Ex Parte 6546446 et al

Patent Trial and Appeal Board

Sep 26, 2012

95001109 (P.T.A.B. Sep. 26, 2012)

UNITED STATES PATENT AND TRADEMARKOFFICE
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.
95/001,109 11/11/2008 6546446 38512.20 9305
22852 7590 09/27/2012
FINNEGAN, HENDERSON, FARABOW, GARRETT & DUNNER
LLP
901 NEW YORK AVENUE, NW
WASHINGTON, DC 20001-4413
EXAMINER
ESCALANTE, OVIDIO
ART UNIT PAPER NUMBER
3992
MAIL DATE DELIVERY MODE
09/27/2012 PAPER
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.
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

MICRON TECHNOLOGY, INC.
Requester, Appellant

v.

RAMBUS INC.
Patent Owner, Respondent
____________

Appeal 2012-001639
Inter Partes Reexamination Control No. 95/001,109 & 95/001,155
United States Patent 6,426,916 B2
Technology Center 3900
____________

Before HOWARD B LANKENSHIP, KARL D. EASTHOM, and
STEPHEN C. SIU, Administrative Patent Judges.

EASTHOM, Administrative Patent Judge.

DECISION ON APPEAL
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

2

This merged proceeding arose out of separate requests by Micron
(95/001,155) and Samsung Electronics Ltd. (95/001,109) for inter partes
reexaminations of U.S. patent 6,546,446 B2 to Farmwald et al., Synchronous
Memory Device Having Automatic Precharge (issued April 8, 2003, which
claims priority to April 18, 1990 based on a series of continuation
applications starting with application number 07/510,898 assigned to
Rambus.
Samsung has not filed a Brief in this proceeding. Appellant,
Requestor Micron, appeals from the Examiner’s decision in the Right of
Appeal Notice (RAN) confirming claims 1-4. The Examiner’s Answer
relies on the RAN, incorporating it by reference.
We have jurisdiction under 35 U.S.C. §§ 134(b) and 315.
We REVERSE the Examiner’s decision to confirm claims 1-4 based
on certain rejections and AFFIRM the Examiner’s decision not to reject the
claims based on other rejections.

STATEMENT OF THE CASE
Rambus and Micron refer to several related judicial and other
proceedings including inter partes and ex parte reexaminations,
International Trade Commission proceedings, and Federal District Court and
Circuit Court proceedings in their briefs and inter partes requests. An oral
hearing of this appeal transpired on June 6, 2012 and was subsequently
transcribed.
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

3
Appellant, Requester Micron, appeals the Examiner’s refusal to
maintain the rejections of claims 1 and 2 as obvious based on Bennett,1 and
either Wicklund2 or Bowater3; claim 3 as obvious based on Bennett, either
Wicklund or Bowater, and either Schanke,4 Inagaki,5 or Novak;6 claim 4 as
obvious based on Bennett, either Wicklund or Bowater, and either Bazes7
or Lofgren;8 claims 1-2 as obvious based on Moussouris,9 Gustavson
(“SCI-A”) ,10 and either Wicklund or Bowater; claims 3-4 as obvious based
on Moussouris, Gustavson, either Wicklund or Bowater, and Schanke
(“SCI-B”). Requester also relies on admitted prior art (“APA”) for most of
the rejections and appeals the refusal to maintain the rejections of claims 1-4
as anticipated by JEDEC11 or as obvious based on Park12 and JEDEC. The
JEDEC and Park rejections turn on priority.
The appealed claims follow:

1 Bennett et al., U.S. 4,734,009 (Mar. 29, 1988).
2 Wicklund et al., U.S. 5,159,676 (Oct. 27, 1992).
3 Bowater et al., U.S. 5,301,278 (Apr. 5, 1994).
4 Schanke, Proposal for Clock Distribution in SCI (May 5, 1989).
5 Inagaki, JP 57-210495 (Dec. 24, 1982) (reference hereinafter is to an
English language translation of record) (attached as Exhibit 2 to NVIDIA’s
Respondent Brief).
6 Novak et al., U.S. 4,663,735 (May 5, 1987).
7 Bazes, U.S. 4,496,861 (Jan. 29, 1985).
8 Lofgren et al., G.B. 2,197,553 A (May 18, 1988).
9 Moussouris, Life Beyond RISC: The Next 30 Years in High-Performance
Computing, Technologic Comp. Let. V. 5, No. 5 (July 31, 1989).
10 Gustavson et al., The Scalable Coherent Interface Project
(Aug. 22, 1988).
11 Joint Electronic Device Engineering Counsel (JEDEC) Standard
No. 21-C, Revision 9 (1999).
12 Park et al., US 5,590,086 (Dec. 31, 1996, effective filing Oct. 1993).
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

4
1. A synchronous integrated circuit device having a memory array which
includes dynamic random access memory cells, wherein the integrated
circuit device comprises:
a clock receiver to receive an external clock signal;
a plurality of sense amplifiers, coupled to the memory array, to sense data
from the dynamic random access memory cells; and
a plurality of input receivers to sample an operation code synchronously
with respect to the external clock signal,
the operation code including precharge information, wherein, in response to
the precharge information, the plurality of sense amplifiers is automatically
precharged after the data is sensed.
2. The integrated circuit device of claim 1 wherein the operation code
specifies a read operation, and wherein the integrated circuit device further
includes a plurality of output drivers to output first and second portions of
the data in response to the operation code specifying a read operation.
3. The integrated circuit device of claim 2 wherein: the plurality of output
drivers output the first portion of the data synchronously with respect to a
rising edge transition of the external clock signal; and the plurality of output
drivers output a second portion of the data synchronously with respect to a
falling edge transition of the external clock signal.
4. The integrated circuit device of claim 2 further including a delay lock
loop, coupled to the plurality of output drivers, to synchronize the output of
the first and second portions of the data with the external clock signal.
ANALYSIS
I. Obviousness – Bennett with Wicklund or Bowater, and Prior Art
Admissions (APA) in the ‘446 Patent
The Examiner determined that claims 1 and 2 are not obvious over
Bennett and either Wicklund or Bowater. One central dispute on appeal
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

