results. Graham v. John Deere Co., 383
U.S. 1, 17–18 (1966) (“the Graham factors”). The level of ordinary skill in
the art usually is evidenced by the references themselves. See Okajima v.
Bourdeau, 261 F.3d 1350, 1355 (Fed. Cir. 2001); In re GPAC Inc., 57 F.3d
1573, 1579 (Fed. Cir. 1995); In re Oelrich, 579 F.2d 86, 91 (CCPA 1978).
For an obviousness analysis, prior art references must be “considered
together with the knowledge of one of ordinary skill in the pertinent art.” In
re Paulsen, 30 F.3d 1475, 1480 (Fed. Cir. 1994) (quoting In re Samour, 571
IPR2015-00302
Patent 6,819,271 B2
11

F.2d 559, 562 (CCPA 1978)). Moreover, “it is proper to take into account
not only specific teachings of the reference but also the inferences which one
skilled in the art would reasonably be expected to draw therefrom.” In re
Preda, 401 F.2d 825, 826 (CCPA 1968). That is because an obviousness
analysis “need not seek out precise teachings directed to the specific subject
matter of the challenged claim, for a court can take account of the inferences
and creative steps that a person of ordinary skill in the art would employ.”
KSR, 550 U.S. at 418; see also Translogic Tech., Inc., 504 F.3d at 1259.
2. Summary of Glover (Ex. 1010)
Glover discloses a data compression coder using subband/transform
coding with a Lempel-Ziv based coder. Ex. 1010, Title. Figure 1 of Glover
is reproduced below.

Figure 1 shows a block diagram of a lossy coder. Ex. 1010, 3:11. In
Figure 1, data is fed into subband transform 4. Id. at 5:16–17. Subband
transform 4 separates the signal into a plurality of subbands 100, 200, 300
and 400. Id. at 5:17–19. Lowband signal 100 is first fed into Digital Pulse
Code Modulator (DPCM) 102, which then produces DPCM subband 103.
Id. at 5:19–21. DPCM subband and the other subbands are then fed into a
IPR2015-00302
Patent 6,819,271 B2
12

plurality of quantizers 105, 205, 305, 405, respectively. Id. at 5:22–24.
Quantizers 105, 205, 305, 405 reduce the variability in the signal by
removing some of the data thereby producing quantized signals 110, 210,
310, and 410 with less variation. Id. at 5:24–27. This data is then fed into a
plurality of run-length coders 115, 215, 315, 415, which group the data into
values thereby producing grouped signals 120, 220, 320, 420. Id. at 5:27–
30. Grouped signals 120, 220, 320, 420 may be compressed using Huffman
coders and Lempel-Ziv coders. Id. at 5:30–38. These compressed subbands
then can be used as inputs to data commutator/transmitter 6, which combines
the subbands and produces coded signal output at 8. Id. at 5:39–6:3.
Glover further discloses decompression in Figure 2. Figure 2 is
reproduced below.

Figure 2 is a block diagram of a lossy decoder. Ex. 1010, 3:13. In Figure 2,
coded data signal 8 is found into data receiver and decommutator 12, which
produces subband signal outputs 135, 235, 335, 435. Id. at 6:4–7. These
subband signal outputs are fed into a plurality of Lempel-Ziv decoders 140,
240, 340, 440 and other decoders. Id. at 6:7–18. The signals then are
IPR2015-00302
Patent 6,819,271 B2
13

combined through inverse subband transform 14 to produce decoded signal
16. Id. at 6:23–25.
3. Analysis
Petitioner challenges claims 1–7, 10, and 57–61 under 35 U.S.C.
§ 103 based on Glover. Pet. 38–47, 53–58. Patent Owner contests
Petitioner’s position. PO Resp. 8–60. As explained below, we have
considered the arguments and evidence presented by both parties, and we
determine Petitioner has shown by a preponderance of the evidence that
claims 1–7, 10, and 57–61 would have been obvious over Glover.
a. Claims 1–3, 5–7, 10, 57, and 59–61
Below we discuss claim 1, which is illustrative of the subject matter
of claims 2, 3, 5–7, 10, 57, and 59–61.
Claim 1 is directed to a data compression system that includes “a
plurality of parallel compression engines.”
Petitioner asserts that Glover teaches a “plurality of parallel
compression engines” because Glover discloses a quantizer and Run Length,
Huffman, and Lempel-Ziv coders for each of subbands 0–3. Pet. 39 (citing
Ex. 1010, Fig. 1, 5:22–37); see Reply 2. Petitioner further asserts these
coders perform compression of the respective subband. Pet. 39 (citing Ex.
1010, 5:14–15, 5:27–38).
Patent Owner responds that Petitioner’s arguments inconsistently
include or exclude the quantizers and/or DPCM as part of the “parallel
compression engine.” PO Resp. 16–20. Patent Owner argues “[r]egardless
of which interpretation Petitioner presents, Glover’s use of the DPCM
process and quantization process is essential to the efficiency of Glover’s
compression algorithm.” Id. at. 18. Patent Owner adds that Glover’s
“passing mention of a lossless compression possibility” does not explain
IPR2015-00302
Patent 6,819,271 B2
14

why or how a skilled artisan would exclude the DPCM or quantizer. Id. at
19 (citing Ex. 1010, 1:63–2:8); see Exhibit 2001 ¶ 73.
Based on the complete record, Petitioner’s arguments are persuasive.
On page 39 of the Petition, Petitioner provides an annotated version of
Glover’s Figure 1 (shown below). Pet. 39.

