VIA Alliance Semiconductor Co., Ltd.

Patent Trials and Appeals Board

Apr 20, 2021

2020002601 (P.T.A.B. Apr. 20, 2021)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
15/098,471 04/14/2016 Zhenhua HUANG 252229-3210 6443
109673 7590 04/20/2021
McClure, Qualey & Rodack, LLP
280 Interstate North Circle SE
Suite 550
Atlanta, GA 30339
EXAMINER
SAADOUN, HASSAN
ART UNIT PAPER NUMBER
2435
NOTIFICATION DATE DELIVERY MODE
04/20/2021 ELECTRONIC
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
dan.mcclure@mqrlaw.com
terri.logan@mqrlaw.com
uspatents@mqrlaw.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte ZHENHUA HUANG, YONG LI, MENGMENG YAN, and
XUEHUA HAN
Appeal 2020-002601
Application 15/098,471
Technology Center 2400
Before DEBRA K. STEPHENS, CARL W. WHITEHEAD JR., and
ADAM J. PYONIN, Administrative Patent Judges.
Opinion for the Board filed by Administrative Patent Judge STEPHENS.
Opinion dissenting filed by Administrative Patent Judge PYONIN.
STEPHENS, Administrative Patent Judge.
DECISION ON APPEAL
Pursuant to 35 U.S.C. § 134(a), Appellant1 appeals from the
Examiner’s rejection. We have jurisdiction under 35 U.S.C. § 6(b).
We AFFIRM.

1 Herein, “Appellant” refers to “applicant” as defined in 37 C.F.R.
§ 1.42(a). Appellant identifies the real party in interest as “VIA Alliance
Semiconductor Co., Ltd. (AKA Shanghai Zhaoxin Semiconductor Co.).”
Appeal Brief 2.
Appeal 2020-002601
Application 15/098,471

2
STATEMENT OF THE CASE
Introduction
The Application is directed to “a Central Processing Unit (CPU) and a
method to verify mainboard data” (Abstract). Claims 1–7 and 9–15 are
pending; claims 1 and 9 are independent (Appeal Brief, Claims Appendix).
Claim 1 is reproduced below for reference (emphases and some formatting
added):
1. A Central Processing Unit (CPU), comprises:
an on-die Read-Only Memory (ROM) to
store trusted root digest information, wherein the
trusted root digest information is not allowed to be modified; and
a core to
compute digest information of a trusted root data
stored in a mainboard using a digest algorithm when receiving a
specific instruction during a power-up process,
compare the digest information with the trusted root
digest information stored in the on-die ROM, and
perform a signature verification algorithm with the
trusted root data to verify the integrity of mainboard data if the
digest information coincides with the trusted root digest
information stored in the on-die ROM;
wherein the mainboard data comprises a ucode patch of
the CPU, and the ucode patch is signed corresponding to the
trusted root data.
References and Rejections
The Examiner relies on the following prior art:
Name Reference Date
Sutton US 2003/0196096 A1 Oct. 16, 2003
Findeisen US 2009/0222653 A1 Sept. 3, 2009
Li US 2014/0129818 A1 May 8, 2014
Miller US 9,479,340 B1 Oct. 25, 2016

Appeal 2020-002601
Application 15/098,471

3
Claims 1, 3–7, 9, and 11–15 are rejected under 35 U.S.C. 103 as being
unpatentable over Findeisen, Li, and Sutton (Final Action 3).
Claims 2 and 10 are rejected under 35 U.S.C. § 103 as being
unpatentable over Findeisen, Li, Sutton, and Miller (Final Action 8).
ANALYSIS
Claim 1 recites “an on-die Read-Only Memory (ROM) to store trusted
root digest information wherein the trusted root digest information is not
allowed to be modified; . . . ” and “ a core to . . . compare the digest
information with the trusted root digest information.”
The Examiner finds Findeisen teaches these recited elements except
for the “on-die” limitation upon which the Examiner relies on Li to teach
(Final Act. 3–5).
Appellant argues, “the Office Action fails to set [forth] the required
objective indicia appropriate to support the rejection” (Appeal Brief 13),
because the rejection “has dissected the claimed feature of ‘comparing the
digest information with trusted root digest information stored in an on-die
Read-Only Memory (ROM) of a Central Processing Unit’ and has
essentially recreated this feature by combining select (and disparate) features
from Findeisen, Li and Sutton” (Reply Brief 7).
More specifically, Appellant contends the combination of relied upon
references fails to teach “compare the digest information with the trusted
root digest information stored in the on-die ROM,” as recited in claim 1
(Appeal Br. 7). According to Appellant “what is stored in the on-die ROM
is not the trusted root data itself, but its corresponding digest information
itself” which is not taught by the references (id.).