5
involves the recitation in claim 1 of a “synchronous integrated circuit device
having . . . dynamic random access memory cells;” i.e., a DRAM chip, and
the ability of the device to synchronously sample an operation code which
includes precharge information. Claim 2 further includes a read request in
the operation code.
Bennett’s Teachings
B1. Bennett’s “paramount object” is to provide communication
between “very large scale integrated VLSI (circuit) elements” (col. 12,
ll. 14-18) – i.e., “VLSIC chips” (col. 9, ll.35-40). Bennett discloses
combining Versatile Bus Interfaces (VBI) and VLSIC “upon the same chip
substrate as the VLSI User Device” (col. 12, ll. 29-32 (emphasis added))
with such a user device including “interfaces intended to be built with a
CPU, IOC or Memory, or similar User device for signal or data exchange”
(col. 35, ll. 59-61 (emphasis added)). (See also col. 14, ll. 19-24 (describing
“interface to the user devices (usually upon the same chip substrate)”.)
As another example demonstrating a preference for a single chip,
Bennett states that “[e]ach Versatile Bus Interface Logics, for example
Versatile Bus Interface Logics 102a [of Fig. 1], interfaces a User module, for
example VLSI Circuit User Device 106a which is pictorially represented in
shadow line within FIG. 1 as existing on the same VLSIC chip substrate as
Versatile Bus Interface Logics 102a.” (Col. 36, ll. 19-24 (emphasis added).)
Bennett’s chips have up to 120 pins as a practical limit. (Col. 9,
ll. 60-61.) Bennett also discloses different memory types as “Fast Memory”
or “Large Memory” with the memory having address widths of 16, 24, or
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

6
32, and one fast memory embodiment having 37 pins (col. 92, ll. 15-56;
Fig. 32). One large memory has at least 16 pins to access 232 addresses by
employing two 16-bit address words over successive clock cycles. (See col.
95, ll. 59-60; Fig. 36.)
B2. Figure 38 shows “memories device” 3802c and 3802d connected
to a “Versatile Bus.” (Col. 97, ll. 8-10.) In the next paragraph, Bennett
refers to “VSLI chips hav[ing] access to all Versatile Bus lines and
therefore, the Versatile Bus protocols.” (Id. at ll. 20-22.)
Bennett elsewhere refers to “memory devices” including, but not
limited to, a ROM: “Not all memory devices can perform all operations; for
example, read only memory (ROM) cannot execute the write operations.”
(Col. 90, l. 66 to col. 91, l. 2.) Bennett then refers to “[s]ample memory
operations in the following paragraphs” (col. 90, ll. 4-5) and thereafter
describes “relatively small fast memories, and . . . larger and relatively
slower memories” (col. 92, ll. 13-14).
Bennett also refers to “VLSIC chip devices” and in the next full
sentence (under the “Section 4.1” heading) states that “[m]any, if not most,
applications of VLSIC technology are likely to include memory devices.”
(Col. 90, ll. 42-43.) Bennett generally discusses these chip devices as
employing the interconnection protocol standards outlined generally in
Section 3 and more specifically discusses memory devices in Section 4 of
the ‘051 patent. (See id. at ll. 36-41.)
For example, as discussed in Section 4 of Bennett, Figures 32 and 33
represent fast memory write operations using data on 16 pins and 16 other
pins for arbitration and slave ID. (See col. 93, l.12 - col. 94, l. 56.)
Figure 36 represents pin and timing for a write operation to a large memory
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

7
device with a 4315335 protocol “configuration.” (See col. 26, ll. 54-57.)
Figures 25a-h, represent more generic slave device configurations as
discussed in Section 3 of Bennett. (Col. 25, l. 58 to col. 26, l; see generally
columns 81-88).
B3. In addition to chips, Bennett also discusses memory cards in
Section 2, “Description of the Prior Art” (see col. 5, l. 52 et seq.), and states
that “the functionality of VLSIC chips is often similar to cards today” but
that “VLSIC technology promises much higher performance than that of
cards,” even though cards hold more memory and chips have higher
development costs. (Col. 9, ll. 43-56.) In the following passage, Bennett
discusses creating larger chips to accommodate a greater numbers of pins.
(Col. 9, l. 66 to col. 10, l. 29.)
B4. Bennett describes a “third physical objective” – the VBI
(versatile bus interface) “should occupy a reasonable VLSI circuit substrate
area” using fast and efficient CMOS technology as the preferred
embodiment. (Col. 13, ll.18-23.) Typically, only about 20 VLSIC devices
will be interconnected. As a “first logical object,” the VBI logics “should
offer a fixed format, simply controlled, powerfully featured interface to the
user devices (usually upon the same chip substrate)” yet with certain options
for use. (Col. 14, ll. 20-30.)
Bennett contemplates simple devices with “as few as three pins”
(Bennett, col. 12, l. 61), or “pass[ing] but a single bit of data from a single
master device to a single slave device . . . [or more bits and devices]. The
versatility is from the trivial to the profound.” (Col. 15, ll. 26, 42-50.)
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

8
B5. Figures 25a-h show an ID/FUNCTION command, which
includes a read or a write (see e.g. Fig. 35, 36).13 (See col. 85, l. 9 to col. 87,
l. 6.). Figure 34 shows multiple functions in a memory write code: i.e., a
read-modify-write code signifying multiple functions in a single code.
B6. Bennett discloses synchronous clocked communication between
bused VLSIC chips over 16 data lines at 25MHz, and notes that synchronous
communication is more efficient than asynchronous communication.
(Col. 13, ll. 3-17; col. 66, l. 9 – col 67, l.18; col. 101, ll. 50-54 (“all
communication . . . is synchronously referenced”); col. 102, ll. 9-27.)
Bennett also states that the clock signals “are normally synchronous.”
(Col. 274, l. 62.)
B7. Bennett employs a dual-phase clocking scheme. (See Fig. 84.)
Bennett explains that the scheme employs a first phase to charge the
capacitance on the bus. (Abstract.) To further explain the system, Bennett
(col. 13, ll. 33-43; col. 105, ll. 50-61) points to, and incorporates by
reference, App. No. 0355,803, which corresponds to Bennett at al.,
U.S. 4,500,988 (Feb. 19, 1985.) Bennett ‘988, in reciprocal fashion,
mentions using its clocking scheme and corresponding driver circuitry in
Bennett. (See, e.g., Bennett ‘988, col. 7, ll. 47-51 (referring to Bennett’s
application number 356,051).)
B8. Bennett ‘988 (col. 7, ll. 62-67) and Bennett (col. 277, ll. 30-33)
each similarly explain that the dual-phases can have a 50% duty cycle and
can be symmetrical. Bennett ‘988 explains that the dual phase scheme