Petitioner explains that “[i]n annotated Figure 1 . . . , each compression
engine is its own color.” Id. at 39. In its Reply, Petitioner confirms that
“[i]n Glover, the parallel compression engines are comprised of the
quantizers and the Run-Length, Huffman, and Lempel-Ziv coders.” Reply
2. Further, Petitioner argues that the parallel compression algorithm is
executed by the coders, but not the quantizer. Reply 8 (“In the case of
Glover, the PCEs implement a PCA by pipelining the Run-Length, Huffman,
and Lempel-Ziv coders. They also implement a quantizer.”).
In other words, Petitioner does not exclude or ignore the DPCM and
quantizer processes. Instead, Petitioner asserts that the DPCM and
quantizers process “uncompressed data in some way, but it is the Run-
Length, Huffman, and Lempel-Ziv coders in Glover that perform
IPR2015-00302
Patent 6,819,271 B2
15

compression.” Id. at 1–2. Based on the complete record, including
Petitioner’s explanation of Glover’s Figure 1, Petitioner’s arguments are
persuasive and explain sufficiently how Glover discloses a “plurality of
parallel compression engines.”
Claim 1 further recites that “each of the plurality of parallel
compression engines operates independently.”
For this limitation, Petitioner argues that each of Glover’s
compression engines operates independently of the other engines to process
and compress a specific subband of a video signal. Pet. 40 (citing Ex. 1010,
5:17–19, 5:38–39, Fig. 1). Petitioner relies on the testimony of Dr. James A.
Storer to support the proposition that the purpose of separating the video
signal into subbands, as disclosed by Glover, is to allow each subband to be
processed independently. Id. (citing Ex. 1001 ¶¶ 222–223).
Patent Owner counters that Petitioner’s arguments are unsupported by
Glover’s disclosure. In particular, Patent Owner argues that Figure 1 and
column 5, lines 17–19 and 38–39 of Glover do not teach compression
engines that do not share resources or that have respective inputs and
outputs. PO Resp. 45–47. Patent Owner adds that there are “many purposes
for separating data into subbands,” and that US Patent No. 5,729,228
(Exhibit 2009, “Franaszek”), which was cited during the prosecution of the
’271 patent, teaches that “someone skilled in the art would implement
Glover with Franaszek’s shared dictionaries.” Tr. 47:14–18; see also PO
Resp. 47 (citing Ex. 2009, Fig. 2, 2:66–3:4) (“[E]ven if data is separated into
subbands . . . it would be advantageous for the compression engines to
depend on each other”).
Further, Patent Owner asserts that the primary objective of Glover is
“increased compression efficiency,” and that independent operation of
IPR2015-00302
Patent 6,819,271 B2
16

Glover’s compression engine would not achieve this purpose because a
skilled artisan would “use a single, shared dictionary for Lempel-Ziv coding
across the subbands.” PO Resp. 47–48, 54 (citing Ex. 2001 ¶¶ 115, 117).
Patent Owner argues that Glover does not disclose compression engines that
“operate independently” because the compression engines in Glover depend
on each other to build shared look up tables, in the case of Huffman coders,
and shared dictionaries, in the case of Lempel-Ziv coders. Id. at 56–57
(citing Exhibit 2001 ¶ 122).
Patent Owner also relies on the testimony of Dr. Richardson, (PO
Resp. 48–53), which states that Dr. Richardson “attempted to model the
preliminary processing stages of Glover” by creating
a Matlab script file (a series of commands), subband_example.m.
Exhibit 2004. This script file makes use of a public-domain
Matlab implementation of a Discrete Wavelet Transform, which
is a transform that separates a signal into subbands, available at
the following web address:
http://fresourcecode.net/sites/default/files/57785.zip. The
specific modules that are called by the subband_example.m
script file are collected as Exhibit 2007.
Ex. 2001 ¶¶ 45, 52. Based on Dr. Richardson’s model, Patent Owner argues
that the purpose of Glover’s processing stages is to produce a similar
statistical profile across each subband. PO Resp. 54 (citing Ex. 2001 ¶ 117).
Patent Owner further argues that “[b]oth Huffman coding and Lempel-Ziv
coding provide more efficient compression performance with larger sets of
statistically similar data.” Id. (citing Ex. 2001 ¶ 117). Patent Owner then
concludes that a skilled artisan would have used a single, shared lookup
table for Huffman coding and a single, shared dictionary for Lempel-Ziv
coding across subbands for efficiency. Id. at 54–57. Finally, Patent Owner
adds that separate dictionaries in Glover would reduce compression
IPR2015-00302
Patent 6,819,271 B2
17