Appeal 2020-002601
Application 15/098,471

4
Initially, we note the Specification does not define explicitly “trusted
root digest information” (see generally Spec.). Nor does Appellant identify
any construction (see generally Appeal Br.). The Specification describes
what is being stored “is just trusted root digest information with smaller size
instead of the entire trusted root data” (Spec. ¶¶ 23, 35) but provides no
details as to the size difference or content aside from discussion associated
with an embodiment of the present invention (id. ¶ 45). Thus, Appellant
does not persuade us of error in the Examiner’s interpretation.
Similarly, “trusted root data” is not defined explicitly in the
Specification. Rather, trusted root data is described as a public key in one
example (Spec. ¶ 5 (“e.g. 2048-bit RSA public key”) and “a root public key”
of a signature verification algorithm (id. ¶ 41). The Specification does not
limit “trusted root data” to being a public key, instead, explicitly stating “the
present invention is not limited thereto . . . In this case, rather than the root
public key, other trusted root data 1110 is stored in the BIOS ROM 111 for
verifying” (id. ¶ 44) (emphasis added). Thus, the Specification sets forth, in
a circular description, that trusted root data may be something other than a
public key, such as trusted root data. We find Appellant does not proffer
sufficient evidence or argument to persuade us the Examiner’s interpretation
is in error.
Appellant contends, “the hash value of Findeisen is NOT directly
stored in the on-die ROM” (Appeal Br. 7). In particular, Appellant contends
“Findeisen’s initial hash value is retrieved by the decryption keys, but not
directly stored in the internal ROM 103, while in the present invention, what
is stored [in] the on-die ROM is a hash value (the claimed trusted root
digest information) but not a decryption key” (Appeal Brief 5). Moreover,
Appeal 2020-002601
Application 15/098,471

5
Appellant contends “what is stored in Findeisen’s non-volatile memory 120
[is] only mainboard data (like portion 121 A and portion 121B) but not
‘trusted root digest information’” (Reply Brief 6 (emphasis omitted)).
Appellant’s contentions have not persuaded us of Examiner error.
The Examiner has relied on the combination of Findeisen and Li to teach the
limitation. More specifically, the Examiner relies on Findeisen’s computed
hash value to teach the computed digest information (id. (citing Findeisen
¶¶ 37–38, Fig. 1)). Findeisen teaches “[u]pon execution of the portion
121A, a hash value may also be calculated over the image 121B copied into
the system memory 110, which may then be compared with the hash value
initially contained in the portion 121A” (Findeisen ¶ 38; Final Act. 4; see
also Findeisen ¶ 43 (“The one or more calculated hash values may then be
compared to the one or more hash values contained in a dedicated location
of the data 121 or 121A”). Thus, Findeisen teaches comparing a computed
hash value with an initial hash value and thus, teaches “compar[ing] the
digest information with the trusted root digest information,” as recited in
claim 1 (Findeisen ¶¶ 38, 43).
Appellant next contends, “the hash value of Findeisen is NOT directly
stored in the on-die ROM” (Appeal Br. 7). Findeisen teaches storage of the
boot information including hash values in the non-volatile memory or
SysROM 120, a non-volatile memory, and storage of data and instructions
that serve as a static root of trust for measurement during initialization, in a
secure memory area that is a static root of trust (Findeisen ¶¶ 27, 30, 34, 36–
38). Thus, Findeisen teaches the trusted root digest information is stored in
Read-Only Memory (ROM), “wherein the trusted root digest information is
not allowed to be modified” –– it is read-only. The Examiner then relies on
Appeal 2020-002601
Application 15/098,471