13 Bennett also refers to functions as operations, which include read or write
operations. (See col. 91, l. 62 to col. 92, l. 8; col. 91, ll. 43-53 (“functional
operations”); Figs. 31, 35, 36.)
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

9
allows smaller interfacing transistors in separate devices to work together in
a wired-OR fashion to pre-charge the bus. (Bennett ‘988, col. 6, ll. 47-51;
col. 7, ll. 7-34; col. 8, ll. 57-63.) The bus is precharged to a high voltage
level of 3 volts (corresponding to a logical 0) at the leading edge of phase
(H) φ1, and then, if, and only if, any one of the data lines seek to represent a
logical 1 data bit, that line/lines is/or driven to a low voltage at the leading
edge of phase (H) φ2. (See Bennett’ 988, Fig. 5, col. 6, ll. 55-65; Bennett
Fig. 84.) In other words, some data lines stay logical 0 as initiated by the
leading clock edge of (H) φ1, and others change to logical 1 as initiated by
the leading clock edge of (H) φ2.
B9. Bennett ‘988 discloses that the dual clocks are both “distributed
by a master clock.” (Bennett ‘988, col. 7, ll. 61-62.)
B10. Bennett explains that faster operations are possible by
“configured nonperformance of arbitration and slave identification/function
or both.” (Col. 107, ll. 61-63.) Bennett compares “[u]nsophisticated User
devices such as memories” to “sophisticated devices such as processors” –
the latter do not know what other devices are on the network. (Col. 58,
ll. 64-65.) As noted supra, Bennett contemplates simple systems having “a
single slave memory.” (Col. 57, l. 57.) Bennett explains that “the number of
[device] locations strongly affects complexity.” (Col. 8, ll. 30-31.) Bennett
distinguishes between slaves and masters: slaves “only respond to
information on the interconnect,” masters “control the interconnect”; thus,
slaves are subordinate to masters. (Col. 8, ll. 30-41.)
B11. Bennett projects speed increases – e.g., systems projected to
“drive signals from chip to chip in 20 to 40 nanoseconds.” (Col. 9, ll. 58-
59.)
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

10
Discussion
Rambus maintains that the Board should confirm the claims for the
same reasons that the Examiner does. (Resp. Br. 2.)14 Micron shows (App.
Br. 12-14; Reb. Br. 2-4) that the Examiner’s articulated reason; i.e., that the
prior art teaches a precharge signal, but fails to suggest modifying Bennett’s
operation code to include a precharge signal, is in error. (See RAN 70-76.)
The ‘446 patent admits, and other portions of the record corroborate
(see e.g., App. Br. 8, n.5 and discussion below), that in a normal read mode,
a precharge signal normally follows a DRAM read signal. For example, the
‘446 patent describes known precharge operations as follows:
In normal mode (in conventional DRAMS and in this
invention), the DRAM column sense amps or latches have been
precharged to a value intermediate between logical 0 and 1.
This precharging allows access to a row in the RAM to begin as
soon as the access request for either inputs (writes) or outputs
(reads) is received and allows the column sense amps to sense
data quickly.
(‘446 patent, col. 10, ll. 18-24.)
The patent further describes “typical;” i.e., known settings: “Typical
settings are ‘precharge after normal access’ and ‘save after page mode

14 Rambus’s contention that Micron’s incorporation by reference to earlier
reexamination documents is improper lacks merit. (See Resp. Br. 1-2.)
Rambus similarly incorporates other documents, the record is large, and the
briefs are limited by page size. (See Resp. Br. 1 n. 4; Resp. Br. 8 (relying on
the Examiner’s findings in related proceedings but only citing Micron’s
Brief); Reb. Br. 1.) As Micron points out, Micron need not raise arguments
or present detailed findings listing claim elements, etc. or other issues and
findings with which Micron and the Examiner agree. (See Reb. Br. 1.)
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

11
access’ but ‘precharge after page mode access’ or save after normal access’
are allowed.” (’446 patent, col. 10, ll. 43-46.)15
As to the operation code, Micron relies on Bennett’s “synchronous
operation (FUNCTION) code of Figs. 31, 34),” and notes that the Examiner
initially adopted the proposed rejection and did not alter the operation code
finding. (See App. Br. 7 (citing the Examiner’s “8/7/09 OA at 98-99);
accord B5, B7).)16 Hence, while the Examiner and Micron agree that
Bennett discloses an operation code, the Examiner reasons that modifying it
to include a precharge signal would not have been obvious:
While the primary reference [Bennett] disclose[s] . . . an
operation code that is sent synchronously with respect to an
external clock signal, the rejection to the claims does not
support a reason to include the precharge information with the
same operation code.
(RAN 69.)