efficiency and unnecessarily increase hardware and implementation
complexity. Id. at 55 (citing Ex. 2001 ¶ 119).
Based on the entire record, we agree with Petitioner’s position that
Glover teaches or suggests “each of the plurality of parallel compression
engines operates independently.” We credit the testimony of Dr. Storer,
who states that Glover teaches “dividing” a video signal into four subbands
and then feeding each subband into a different compression engine “to allow
compression of each subband to occur independently of the other subbands.”
Ex. 1001 ¶¶ 222–223; see Ex. 1024, 171:6–16. Dr. Storer’s testimony is
consistent with and supported by Glover’s disclosure, which provides that an
incoming video signal is separated into subbands, and each subband is
processed (compressed or decompressed) by separate sets of
coders/decoders (i.e., “engines”). Ex. 1010, 5:17–38, 6:4–21, Figs. 1, 2. For
example, in Figure 1 and column 5, Glover teaches that the subband
transform separates the signal into subbands 100, 200, 300, and 400. Ex.
1010, 5:17–19, Fig. 1. Each of these subbands is fed separately (e.g., input)
into a quantizer and coders, and then compressed and output as separate
compressed subbands 130, 230, 330, and 430 that are combined as a coded
signal. Ex. 1010, 5:19–6:3.
Additionally, we agree with Petitioner’s argument that Glover also
“discloses treating each subband independently by identifying ‘four
quantizer designs,’ and tailoring the quantizer used in each subband to the
data in that subband to balance improved compression with image quality.”
Reply 20 (citing Ex. 1010, 2:55–62; 6:35–37, claims 3–4; Ex. 1029 ¶ 33).
For example, Glover shows in Table I that the design of the quantizer for
each band is different based on the subband. Ex. 1010, Table I, 2:55–62
(describing “fine,” “coarse,” and “very coarse” quantizers).
IPR2015-00302
Patent 6,819,271 B2
18

Furthermore, we do not agree with Patent Owner that one of ordinary
skill in the art must read Glover as using a shared dictionary/lookup table for
compression. First, Glover discloses the use of a “dictionary based
compression scheme,” but does not describe this “scheme” as using a shared
dictionary or a lookup table. Ex. 1010, 3:63–66; see Tr. 51:16–19. Second,
with respect to Patent Owner’s reliance on Dr. Richardson’s modeling of
Glover’s system, (PO Resp. 48–54), Dr. Richardson states that his
simulation is “illustrative in nature . . . and [is] not intended to be an exact
implementation of Glover.” Ex. 2001 ¶ 46 (emphasis added). While Dr.
Richardson’s model is illustrative, when considered in the context of the
complete the record before us, we are persuaded that Petitioner has the better
position, especially in light of Glover’s actual disclosure that describes,
among other things, separate subbands compressed by separate sets of
coders and a specialized quantizer for each subband. Ex. 1010, Fig. 1, Table
I.
Third, Patent Owner’s reliance upon Franaszek is unpersuasive
because, as Petitioner pointed out during Oral Hearing, Franaszek was filed
after Glover issued on May 2, 1995. Ex. 2009 (filing date July 6, 1995). We
are not persuaded by Patent Owner’s argument that Franaszek’s later-filed
disclosure informs how one of ordinary skill in the art would have
understood Glover’s compression engines at the time of Glover’s invention.
Nevertheless, even considering Franaszek’s disclosure of a shared
dictionary, we are not persuaded that the term “operates independently”
excludes the use of a shared dictionary. Our claim construction does not
require this exclusion and Patent Owner has not presented any argument,
evidence, or explanation supporting an alternative construction. PO Resp. 7,
44; see supra Claim Construction.
IPR2015-00302
Patent 6,819,271 B2
19

Finally, Dr. Richardson’s testimony that separate dictionaries in
Glover would reduce compression efficiency and unnecessarily increase
hardware and implementation complexity is unpersuasive. See Ex. 2001
¶ 119; Tr. 43:18–44:8 (“[T]o achieve this maximum efficiency and
compression with the statistically similar data across all those subbands, you
would use shared lookup tables for each of the subbands and shared
dictionaries for each of the subbands”). Dr. Richardson’s testimony
regarding the difficulties and efficiencies of implementing separate
dictionaries is unpersuasive as our construction of “operates independently”
does not require a particular type of dictionary scheme (e.g., separate
dictionaries). See supra Claim Construction. Moreover, even considering
Dr. Richardson’s testimony in this regard, we observe that during cross-
examination, Dr. Richardson conceded that it was within the skill of one of
ordinary skill in the art to implement Glover with a separate “tree” for each
Huffman coder and a separate dictionary for each Lempel-Ziv coder. Ex.
2014, 195:22–197:21. Accordingly, we are persuaded that Petitioner has
explained sufficiently how Glover discloses “each of the plurality of parallel
compression engines operates independently” as required by claim 1.
Claim 1 further requires that each of the plurality of parallel
compression engines “implements a parallel data compression algorithm.”
For this limitation, Petitioner argues that one of ordinary skill in the
art would have recognized that in each compression engine in Glover, the
compression algorithm sequence would have used “pipelining” to reduce
processing time. Pet. 40–41 (citing Ex. 1001 ¶¶ 225–36; Ex. 1010, 5:27–38,
Fig. 1). Petitioner asserts “[p]ipelining is a well-known form of parallel
processing that increases the speed at which data can be processed.” Pet. 41
(citing Ex. 1011, 436; Ex. 1001 ¶¶ 228–230).
IPR2015-00302
Patent 6,819,271 B2
20