6
Li to teach the on-die ROM (Final Act. 5). In particular, Li teaches “the
read-only memory 105 can be an On-die ROM” (Li ¶ 21, Fig. 1A). Thus,
we find the combination of Findeisen and Li teaches “compar[ing] the digest
information with the trusted root digest information stored in the on-die
ROM,” as recited in claim 1. Appellant’s contention that the Examiner
made conclusory assertions because the Examiner quoted disclosure of “may
be considered” from the Findeisen reference is not persuasive (Reply Br. 3).
Indeed, Appellant does not persuade us an ordinarily skilled artisan would
not understand the secure memory area is a static root of trust in an
embodiment described by Feindeisen.
Appellant further argues the Examiner has improperly combined the
references –– Findeisen, Li, and Sutton –– because the “rationales are both
incomplete and improper in view of the established standards for rejection
under 35 U.S.C. § 103” (Appeal Br. 9–10). Appellant contends the
Examiner cited paragraphs of Li and Sutton, “which point out certain
benefits of those invention”; however, according to Appellant, “those
alleged benefits have nothing to do with the feature . . . being combined into
the other reference(s)” (Appeal Br. 12).
Appellant does not persuade us the Examiner failed to articulate
reasoning with a rational underpinning as to why an ordinarily skilled artisan
would have combined the teachings of Findeisen, Li, and Sutton (Final Act.
6). More specifically, with respect to Li, the Examiner has determined an
ordinarily skilled artisan would have combined the teaching of Li with
Findeisen “to have an invention where the read-only memory is arranged to
store a security verification code to verify the BIOS code stored in the flash
memory” (Final Act. 5 (citing Li ¶ 20)). Findeisen teaches storing the hash
Appeal 2020-002601
Application 15/098,471

7
values in read-only memory and storing secure information in secure ROM
103 (Findeisen ¶¶ 29, 37–38). Li teaches the read-only memory can be an
on-die read-only memory (Li ¶ 21). Li further teaches the read–only
memory 105 is arranged to store a security verification code and public key
to verify code stored in other memory (id.). The Examiner determines Li
thus teaches “an on-die Read-Only-Memory to store trusted digest
information” (Ans. 14). Appellant has not persuaded us an ordinarily skilled
artisan would find locating the Read-Only-Memory to be an on-die ROM,
for the ROM taught by Findeisen would have been uniquely challenging or
otherwise beyond the skill of an ordinarily skilled artisan (see KSR Int’l Co.
v. Teleflex Inc., 550 U.S. 398, 417 (2007)). Indeed, we determine Findeisen
teaches storage of trusted information in both read-only memory 120 and
read-only memory 103 (Findeisen ¶¶ 27, 30, 34, 36–38). Thus, both
Findeisen and Li disclose use of a ROM for storing trusted information; Li
specifies the location of the ROM for storing trusted root digest information.
Appellant next contends, “in applying the teachings from Sutton, the
Examiner has impermissibly broadened the teachings of Sutton” (Appeal Br.
10). In particular, Appellant argues the claimed terms “root data” and
“ucode” “don’t appear anywhere in Sutton” thereby distorting “the
application of the cited art by combining the strained teachings of Sutton
with the Findeisen/Li combination” (id. at 10, 13). Appellant, however,
appears to be improperly suggesting identity of terminology is required to
reject a claim (see In re Schaumann, 572 F.2d 312, 317 (CCPA 1978); see
also In re Bond, 910 F.2d 831, 832 (Fed. Cir. 1990)). Appellant sets forth
insufficient argument or evidence to support the assertion that the teachings
of Sutton have been “impermissibly broadened” (see generally Appeal Br.).
Appeal 2020-002601
Application 15/098,471