Contrary to the Examiner’s quoted rationale, Micron does
provide persuasive reasons, for example, based on the above-quoted
admissions in the ‘446 patent – i.e., that precharging “‘allows the
column sense amps to sense data quickly.’” (App. Br. 8 (quoting the
‘446 patent).) Micron also points out that the “‘446 patent admits that

15 The Board addressed a similar issue in a related case: BPAI 2012-000142
(see original decision and rehearing decision). The findings and rationale
there related to Bennett, Bowater and Wicklund are adopted and
incorporated by reference.
16 The merged Office Action (OA) appears in the 95/001109 reexamination
file and refers to Micron’s supplemental inter partes reexamination request
(March 11, 2009) which appears in the 95/001155 reexamination file.
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

12
both page mode and normal mode were known separately in the …
art.” (App. Br. 8.)
As Micron further reasons, “instead of leaving the row (page)
of data in the sense amplifiers for a subsequent access as is done in
page mode, a normal mode operation is followed by a precharge.”
(App. Br. 8.) Rambus similarly describes the known operations,
noting that “[p]recharging occurs when a row in the DRAM is closed”
and that “[i]n asynchronous DRAMS, precharging is performed in
response to a further instruction—when the /RAS signal is
[deasserted] . . . at the end of a sequence of one or more accesses to
data in a particular row.” (Resp. Br. 4; accord id. n.7 (“Once all
accesses in a row are complete, the row is closed and precharging in
preparation for the next access occurs.”).)
As Micron also points out, Mr. Murphy, Rambus’s expert in
this and other litigation, testified that “‘[t]he normal mode read is
essentially equal to, or similar to a read with auto precharge.’” (App.
Br. 8 (emphasis by Micron, quoting Hynix Trial Tr. at 434:2-11).)
Hence, since it was well-known that a signal to precharge follows the
normal mode read in prior art systems, it would have been obvious to
include it with a normal mode read signal, as a modification to
Bennett’s (FUNCTION) operation code, in order to eliminate sending
another or redundant signal.
Moreover, Bennett further suggests the modification by
disclosing more than one function in a single operation code. (See B5
(read-modify-write function).) Since there is no reasonable dispute
that a DRAM row in the prior art typically closes pursuant to a
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

13
precharge signal after the normal mode read signal, and Bennett
discloses multiple signals in a single op code, closing the row
automatically after performing the read in response to a single
operation code would have been obvious so that the row would be
ready for another normal mode read.
Micron also relies on teachings in Wicklund and Bowater. (See App.
Br. 9-11.) The teachings there are largely duplicative to what is known in
the art. For example, Micron points out that “Wicklund describes the same
goal identified in the ‘446 patent.” (App. Br. 9.) Rambus’s assertions, and
the Examiner’s findings, directed to Wicklund’s and Bowater’s teachings,
fail to account adequately for the rationale advanced by Micron based on the
totality of the teachings including the ‘446 patent admissions.
Micron buttresses its rationale by pointing out that Wicklund provides
further reasons for closing a row and automatically precharging – to avoid
the penalty of a page miss by accurately predicting page or normal mode
accesses. (See App. Br. 9-11.) Micron reasons that Wicklund’s predictive
system renders obvious automatically closing a DRAM row (i.e.,
precharging) after a previous read access. (See App. Br. 9-11.) Micron’s
rationale, buttressed by the ‘446 patent admissions, is persuasive and
supports the obviousness of sending a signal as part of Bennett’s
synchronous op code. The modified system would have been more efficient
by accurately predicting that the data would be on a different row on a
subsequent access, thereby suggesting a row close (i.e., a precharge) on the
current row prior to accessing the different row. (See App. Br. 11-12.)
Rambus also maintains that Bennett does not disclose a synchronous
memory device. (Resp. Br. 2-3.) Rambus similarly argues that Bennett's
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

14
operation codes “are provided to the VBI, not the DRAMs that Micron
alleges.” (Id. at 4.) These related arguments reduce to the assertion that
Bennett does not disclose a single chip memory device, because Bennett
discloses memory boards, and Bennett does not disclose a DRAM. Bennett
at least discloses a synchronous single chip memory device (B1-B6) and
renders a synchronous DRAM chip obvious, where, for example, Wicklund
describes DRAMs as “the most popular form of read/write memory” (col. 1,
ll. 37-38) and Bennett discloses read/write memory devices, while noting
that ROM chips do not perform writes (B4). Finally, the Board, and a
District Court, has separately addressed similar arguments in related
proceedings, and the findings and rationale there are adopted and
incorporated herein by reference.17
Moreover, Rambus’s argument that Bennett’s memory device does
not signify, to skilled artisans, a single memory chip, but rather, only

17 For example, see BPAI 2011-000142) (2012) (finding that “Bennett
discloses synchronous memory chips” and concluding that employing a
notoriously well-known DRAM memory chip in place of Bennett’s generic
memory chip would have been obvious); BPAI 2011-009664 at 14-18
(2012) (finding that Bennett discloses a synchronous memory device chip
concluding obviousness with respect to a similar issue involving claims
directed to a read operation based, inter alia, on Bennett, Wicklund or
Bowater); BPAI 2012-001638 at 3-22 (2012) (relying, in part, on a District
Court December 15, 2008 Order Granting in Part and Denying in Part
Rambus’s Motion to Strike; Denying Motion for Summary Judgment No. 1
of Invalidity; and Striking Motion for Summary Judgment No. 2 of
Invalidity, Hynix Semiconductor Inc. v. Rambus Inc., No. 00-2905, Rambus
Inc. v. Hynix Semiconductor Inc., No. 05-334, Rambus Inc. v. Samsung
Electronics Co., Ltd., No 05-2298 RMW, and Rambus Inc. V. Micron
Technology, Inc., No. 06-244 (N.D. Cal.) (stayed, Judge R. Whyte) (attached
as Ex. O-3 to the Rambus’s respondent brief in the 2012-001638
reexamination).
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

15
signifies a memory board (see Resp. Br. 4), contradicts other arguments
made elsewhere before the PTO and related proceedings.18 Bennett’s
disclosure of the same term, “memory device” (B2), references to “VLSIC
upon the same chip substrate,” “interfaces intended to be built with . . .
Memory,” (B1) and other similar references to VLSIC, chips or “same”
substrates (B1-B4), combined with a limited discussion of memory cards as
prior art (B3), all show that Bennett’s memory device includes a single chip
embodiment (even if the term also signifies other memory forms of memory
as Rambus argues). Micron’s similar reliance on the memory chip of
Bennett’s Figure 1 and memory devices in Figure 38 bolsters the Examiner’s
findings. (See Reb. Br. 3; accord B1, B2.)
Rambus facially presents arguments with respect to claim 2, but the
arguments do not differentiate in a meaningful manner from arguments
regarding claim 1. (See Resp. Br. 7-8; accord Resp. Br. 1 (grouping claims
1 and 2 together).) For example, Rambus’s observation that “in addition to
including precharge information, the operation code [in claim 2] specifies a
read operation” facially requests the Board to decide that the combination
does not disclose or suggest a read operation. (See Resp. Br. 7.) But
Bennett’s system sends a read op code to memory devices as Micron points
out. (See App. Br. 7, 12 (relying on the Bennett’s FUNCTION code as
including a read and further relying on output drivers cited in Micron’s
Request); B5.)