Further, based on Dr. Storer’s testimony, Petitioner provides an
example of how one of ordinary skill in the art would have understood
pipelining in Glover. Id. (citing Ex. 1001 ¶¶ 231–236). In this example,
Petitioner notes that
[t]his parallel implementation of the compression algorithm
sequence disclosed in Glover would have been plainly obvious
to one of ordinary skill in the art, and would have proceeded as
follows:
First, each subband of a first video signal would have been
processed by the Run Length coders. (Ex. 1001 ¶¶ 231-232.)
Next, each subband of the first video signal would have been
processed by the Huffman coders. Simultaneously, each subband
of a second video signal would have been processed by the Run
Length coders. (Ex. 1001 ¶ 233.)
Next, each subband of the first video signal would have
been processed by the LZ coders. Simultaneously, each subband
of the second video signal would have been processed by the
Huffman coders, and each subband of a third video signal would
have been processed by the Run Length Coder. (Ex. 1001 ¶ 234.)
This process would have continued until all video signals
exited the pipeline. (Ex. 1001 ¶¶ 235-236.)
Id.; see Tr. 19:10–20:17. Petitioner concludes that “[t]his basic pipelining
approach would have provided an efficient way to compress video signals
using a parallel compression algorithm.” Pet. 41 (citing Ex. 1001 ¶¶ 228–
230).
In response, Patent Owner argues that Glover’s disclosure does not
teach pipelining or “motivate one of ordinary skill in the art to seek methods
of concurrent implementation or efficient parallel processing.” See PO
Resp. 32–33 (“Nowhere does Glover direct the reader towards seeking
parallel implementation efficiencies.”). Patent Owner argues that the most
straightforward way to implement Glover is through serial algorithms. Id. at
IPR2015-00302
Patent 6,819,271 B2
21

36. Patent Owner adds that pipelining Glover would require “significant”
storage and “complex additional hardware” such as multiple buffer
memories and synchronization mechanisms. Id. at 37–41 (citing Ex. 2001
¶¶ 94–108, 110). Patent Owner further argues that Petitioner has not
explained how a skilled artisan would address these problems, which are
“beyond the capabilities of what Petitioner has alleged is one of ordinary
skill in the art in the context of pipelining Glover, and even further, any
solution would require significant experimentation.” Id. at 41 (citing Ex.
2001 ¶ 109; Ex. 1011, 520).
Petitioner counters that pipelining was a well-known method of
parallel processing and a skilled artisan would have been motivated to
improve the speed and efficiency of Glover by pipelining the operation of
the coders. Reply 10–11 (citing Ex. 1001 ¶¶ 228–230; Ex. 1011, 436; Ex.
2014, 207:23–208:5); Tr. 68:1–11 (“[A] component of efficiency is speed.”).
Petitioner further argues that Patent Owner overstates the difficulties of
using pipelining with Glover, and this modification would have been within
the abilities of a skilled artisan. Reply 12–14 (citing Ex. 1029 ¶¶ 10–17).
Petitioner’s arguments are persuasive. There is no dispute among the
parties that pipelining was known at the time of the invention. Ex. 1001
¶ 228; Ex. 2014, 207:23–208:5. More particularly, Patent Owner’s
declarant, Dr. Richardson, acknowledges in his cross-examination testimony
that, by 1999, there were existing compression systems that implemented
pipeline processing. Ex. 2014, 207:23–208:5. Thus, we agree with
Petitioner that the background knowledge of a skilled artisan in the relevant
timeframe included pipelined compression systems. See Reply 10–11.
Further, we agree with Petitioner that one of ordinary skill in the art
would have understood that a “basic pipelining approach [in Glover] would
IPR2015-00302
Patent 6,819,271 B2
22

have provided an efficient way to compress video signals using a parallel
compression algorithm.” Pet. 41 (citing Ex. 1001 ¶¶ 228–230). In this
regard, Dr. Storer’s testimony is persuasive. Specifically, Dr. Storer
explains that
[i]n a pipeline, a set of processing elements in a series are ordered
such that the output of one is the input to the next. Each of these
elements is generally called a “stage” of the pipeline. As data (or
a car, in the case of the Ford assembly line) moves through the
pipeline, each processing element operates at the same time to
perform its function. Then, each processing element passes the
data to the next processing element. In this way, the processing
occurs in parallel.
Ex. 1001 ¶ 229. Dr. Storer further explains persuasively how Figure 1 of
Glover teaches or suggests a pipeline structure having a set of processing
elements in a series (e.g., coders) in which data output from one processing
element is input into the next processing element. Ex. 1001 ¶¶ 229, 231–
236. Additionally, we agree with Petitioner that compression speed factors
into efficiency, especially for low data rates where a longer compression
time may further delay transmission. See Tr. 39:13–20, 68:1–11; Ex. 1001 ¶
230 (“Pipelining increases the speed at which data can be processed.”).
Moreover, we are not persuaded that modifying Glover to include
pipelining would have presented significant or insurmountable challenges.
See PO Resp. 41. Dr. Richardson acknowledges that any implementation of
Glover’s coders would require storage/memory, and that generally one of
ordinary skill in the art would have be able to implement Glover’s system in
hardware with a few “design choices.” Ex. 2014, 159:22–163:15.
Additionally, we are persuaded by Dr. Storer’s testimony that existing
coders at the time of the invention included methods for dealing with
variable rates of compression. Ex. 1029 ¶¶ 13–14. Thus, to the extent that
IPR2015-00302
Patent 6,819,271 B2
23