8
Nor does Appellant set forth sufficient argument or evidence as to why
Sutton fails to teach the recited “wherein the mainboard data comprises a
ucode patch of the CPU, and the ucode patch is signed correspond to the
trusted root data,” as recited in claim 1.
Appellant further argues the Examiner failed to set forth “the rationale
for the motivation of its selective features with the Findeisen/Li
combination” (id. at 13). The Examiner has set forth an ordinarily skilled
artisan would have found it obvious to combine the teachings of Sutton with
Findeisen and Li “to have a system to validate and install microcode
patches” (Final Act. 6 (citing Sutton ¶ 15)). The Examiner further cites to
Sutton itself as support (Ans. 14). Sutton discloses that “microcode is
typically placed in a read-only memory (ROM) within the processor at the
time the processor is manufactured. However, microcode sometimes needs
to be modified after the processor is manufactured, and even after the
processor has been placed into operation” and thus, microcode patches are
installed using keys embedded in the processor (Sutton ¶¶ 1, 15, 20). The
Examiner then notes Sutton describes “‘[e]mbedded’ keys are manufactured
into the processor 110 in a manner that prevents them from being changed by
system[ ] software and that prevents them from being read by non-secure
software” (id. (citing Sutton ¶ 20) (emphasis omitted, emphasis added)).
As previously noted, the Examiner has set forth an ordinarily skilled
artisan would have found it obvious to combine the teachings of Sutton with
Findeisen and Li “to have a system to validate and install microcode
patches” (Final Act. 6 (citing Sutton ¶ 15)). The Examiner also relies on
Sutton’s description that embedded keys are manufactured into a processor
Appeal 2020-002601
Application 15/098,471

9
to prevent them from being changed by system software (Ans. 14 (citing
Sutton ¶ 20)).
As stated by the Supreme Court, “[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” (see KSR Int'l
Co. v. Teleflex Inc., 550 U.S. 398, 418 (2007)). Here, the Examiner has
articulated reasoning with a rational underpinning as to why Sutton’s
technique could have been combined with the teachings of Findeisen and Li.
In the Reply, Appellant argues Sutton fails to teach the concept of
trusted root data arguing the “hash value of a public key (disclosed in
Sutton) cannot properly correspond to the claimed trusted root digest
information” (Reply 7). The Examiner, however, does not rely on Sutton to
teach the claimed trusted root digest information (see Final Act. 3–6). As
noted supra, Appellant has not proffered sufficient evidence or argument to
persuade us the Examiner’s interpretation of “trusted root data” or “trusted
root digest information” are in error.
We additionally are not persuaded by Appellant’s allegation of
“improper hindsight reconstruction” (see Appeal Br. 13). “Common sense
teaches [] that familiar items may have obvious uses beyond their primary
purposes, and in many cases a person of ordinary skill will be able to fit the
teachings of multiple patents together like pieces of a puzzle” (KSR 550 U.S.
at 420).
Appellant does not proffer sufficient evidence or argument to
persuade us of error in the Examiners’ conclusion of obviousness, and in
particular, why the Examiner’s articulated reasoning is in error.
Appeal 2020-002601
Application 15/098,471

10
After considering Appellant’s assertions, we determine Appellant’s
arguments do not take into account what the collective teachings of the prior
art would have suggested to one of ordinary skill in the art and are therefore
ineffective to rebut the Examiner's articulated reasoning for combining the
teachings of Findeisen, Li, and Sutton (see In re Keller, 642 F.2d 413, 425
(CCPA 1981)(“The test for obviousness . . . is what the combined teachings
of the references would have suggested to those of ordinary skill in the art.”)
(citations omitted)). Accordingly, Appellant has not persuaded us the
Examiner improperly combined the teachings of Sutton, Li, and Findeisen.
Accordingly, we affirm the Examiner’s obviousness rejection of
independent claim 1, independent claim 9 that recites similar limitations, and
the claims dependent thereon, not separately argued (see Appeal Br. 9).

DECISION SUMMARY
Claim(s)
Rejected
35 U.S.C.
§
Reference(s)/Basis Affirmed Reversed
1, 3–7, 9,
11–15
103 Findeisen, Li, Sutton 1, 3–7, 9, 11–
15

2, 10 103 Findeisen, Li, Sutton,
Miller
2, 10
Overall
Outcome
1–7, 9–15

TIME PERIOD FOR RESPONSE
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a). See 37 C.F.R.
§ 1.136 (a)(1)(iv).
AFFIRMED