18 See In re Rambus, App. No. 2011-1247 (Fed. Cir. Aug. 17, 2012)
(involving an appeal of BPAI 2010-0011178- the Board and the Federal
Circuit holding that the term “memory device” includes, but is not limited
to, a single chip contrary to Rambus’s arguments otherwise).
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

16
Based on this record, Micron persuasively demonstrates that the
Examiner erred in failing to maintain the rejection of claims 1 and 2 based
on Bennett, APA, and Wicklund or Bowater.

II. Obviousness – Bennett, APA, Wicklund, and Inagaki or Novak
Claim 3 depends from claim 2, and requires outputting data
synchronously with respect to rising and falling edges of a clock pulse. The
Examiner relies on reasons for confirming claims 1 and 2. (See RAN 70,
77.) Rambus presents additional reasons for supporting the Examiner’s
decision to confirm claim 3. (See Resp. Br. 8-9.) For example, Rambus
points to the Examiner’s reasoning about Bennett’s clock system in a related
reexamination proceeding in which the (same) Examiner confirms a claim
which is similar to claim 3 here. (See Resp. Br. 8.) Micron’s responses,
which rely on the secondary teachings of Inagaki or Novak, show the
obviousness of employing rising and falling edges to input or output data, as
discussed further below. (See App. Br. 12-15.)
Inagaki’s Teachings
I1. Inagaki discloses a method for increasing data rates in block
access memory. As background, Inagaki teaches that conventional methods
to increase data transfer rates in RAMs involved increasing the data bus
width, which adds cost of packaging and pin count, or to increase the clock
rate. (Inagaki 2.) Inagaki’s solution is to use dual edges of a clock as
quoted as follows.
I2. “The rise and fall of external clock φ are detected, and clocks φ1
and φ2 are generated. Clocks φ1 and φ2 drive shift pulses of the shift
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

17
register. . . . In this way, since one bit is output on each half-cycle, the
operating speed is twice that of the conventional speed.” (Id. at 4.)
I3 “[T]he present invention presents block access memory that
transfers data with a speed that is twice the conventional speed, by
performing I/O [input/output] of data on every half-cycle of the external
clock that drives the I/O shift register.” (Id. at 3.)
Discussion
Rambus maintains that using Inagaki’s clocking scheme in Bennett
would not have been obvious for various reasons. Rambus initially maintains
that Inagaki uses pulses and not a periodic clock. (Resp. Br. 9.) Rambus
makes a similar contention about Novak, asserting that an asynchronous
device does not have an external clock. (See id.)
But Inagaki refers to an “external clock” repeatedly and the clock is
periodic at least while it operates: “The rise and fall of external clock φ are
detected, and clocks φ1 and φ2 are generated. . . . . In this way, since one bit
is output on each half-cycle, the operating speed is twice that of the
conventional speed.” (Inagaki at 4 (emphasis added.) Inagaki’s system
performs “data input or output every half-cycle based on an external clock.”
(Id. at 2.) Inagaki’s numerous computer clock references point skilled
artisans to and embrace the well-known computer clock – in other words,
the same type of external computer clock generically claimed in the
‘446 patent.19 See In re Paulson, 30 F.3d 1475, 1480-81 (Fed. Cir. 1994)

19 clock . . . A source of accurately timed pulses, used for synchronization in
a digital computer . . . .” McGraw-Hill Dictionary of Scientific and
Technical Terms 387 (Fifth Ed. 1994). This reference indicates that clock
signals and clock pulses are the same: i.e., “clock signals. See Clock
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

18
(“a prior art reference must be ‘considered together with the knowledge of
one of ordinary skill in the pertinent art’,” and where the skill level was
“‘quite advanced’ . . . ‘one of ordinary skill certainly was capable of
providing the circuitry necessary to make the device operable for use as a
computer’”) (citations omitted).
Also, Bennett employs different periodic external clocks (see B6;
Bennett Fig. 84; Abstract) and Rambus’s arguments amount to an
unpersuasive separate attack on the references. Micron relies on the
combination to suggest using the rising and falling edges of a clock. Micron
adds that Inagaki teaches using dual edges of a single clock in order to
increase speed or reduce the number of data pins or paths in a memory
device. Micron relies on similar teachings in Novak. (See App. Br. 12; I1-
I3; Novak, col. 6, ll. 34-35; 4f, 4d (“for read operations it takes only
128 cycles of the clock φ to output 256 bits”.)20
Rambus also maintains that Bennett discloses an “intricate clocking
scheme” (Resp. Br. 9) which precludes the proposed modification of
employing leading and falling clocking edges because such a modification
would destroy Bennett’s principle of operation and would lack any
reasonable expectation of success. (Id. at 8-9.) The Federal Circuit recently

pulses.” Id. “[C]lock pulses. . . . Electronic pulses which are emitted
periodically, usually by a crystal device, to synchronize the operation of
circuits in a computer. Also known as clock signals.” Id.