synchronization presented a technical challenge, we agree with Petitioner
that resolving such challenges was well within the abilities of one of
ordinary skill in the art.
Accordingly, we are persuaded that Petitioner has explained
persuasively how Glover discloses that each of the plurality of parallel
compression engines “implements a parallel data compression algorithm.”
Claim 1 further recites
wherein each of the plurality of parallel compression
engines is operable to:
receive a different respective portion of uncompressed
data; and
compress the different respective portion of the
uncompressed data using the parallel data compression algorithm
to produce a respective compressed portion of the uncompressed
data; and
output the respective compressed portion.
For these limitations, Petitioner asserts Glover discloses that
uncompressed video signal 2 is separated into four subbands 0–3 where each
subband is provided to a set of compression encoders. Pet. 42 (citing Ex.
1010, 5:16–30, Fig. 1). Petitioner argues the four sets of encoders each
receive a different subband of the uncompressed video signal. Id. Petitioner
further argues that Glover discloses that each set of encoders compresses its
respective portion of the uncompressed video signal using a sequence of
compression algorithms to produce a compressed portion of the video signal.
Id. at 42–43 (citing Ex. 1010, 5:38–39). Petitioner adds that Glover suggests
implementing the Run Length, Huffman, and Lempel-Ziv data compression
algorithm sequence in each set of encoders to perform compression, using
pipelining. Id. (citing Ex. 1001 ¶¶ 226–36). Petitioner argues Glover
IPR2015-00302
Patent 6,819,271 B2
24

discloses its “coding technique then produces compressed subbands at 130,
230, 330, and 430. These compressed subbands can then be used as inputs
to a data commutator/transmitter.” Id. at 43 (citing Ex. 1010, 5:38–6:3, Fig.
1).
Patent Owner responds that Glover does not disclose that each of the
plurality of parallel compression engines is operable to “receive a different
respective portion of uncompressed data” because the data received from the
subband transformer, as shown in Figure 1 of Glover, is not original,
unmodified, and uncompressed. PO Resp. 20–22. Patent Owner argues that
an example described in Dr. Glover’s technical memo (Ex. 2005) illustrates
that subbands in Glover (Ex. 1010) would have been modified and filtered.
Id. Patent Owner asserts that “the lowest subband in Fig. 5 [of Ex. 2005] is
a modified version of the original image that has been compressed and
contains only a fraction of the information contained in the original image.”
Id. at 22 (citing Ex. 2001 ¶ 78).
Patent Owner’s argument is unpersuasive. To start, the language of
claim 1 does not require original or unmodified data. Claim 1 only requires
that each of the recited compression engines is operable to receive
“respective portion of uncompressed data.” Emphasis added. Further, in
describing subband coding, Dr. Glover’s technical memo states
[t]he original signal is split up into N frequency bands
(subbands). . . . If the proper conditions are met, the resulting
total number of values in the subbands is the same as in the
original signal. No compression has been achieved yet; the
subbanding process prepares the data for compression coding.
Ex. 2005, 3 (emphases added). This is consistent with Glover’s disclosure
that incoming data is “separated into subbands,” and “[o]nce the data is
separated into these subbands it can be coded and then decoded by statisical
IPR2015-00302
Patent 6,819,271 B2
25

coders such as the Lempel-Ziv based coder.” Ex. 1010, Abstract. Thus, we
are not persuaded that the discussion in Exhibit 2005 supports Patent
Owner’s position that the subband transform disclosed in Glover (Ex. 1010)
compresses data before the data is received by the DPCM, quantizers, and
coders shown in Figure 1 of Glover.
Next, Patent Owner argues that the data output from DPCM 102 (Ex.
1010, Fig. 1) is compressed. PO Resp. 22–25. Patent Owner further argues
that Glover would be rendered inoperable for its intended purpose if subband
0 or the DPCM process were ignored. Id. at 28. Additionally, Patent Owner
asserts Dr. Richardson’s simulation of the Lena image shows that
subband 0 is now much more similar to the other bands than if
the DPCM prediction equation disclosed in Glover had not been
used. One would expect that the increased similarity of this data,
after using DPCM, would result in an increase in efficiency when
using dictionary-based compression.
Id. at 25 (citing Ex. 2001 ¶ 62).
These arguments are also unpersuasive. Patent Owner relies on Dr.
Richardson’s testimony for the proposition that DPCM is a form of
compression. Id. at 23 (citing Ex. 2001 ¶ 80). Paragraph 80 of Dr.
Richardson’s declaration states that
The DPCM process described in Glover is a well-known form of
compression. As such, the output of DPCM 102 is not
uncompressed data, but rather is transformed and DPCM
compressed data. See also Paragraphs 60-62 above (showing
how DPCM compresses data and makes the subband data for
Band 0 similar to that of Bands 1, 2, and 3).
Id. Initially, we observe that the statement “DPCM process described in
Glover is a well-known form of compression” is unsupported. Further,
looking to paragraphs 60 through 62 of Dr. Richardson’s declaration, Dr.
Richardson testifies that the DPCM process makes subband data for Band 0
IPR2015-00302
Patent 6,819,271 B2
26