Appeal 2020-002601
Application 15/098,471

11
PYONIN, Administrative Patent Judge, dissenting.
I respectfully disagree with the majority opinion in affirming the
rejection of the pending claims. Based on the record before us, I am
persuaded the Examiner’s rejection contains reversible error. See Reply Br.
5, 6.
Claim 1 recites “trusted root digest information” that is, inter alia,
both (1) not allowed to be modified, and (2) compared to digest information
computed from data stored in a mainboard. The Examiner relies on
Findeisen for these claim features. Appeal Br. 13, 14. Specifically, the
Examiner finds two different elements of Findeisen meet the trusted root
digest information limitation: (1) Findeisen’s “‘trusted set of data 103A,
which is substantially resistant against any external manipulation’ can be
interpreted as ‘wherein the trusted root digest information is not allowed to
be modified,’” or, alternatively, (2) Findeisen’s “‘boot information 121 . . .
also can be interpreted as . . . ‘trusted root digest information’” that is
“compare[d] [to] digest information.” Ans. 11–12 (emphases omitted);
Findeisen ¶¶ 30, 36.
I agree with Appellant that the Examiner has improperly “dissected
the claimed feature” of the trusted root digest information. Reply Br. 7.
Neither of the Examiner’s citations to Findeisen (i.e., the (1) trusted set of
data 103A or the (2) boot information 121) teach or suggest the trusted root
digest information, in the context of claim 1. See Reply Br. 5, 6. First,
Findeisen’s trusted set of data 103A comprises “instructions and data values
representing a secure load routine . . . [and] one or more decryption keys”
(Findeisen ¶ 32), which is used to “enabl[e] a decryption of the encrypted
hash values” (Findeisen ¶ 38). Findeisen provides no mention of comparing
Appeal 2020-002601
Application 15/098,471

12
data 103A to other data, as required by claim 1. See Appeal Br. 5; Findeisen
¶¶ 42, 43.
Second, I find that one of ordinary skill would not understand
Findeisen’s boot information 121 “is not allowed to be modified,” as
required by claim 1. Findeisen implicitly suggests the boot information 121
can be changed, as the reference goes to great lengths to describe trusted set
of data 103A is protected, but makes no similar disclosures with respect to
boot information 121. See, e.g., Findeisen ¶ 27.
Moreover, Findeisen’s boot information 121 is BIOS information
stored in “non-volatile memory 120, such as a flash memory and the like.”
Findeisen ¶ 33; see also Findeisen Fig. 1c (depicting the non-volatile
memory as Flash ROM). Such memory is an arrangement for storing
computer BIOS that provides flexibility by allowing for updates (i.e.,
modifications).2 See Appeal Br. 5. Li teaches “BIOS code is stored in flash
memory of the electronic system,” which allows for “tampering with the
BIOS code.” Li ¶ 5. As explained similarly in Sutton, “an encoded patch
may be stored in non-volatile memory such as the BIOS,” and “key(s) [can
be] . . . placed into BIOS with a BIOS upgrade.” Sutton ¶¶ 23, 27. Thus, I
agree with Appellant that Findeisen’s boot information 121 does not teach or

2 “Flash memory is an electronic non-volatile computer memory storage
medium that can be electrically erased and reprogrammed.” Wikipedia,
https://en.wikipedia.org/wiki/Flash_memory (retrieved March 31, 2021); see
also PC Magazine, https://www.pcmag.com/encyclopedia/term/flash-rom
(“Another term for flash memory. Since flash chips can be updated over and
over, the ROM (read only) designation is somewhat misleading.”) (retrieved
March 31, 2021).
Appeal 2020-002601
Application 15/098,471

13
suggest “trusted root digest information [that] is not allowed to be modified”
as claimed. See Appeal Br. 8, 9; Reply Br. 6.
I am persuaded Findeisen does not teach an element that satisfies all
requirements of the claimed trusted root digest information. See Appeal Br.
9. The Examiner does not rely on Li or Sutton for these teachings; nor does
the Examiner provide a reason to modify Findeisen to obtain “trusted root
digest information” having all the required features of the claim. See Ans.
10–14.
Accordingly, I am persuaded the Examiner errs in finding all
limitations of claim 1 are obvious in view of the cited art. See Reply Br. 5,
6. Respectfully, I would not sustain the Examiner’s obviousness rejection of
the claims.