20 Several previous Board decisions reasoned that Inagaki discloses a clock
and using dual edges therefore. See, e.g. BPAI 2012-000171 (finding
obviousness with respect to combining Inagaki with a prior art reference
using a dual clocking scheme). The findings and rationale involving Inagaki
are adopted and incorporated by reference.
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

19
rejected a similar argument under analogous circumstances based on the
observation that an asserted “difference does not affect the operability of
Mouttet’s [i.e., the applicant’s] broadly claimed device—a programmable
arithmetic processor.” In re Mouttet, 686 F.3d 1322,1332 (Fed. Cir. 2012)
(citations omitted) (also reasoning that physical incorporation is not required
to support obviousness).
The principle of operation of the broadly claimed memory device here
involves synchronously outputting data from a memory device on both clock
edges of the clock. Bennett’s principle involves “its high level ability to,”
see Mouttet at 1332, transfer data synchronously (B6) to and from a single
memory device or a group of such devices, using a “simply controlled,” yet
versatile, system (B4, accord B10 (“single slave device”)). For example,
Bennett’s system is “configured so simply as to pass but a single bit of data
from a single master device to a single slave device, or with ten deep
pipelining of eight phases of time-based arbitration (between 256 devices)
time overlapped with slave identification/function . . . . The versatility is
from the trivial to the profound.” (Col. 15, ll. 42-50 (emphasis added);
accord B4.)
Using one or two external clocks does not undermine this broad
principle of versatile data transfer and coalesces with the broad principle
underlying claim 3. Moreover, Bennett’s dual clocking scheme mainly
serves to precharge the bus lines. (B8.) Such a scheme is not required to
read data from a single device and is not required to satisfy claim 3. For
example, Bennett points out that the precharging scheme allows for reduced
interface transistor sizes which are used to drive the bus in a wired-OR
fashion – i.e., used to with multiple slave devices. (See B8.) As such,
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

20
skilled artisans would have recognized the pre-charging scheme would not
be required for simpler systems such as Bennett’s single slave memory chip
(see B11) since a large interface transistor and a single clock scheme could
be employed to drive a handful of chips - as Inagaki teaches and Bennett
suggests.
Even if Bennett suggests that precharging is required or advantageous
as implemented on the leading edge of phase (H) φ1, the second falling
transition of phase (H) φ1 could be employed to change other portions of the
data to a low voltage value as Inagaki suggests - instead of the leading edge
of phase (H) φ2 as Bennett discloses in the Figure 84 embodiment. (See B8.)
Bennett’s system further suggests, if not discloses, changing other portions
of data on the trailing (H) φ1 edge because the trailing (H) φ1 edge coincides
with the leading (H) φ2 edge in Bennett’s 50% duty cycle option. (See B7,
B8.)
Rambus’s arguments vaguely refer to corrupted data, but the
arguments strike down a straw man and address another hypothetical
modification in which both edges of H (φ1) are used to pull the data lines to
a high voltage. Such arguments do not explain how the simple
modifications described here would render Bennett inoperable. (See Resp.
Br. 8.) Rambus’s arguments appear to be based on the false premise that
Bennett does not contemplate symmetrical (50 % duty cycle) (H) φ1 and (H)
φ2 wave forms. Bennett’s 50 % duty cycle option essentially teaches or at
least suggests that the leading edge of one clock phase coincides with the
falling edge of the other – Inagaki’s internal clocks overlap in the same
fashion. (Compare B7, B8 and Bennett’s Figure 84 with Inagaki Fig. 4.)
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

21
In other words, Bennett’s principle of transferring data “is not unique
to its . . . [specific dual clocking] operation” as disclosed in the Figure 84
embodiment. See Mouttet at 1332. As also noted, Bennett’s scheme
envisions “trivial” schemes, including those embraced by the broad reach of
claim 3, which requires reading data synchronously from a single memory
chip using the dual edges of a clock as Inagaki or Novak suggest. (See B4,
B10.) Bennett also describes the clocking scheme as merely “a preferred
embodiment of the invention,” and indicates that “the electrical timing is
capable of being altered.” (See Bennett, col. 277, ll. 24-26.)21
Rambus has not demonstrated that skilled artisans, motivated by
Inagaki’s or Novak’s teaching of using rising and falling clock edges for
increasing data transfer speed from known DRAM memory devices, would
have been unable to modify Bennett’s system to arrive at the broadly
claimed invention. Inagaki’s or Novak’s dual clocking scheme provides the
fastest possible signal transfer on a bus without increasing the pin count or
data path width, thereby providing the motivation for the modification. (See
I1-I3; Novak, col. 6, ll. 34-35 Fig. 4f, 4d.)
Inagaki and Novak provide evidence of a reasonable expectation of
success in using dual clock edges on data for increased speed, especially for
“trivial” systems using a single or a handful of memory devices, as Bennett
teaches. Increased speed and compactness by reducing bus width and
corresponding pin number while saving cost (see I1) constitute universal

21 While the full passage implies altering the duty cycle of both clocks,
skilled artisans would have recognized that simple single memory device
systems as claimed would not require two clocks as Inagaki and Novak
make clear and as discussed supra.
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

22
motivators. Dystar Textilfarben GmBH & Co. Dutschland KG v. C.H.
Patrick Co., 464 F.3d 1356, 1368 (Fed. Cir. 2006) (“[A]n implicit
motivation to combine exists … when the ‘improvement’ is technology-
independent and the combination of references results in a product or
process that is more desirable, for example because it is stronger, cheaper,
cleaner, faster, lighter, smaller, more durable, or more efficient.”) Also, “if a
technique has been used to improve one device, and a person of ordinary
skill would recognize that it would improve similar devices in the same way,
using the technique is obvious unless its application is beyond his or her
skill.’” KSR Int’l Co. v. Teleflex, Inc., 550 U.S. 398, 417 (2007)(citation
omitted).
Despite Rambus’s related arguments that Micron has not shown how
to modify Bennett’s two phase system (see Resp. Br. 8-9), Rambus has not
shown that the proposed modification would have been beyond the skill of
an ordinary artisan, and the Court has also recognized that “the interaction of
multiple components means that changing one component often requires the
others to be modified as well.” KSR, 550 U.S. at 424. In other words, as
KSR implies, making other required modifications to increase the data speed
by using both clock edges, as Inagaki or Novak, does not defeat obviousness
or show inoperability.
Further, Mr. Murphy states that the clocking scheme was “often
referred to as ‘dual edge clocking’ and allows for data transfer at twice the
rate of the external clock signal.” (Murphy Dec ¶ 26) (filed May 13, 2009 in
the 95/001109 reexamination proceeding).) Mr. Murphy also explains that
“[o]ne of ordinary skill in the art would understand a synchronous bus, by its
very nature, relies on a clock, but the way that clock is distributed in the
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