similar to that of Bands 1, 2, and 3. Ex. 2001 ¶¶ 60–62. However, Dr.
Richardson does not explain how increasing data similarity is the same as
compression. See Reply 5. Moreover, this position is contradicted by
Glover’s disclosure, which teaches that compression occurs after
quantization. Ex. 1010, 5:9–15 (“After quantization, an adaptive Huffman
coder or Lempel-Ziv based coder is used to perform compression.”).
According to Glover, the DPCM process takes place before quantization.
Ex. 1010, 5:19–24.
Patent Owner further argues that the data received by the coders is not
“uncompressed” because quantizers 105, 205, 305, and 405 output
compressed data. PO Resp. 29–31. This argument is unpersuasive because
Petitioner argues that the quantizers are part of the Glover’s parallel
compression engine. Reply 7. Further, as discussed, we are persuaded that
data received by the quantizers, which is output from the subband transform
and DPCM, is uncompressed. See Ex. 1010, 5:19–24.
Claim 1 further recites “wherein the plurality of parallel compression
engines are configured to perform said compression in a parallel fashion to
produce a plurality of respective compressed portions of the uncompressed
data.” Emphasis added.
Petitioner argues that Glover discloses compression engines that are
configured in parallel “so that each PCE can compress its subband without
interfering with the other PCEs. (Ex. 1010 at Fig. 1, 5:16-39; Ex. 1001
¶¶ 218-220, 222-223, 241-243, 251.).” Pet. 44. Petitioner asserts that this
architecture allows Glover’s compression engines to operate in parallel to
compress their respective portions of the uncompressed data and produce a
respective compressed portion of that uncompressed data. Id.
Patent Owner contends Glover does not mention “parallel” or
IPR2015-00302
Patent 6,819,271 B2
27

“concurrent” operation, and that there is no evidence that the processing
chains in Figure 1 of Glover would perform compression in a parallel
fashion or concurrently. PO Resp. 58. Patent Owner further argues that
implementing parallel processing in Glover requires undue experimentation
because additional storage and processing resources would be required. PO
Resp. 58–59 (citing Ex. 2001 ¶¶ 124–125).
Petitioner responds that a skilled artisan would understand that the
architecture shown in Figure 1 of Glover depicts operation in a “parallel
fashion” or “concurrently.” Reply 22–23 (citing Ex. 1010, Fig. 1; Ex. 1029
¶¶ 34–35). Petitioner further argues that any implementation of Glover,
serial, parallel, concurrent etc., would require storage and buffers not
identified in Glover. Reply 24. Additionally, Petitioner asserts that “[e]ven
if Glover did not disclose PCEs that are in operation at the same time, a
PHOSITA would have been motivated to implement the PCEs such that they
were in operation at the same time to increase the efficiency of Glover.”
Reply 23; see id. at 24 (“the additional processing power needed to operate
the PCEs at the same time would increase the efficiency and throughput of
the system while having a minimal impact on the design. Ex. 1029 ¶¶ 10-12,
35.”).
Petitioner’s arguments are persuasive. Although Glover does not
explicitly use the terms “parallel” or “concurrently,” Figure 1 of Glover
depicts the flow of data in a parallel manner with each subband flowing
through a set of quantizers and coders. Ex. 1010, Fig. 1. The accompanying
description of Figure 1 is consistent with this reading and does not dictate
serial processing of one subband after another. For example, Glover teaches
that the lowband signal 100 is first fed into DPCM 102. Ex. 1010, 5:19–20.
However, when describing subsequent processing, Glover does not
IPR2015-00302
Patent 6,819,271 B2
28

distinguish any of the subbands as processed in any particular order. See id.
at 5:19–6:3. Moreover, we additionally agree with and adopt as our own
Petitioner’s argument that a skilled artisan would have been motivated to
implement Glover’s compression engines such that they were in operation at
the same time to increase the efficiency of Glover. See Reply 23–24.
Further, we are not persuaded by Patent Owner’s argument that
implementing parallel processing in Glover would require undue
experimentation. Here, the issue is whether the amount of required
experimentation is undue, not whether any experimentation is necessary. In
re Vaeck, 947 F.2d 488, 495 (Fed. Cir. 1991). In re Angstadt, 537 F.2d 498,
504 (CCPA 1976).
The testimony of Dr. Richardson does not explain, with sufficient
specificity, what and how much experimentation would have been required
by one with ordinary skill, and why that amount of experimentation should
be regarded as “undue.” See Ex. 2001 ¶¶ 123–127. Moreover, Dr.
Richardson agrees any implementation of Glover’s coders would require
storage/memory, and that generally one of ordinary skill in the art would
have been able to implement Glover’s system in hardware with a few
“design choices.” Ex. 2014, 159:22–163:15. Additionally, we are
persuaded by Dr. Storer’s testimony that existing coders at the time of the
invention included methods for dealing with variable rates of compression.
Ex. 1029 ¶¶ 13–14. Further, we observe that much discussion has been
made of Franaszek with regard to the use of a shared dictionary for
compression. In this context, Patent Owner has described Franaszek as
disclosing “parallel compression” with the use of individual compressors.
Tr. 46:14–47:3. Patent Owner’s reading is consistent with disclosure of
Franaszek, which is titled “Parallel Compression and Decompression Using
IPR2015-00302
Patent 6,819,271 B2
29