23
system need not be limited to any particular distribution scheme.” (Id. at ¶
39.) Mr. Murphy’s testimony indicates that skilled artisans knew how to
implement what was often referred to as “dual edge clocking” to double the
data rate and knew how to apply different clocking schemes to clocked
systems like Bennett’s synchronous clocking scheme.
Given the claim breadth and high level of ability involved, skilled
artisans easily could have modified Bennett’s system in view of Inagaki’s or
Novak’s clocking scheme to create a cleaner system, dropping any unneeded
functions in a single device memory system (e.g., arbitration signals, etc.),
where Bennett’s system provides broad flexibility and versatility. Such a
“cleaner” memory device to handle single direction data transfers as
embraced by broad claim 3 constitutes a universal motivator under Dystar.
Also, Rambus’s reliance on Bennett’s Figure 84 to show that Bennett
requires a two phase intricate clocking system (see Resp. Br. 8) improperly
assumes that any structure implied by the example of Figure 84 must be
physically combinable with a known dual clocking structure – a proposition
Mouttet and other cases refute. See also In re Sneed, 710 F.2d 1544, 1550
(Fed. Cir. 1983) (“[I]t is not necessary that the inventions of the references
be physically combinable to render obvious the invention under review.”); In
re Nievelt, 482 F.2d 965, 968 (CCPA 1973) (“Combining the teachings of
references does not involve an ability to combine their specific structures”).
Assuming for the sake of argument that such a physical combination
is required to show obviousness, skilled artisans would have recognized that
Bennett’s Figure 84 system could have been modified as discussed above, or
under another alternative, to include the dual edges of slower external
Inagaki or Novak clocks as triggers for the leading edges of the faster clocks
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

24
φ1 and φ2 represented in Figure 84. That is, in Inagaki (see Fig. 4), a slower
external clock φ has rising and falling dual edges corresponding to and
triggering single leading edges of faster clocks φ1 and φ2 , suggesting a
similar external clock to trigger the fast clocks in Bennett. (See e.g. Inagaki
Fig. 4; I3.)22 Bennett employs multiple clocks, as does Inagaki, rendering
such a combination obvious. Still further, Bennett ‘988 discloses that
Bennett’s dual clocks are both “distributed by a master clock.”
(Bennett‘988, col. 7, ll. 60-61.) The ‘446 patent similarly discloses
“clocking at half the bus cycle data rate.” (See col. 19, ll. 33-45 (describing
the 500 MHz bus as using a 250MHz clock).)
Also, the thrust of Rambus’s arguments are not commensurate in
scope with claim 3, which does not require the “intricate clocking system”
(Resp. Br. 8) of Bennett. As discussed, claim 3 broadly embraces a memory
device which outputs one-way data synchronously. The record indicates the
structural features of claim 3, including sense amplifiers, input receivers, and
output drivers, constitute well-known components of typical DRAM
memory devices at the time of the invention. (See e.g., ‘446 patent, Fig. 15;
col. 23, ll. 43-65 (describing a “conventional 4 Mbit DRAM” and explaining
that “[m]any of these details have been implemented selectively in certain
fast memory devices”); accord Micron’s Request 35 (relying on such known
admitted DRAM features).) Rambus has not demonstrated that skilled
artisans, motivated by the known use of rising and falling clock edges to

22 Bennett also discloses varying the duty cycles of the two clocks as noted
above so that the clocks would coincide just as Inagaki’s clocks do.
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

25
increase data output speed from known DRAM memory devices on a single
bus, would not have been able to arrive at the broadly claimed invention.23
Based on the foregoing discussion, Micron has shown that the
Examiner erred in not maintaining the rejection of claim 3 based on Bennett,
APA, Wicklund, and Inagaki or Novak.

III. Obviousness Based on Bennett, Wicklund, and Lofgren or Bazes
With respect to claim 4, Rambus argues that Bazes and Lofgren do
not cure asserted deficiencies in the rejection of claim 1. (Resp. Br. 9-10).
The Examiner relies on reasons with respect to claim 1 for not rejecting the
claims – primarily the asserted lack of a teaching or suggestion for an
automatic precharge signal. (See RAN 71-72, 78-84.)

23 In a related hearing before the Board on September 12, 2012 (see hearing
transcript in BPAI 2012-002081 & 2012-001976 (argued together)), Rambus
raised a new argument premised on the Federal Circuit’s recent decision in
In re Rambus, App. No. 2011-1247 (Fed. Cir. Aug. 17, 2012) (cited supra).
Rambus argued that In re Rambus precludes Bennett’s interface because it is
akin to a complicated processor such as a BIU which the Federal Circuit
reasoned was not included in the term “memory device.” To the contrary,
Rambus itself distinguishes masters from slaves, as the Federal Circuit’s
reasoning points out. See id. at 6. In re Rambus only precludes “a global
bus controller or CPU,” id. at 7, but it does not preclude Bennett’s single
chip slave devices which merely respond to master bus processor controllers.
Bennett also teaches that any device functionality can be disabled to increase
speed and distinguishes such “unsophisticated” slave memory devices from
“sophisticated” processor master controller devices. (See B4, B10.) As
such, assuming arguendo Bennett’s chip interface has too much control, it
would have been obvious to eliminate unneeded functions in the chip
interface for the simple single memory chip at issue here. See In re Sovish,
769 F.2d 738, 743 (Fed. Cir. 1985) (“This argument presumes stupidity
rather than skill.”).
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