a Cooperative Dictionary.” Ex. 2009, Title (emphasis added). At a
minimum, Patent Owner acknowledges that the state of the prior art at the
time of the invention included parallel operation of individual compressors.
Even if the required experimentation may be complex, that does not make it
undue, if the art typically engages in such experimentation. In re Wands,
858 F.2d 731, 737 (Fed. Cir. 1988).
Accordingly, for the foregoing reasons, Petitioner has demonstrated
by a preponderance of the evidence that claim 1 is unpatentable under 35
U.S.C. § 103 over Glover.
Additionally, we determine that Petitioner provides detailed
explanations of how each limitation recited in claims 2, 3, 5–7, 10, 57, and
59–61 are obvious over Glover. Pet. 44–47, 53–58. For these claims, we
agree with and adopt Petitioner’s arguments as our own. In the Patent
Owner’s Response, Patent Owner relies on arguments presented for claim 1
that are discussed above. PO Resp. 8–60. For the reasons set forth in the
Petition (see Pet. 44–47, 53–58), we conclude that Petitioner has established
by a preponderance of the evidence that these claims are rendered obvious
by Glover. Accordingly, based on the complete record, Petitioner has
established by a preponderance of the evidence that claims 2, 3, 5–7, 10, 57,
and 59–61 are unpatentable under 35 U.S.C. § 103 over Glover.
b. Claims 4 and 58
Claims 4 and 58 depend from claims 1 and 57 respectively. Claims 4
and 58 further require a parallel lossless data compression algorithm.
For these limitations, Petitioner argues that Glover discloses parallel
compression engines that use Run Length, Huffman, and Lempel-Ziv data
compression algorithms and suggests using a parallel implementation via a
pipelined architecture. Pet. 46 (citing Ex. 1010, 5:27–38, Fig. 1; Ex. 1001
IPR2015-00302
Patent 6,819,271 B2
30

¶¶ 225–236). Petitioner argues that Glover further discloses that the Run
Length, Huffman, and Lempel-Ziv algorithms are lossless and would
implement a lossless parallel data algorithm. Pet. 46 (citing Ex. 1010, 2:1–2,
2:17–19), 52. In Petitioner’s Reply, Petitioner also argues that the claim
language at issue does not require the recited compression engines to only
implement a lossless parallel data compression algorithm. Reply 15; Tr.
20:19–22:9. Petitioner argues that Glover’s coders perform lossless
compression even if the quantizer may separately execute a lossy process.
Reply 15.
Patent Owner counters that Glover’s quantizers perform lossy not
lossless compression. PO Resp. 42–43. At the Oral Hearing, Patent Owner
argued for the first time in the proceeding that Kustom Signals, Inc. v.
Applied Concepts, Inc., 264 F.3d 1326 (Fed. Cir. 2001) stands for the
proposition that the claim language in the instant proceeding cannot be read
to include a lossy algorithm in addition to a lossless algorithm.4 Tr. 55:4–
16.
We agree with Petitioner that Glover discloses the Run Length,
Huffman, and Lempel-Ziv algorithms are lossless and would implement a
lossless parallel data algorithm. See Pet. 46 (citing Ex. 1010, 2:1–2, 2:17–
19), 52. Furthermore, based on the facts before us in the instant case,
Kustom Signals is not controlling. In Kustom Signals, the patent at issue
covered a traffic radar system having user-selectable modes of operation,
whereby the operator selects whether to identify and display the speed of
either the strongest target or the fastest target vehicle. Kustom Signals, 264

4 Patent Owner did not rely on this case prior to the Oral Hearing.
Nonetheless, in this Decision, we address Patent Owner’s arguments raised
for the first time at the Oral Hearing.
IPR2015-00302
Patent 6,819,271 B2
31

F.3d at 1329. The Federal Circuit determined that the disjunctive claim
terms “or” and “either-or” excluded an accused infringing device that
performs both a strongest and a fastest analysis of the return signal, and,
which was not subject to operator selection. Id. at 1333. Here, the claim
language before us only requires that the recited compression engines
implement a lossless parallel data compression algorithm, but does not
exclude other types of algorithms through the use disjunctive language,
which was the focus in Kustom Signals.
Accordingly, for the foregoing reasons, Petitioner has demonstrated
by a preponderance of the evidence that claims 4 and 58 are unpatentable
under 35 U.S.C. § 103 over Glover.
C. Claims 35–37, 40–42, and 45–47 — Asserted Obviousness over
Glover and Cheng (Ex. 1023)
1. Summary of Cheng (Ex. 1023)
Cheng discloses Ziv-Lempel data compression system using variable
code fields. Ex. 1023, Title. Cheng further discloses various systems that
incorporate the described compression and decompression systems. Id. at
4:50–65. As an example, Figure 11 of Cheng is reproduced below.
IPR2015-00302
Patent 6,819,271 B2
32

Figure 11 shows computer system 100 includes central processing
unit (CPU) 102 that communicates with system memory 104. Id. at 17:8–
10. Computer system 100 further includes compression/decompression
engine 124 that contains data compression engine 126 and data
decompression engine 128. Id. at 17:19–22. Cheng further teaches that
compression/decompression engine 124 includes means necessary to
implement the data compression and decompression process, which may be
embodied as program objects and data objects stored in a memory means to
direct CPU 102 to perform processes. Id. at 17:22–27. Engine 124 may use
system memory 104 or optional memory 132 while performing the desired
compression or decompression processes. Id. at 17:33–35. Additionally,
referring to Figure 13, Chen further discloses that encoder 176 incorporating
microprocessor 178 is coupled to data memory 180, which includes the
program and data objects necessary to the data compression process. Id. at
18:5–8.
Further, according to Petitioner, Cheng discloses a network device
IPR2015-00302
Patent 6,819,271 B2
33

that includes a terminal, an interface, and an encoder (compression system)
“arranged to transmit and receive data over a communications link.” Pet. 52
(citing Ex. 1023, 18:19–22; Fig. 15). This network device “operates
transparently to increase the data transmission capacity of the
communication link by maximizing the information content per bit in the
link data stream.” Id. (citing Ex. 1023, 18:23–26, Fig. 15).
2. Analysis
Petitioner asserts claims 35–37, 40–42, and 45–47 are unpatentable
under 35 U.S.C § 103 over the combination of Glover and Cheng. Pet. 47–
53. Based on the entire record, we are persuaded that Petitioner has
demonstrated by a preponderance of the evidence that claims 35–37, 40–42,
and 45–47 are unpatentable over the combination of Glover and Cheng.
Illustrative claim 35 is discussed below.
Claim 35 is directed to a memory controller having “memory control
logic for controlling a memory.” For this limitation, Petitioner asserts
Cheng discloses that a compression engine can be inserted as a memory
controller that is embodied in a memory means and uses system memory.
Pet. 48 (citing Ex. 1023, 17:22–27, 17:33–35, Fig. 11). Petitioner adds that
Cheng discloses “the computer systems discussed above in connection with
[Figures] 11–15 may be programmed or otherwise designed to facilitate” the
compression system of Cheng. Id. (citing Ex. 1023, 18:27–32). Petitioner
reasons that one of ordinary skill in the art would have understood that the
memory controller included memory control logic in order to use the system
memory, and that Cheng, thus, discloses memory control logic for
controlling a memory. Id. (citing Ex. 1001 ¶ 325).
For the remaining limitations recited in claim 35, Petitioner relies on
the disclosure in Glover that Petitioner previously argued teaches or suggests
IPR2015-00302
Patent 6,819,271 B2
34

similar limitations recited in claims 1, 3, and 4. See Pet. 48–49. Petitioner
further argues that one of ordinary skill in the art “would have understood
that the system taught by Glover could have been implemented on a network
device, memory module, or memory controller, as disclosed in Cheng, to
provide compression when storing or transmitting data, the two most
common applications of compression.” Id. at 38 (citing Ex. 1001 ¶¶ 23,
357–58).
Patent Owner does not present arguments for this ground separate
from those discussed with respect to Petitioner’s challenge based on Glover
alone. See PO Resp. 1–2, 8–60.
Based on the complete record, we agree with Petitioner’s arguments
and, for these same reasons, determine that Petitioner has demonstrated, by a
preponderance of the evidence, how the combination of Glover and Cheng
teaches or suggests teaches or suggests each limitation recited in claims 35–
37, 40–42, and 45–47. See Pet. 47–53.
Further, with respect to claims 35, 40, and 45, these claims require
lossless parallel data compression algorithm. Petitioner relies on arguments
presented for a similar limitation recited in claim 4. Pet. 45–46, 48–52.
Particularly, Petitioner argues that Glover discloses that the Run Length,
Huffman, and Lempel-Ziv algorithms are lossless and would implement a
lossless parallel data algorithm. Pet. 46 (citing Ex. 1010, 2:1–2, 2:17–19).
For the same reasons discussed with respect to claims 4 and 58, we are
persuaded by Petitioner’s arguments.
Accordingly, based on the complete record, we are persuaded that
Petitioner has demonstrated by a preponderance of the evidence that claims
35–37, 40–42, and 45–47 are unpatentable under 35 U.S.C. § 103 over the
combination of Glover and Cheng.
IPR2015-00302
Patent 6,819,271 B2
35

D. Patent Owner’s Motion to Exclude
Patent Owner filed a Motion to Exclude. Paper 30 (“PO Mot.
Exclude”). Petitioner filed an Opposition to Patent Owner’s Motion to
Exclude (Paper 36, “Pet. Exclude Opp.”), and Patent Owner filed a Reply
(Paper 38, “PO Exclude Reply”).
Patent Owner moves to exclude Exhibits 1025–1028. PO Mot.
Exclude 1–9. Patent Owner’s Motion to Exclude these exhibits is dismissed
as moot because this Decision does not rely on these exhibits.
IV. ORDER
In consideration of the foregoing, it is hereby:
ORDERED that claims 1–7, 10, 35–37, 40–42, 45–47, and 57–61 of
the ’271 patent have been shown to be unpatentable;
FURTHER ORDERED that Patent Owner’s Motion to Exclude
Exhibits 1025–1028 is dismissed as moot; and
FURTHER ORDERED that any party to the proceeding seeking
judicial review of this Final Written Decision must comply with the notice
and service requirements of 37 C.F.R. § 90.2.

IPR2015-00302
Patent 6,819,271 B2
36

PETITIONER:

Kenneth R. Adamo
Joel R. Merkin
Brent P. Ray
Eugene Goryunov
KIRLAND & ELLIS LLP

kenneth.adamo@kirland.com
joel.merkin@kirkland.com
brent.ray@kirkland.com
eugene.goryunov@kirkland.com

PATENT OWNER:

Henry A. Petri, Jr.
James P. Murphy
Margaux A. Savee
Ryan M. Murphy
POLSINELLI PC

hpetri@polsinelli.com
jpmurphy@polsinelli.com
msavee@polsinelli.com
rmurphy@polsinelli.com

Donald J. Coulman
INTELLECTUAL VENTURES I LLC

dcoulman@intven.com