26
For reasons explained supra, that contention is not persuasive. With
respect to the added limitation in claim 4, Micron persuasively shows that
delay lock loops (DLLs) were standard in the art prior to the invention. For
example, Micron explains that Bazes teaches that for “‘integrated circuit
memories.’” including “‘dynamic, random-access memories, many clock
signals are required for each memory cycle to latch addresses, decode the
addresses, access the array, precharge nodes, control refreshing, etc.’” (App.
Br. 15 (quoting Bazes at col. 1, ll. 11-16).) Micron points out that Bazes
also teaches “‘generating these signals ‘on-chip’” using a “‘precision
synchronous delay line.’” (App. Br.15 (quoting and citing Bazes at col. 1,
ll. 17-19; col. 2, ll. 21-23.) Bazes also teaches that such a “synchronized
delay line” is “insensitive to voltage changes, temperature changes and
wafer processing variations. It is ideally suited for providing on-chip timing
signals derived from a reference clock for MOS integrated circuits.”
(Abstract.) As Micron also reasons, Bazes states that providing a DLL on-
chip obviates the need for high precision external clocks. (See App. Br. 15-
16; (citing Baze at col. 1, ll. 17-19).)
Rambus disputes Micron’s statement that “‘as demonstrated in
Micron’s reexamination request, it was known in the prior art to utilize a
DLL on a memory chip.’” (Resp. Br. 10 (citing and relying on “03/26/10
OA [office action] in [related] control no. 90/010,420 at 29-36”).) The
examiner in that ‘420 case finds, inter alia, that Lofgren does not disclose an
external clock signal for controlling the DLL and that Lofgren and the iAPX
Manual only suggests that “precise delays are for controlling data to the
memory” – i.e., controlling data from the iAPX MCU controller module to
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

27
DRAM chips and involving control of RAS and CAS control signals to
memory arrays. (See the ‘420 OA at 32 (emphasis added).)
But here, Bennett involves a single chip memory so that any
distinctions between control to and from a memory device do not flow
from the examiner’s logic in the ‘420 case. Thus, Rambus’s reliance on the
‘420 reexamination is misplaced. Here, more importantly, the Examiner
initially adopted the proposed rejections but subsequently refused to
maintain them based on asserted deficiencies in claim 1 with respect to
Wicklund’s precharge teachings. (See App. Br. 16 (relying on the
Examiner’s ‘155 “8/7/09 OA at 100.”) Further, the Examiner made a
supportive finding in another reexamination that culminated in an appeal to
the Board, whereas the Examiner’s findings in the ‘420 reexamination were
not at issue in an appeal of that reexamination.24 As indicated, the Examiner
recently found that combining Lofgren and the iAPX system would have
been obvious with Lofgren essentially teaching an accurate timer using a
delay line. The Board, relying on the Examiner’s findings, including the
finding that Rambus essentially equated phase-locked loops and delay-
locked loops, agreed with the examiner that combining Lofgren with the
iAPX Manual (also at issue in the ‘420 reexamination) would have been
obvious. (See BPAI 2011-008431 at 30-31.) The prior ‘8431 Board
decision and reasoning related to Lofgren is adopted and incorporated herein
by reference.

24 The ‘420 reexamination appeal culminated in In re Rambus Inc.,
2011-1247, (Fed. Cir. Aug. 18, 2012) (holding that the claimed memory
device is not limited to a single chip so that the iAPX Manual anticipates the
sole claim at issue there).
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

28
Micron’s Inter Partes Request further explains that Bennett receives
external clock signals and generates internal clock signals “on the device
including when the Driver/Receiver (output driver) circuitry elements output
data.” (Micron’s Request 52 (citing Bennett, col. 274, ll. 59-67; col. 267, ll.
13-16).) Micron also explains, inter alia, that Bazes teaches generating
precision clocking as necessary for integrated memories and that a delay
loop overcomes known problems in timing due to wafer processing, supply
voltage variations, and operating temperature. (Micron’s Request 53.)
Micron reasons that it would have been obvious to employ a delay lock loop
“in order to address the known prior art timing problems.” (Id.) Micron
presents similar rationale with respect to Lofgren. (See id. at 44, 54-55.)
Micron also explains that Bennett discloses internal delays on the memory
chip – i.e., “the internal clocks are distributed throughout the device.” (Id. at
44.)
Rambus fails to rebut Micron’s contentions with persuasive evidence
or reasoning. (See Resp. Br. 9-10.) While Rambus asserts that Micron does
not explain how the DLL would be implemented with Bennett’s timing
scheme, Micron’s explanation shows that, on this record, since Bennett
discloses internal clock delays implemented with that scheme, providing a
DLL to better control the internal timing synchronization and account for
wafer process variations would have been obvious. Moreover, as explained
above, contrary to Rambus’s arguments which attack the references
separately and allege that Micron has not shown how to incorporate a DLL
with Bennett’s dual clocking scheme (see Resp. Br. 10), under Mouttet and
other cited cases supra, the prior art references need not be physically
combined to support obviousness. Rambus’s assertions about complexity
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

29
and cost simply show that trade-offs existed in adding circuitry to chips, but
Rambus does not show that skilled artisans, motivated by providing
precision timing on wafers as the DLL teachings provide, would have been
unable to overcome any cost or complexity as required to add the known
DLL circuits to known DRAM devices.
Based on this record, Micron has shown that the Examiner erred in
failing to maintain the rejection based on Bennett, APA, Wicklund, and
Lofgren or Bazes.

IV. Obviousness Based on Moussouris, SCI-A, APA, SCI-B, and either
Wicklund, Bowater, or Olson
Public Dissemination
The record supports the Examiner’s finding that SCI-A (by Gustavson
et al., see note 10 supra) was probably publicly disseminated and that SCI-B
(by Schanke et al., see note 5 supra) was not. (See RAN 60-62.) As to
SCI-A, it appears that a previous district court judge found the document to
have been publicly disseminated based partially on at least one witness
who participated as one of the authors of the document. (See Hynix Trial
Tr. 1254:12-13.)25 As the Examiner also finds, SCI-A was also listed on
another SCI document which was published by a group of authors at a
European Conference, further corroborating its dissemination. (RAN 60

25 Hynix et al. v. Rambus, Inc., C-00-20905 RMW (Cal. D. Ct.
Mar. 27, 2006) (Transcript of Proceedings before Judge R. Whyte) (10 page
partial submission in the 95/001,109 reexamination proceeding, filed June
12, 2009 as document 238, Ex. D, in the PTO EDAN computer system).
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

30
(citing the Kristiansen et al. publication).)26 It appears that another
(unnamed) witness, apparently another author of SCI-A, testified that he was
“quite certain” that some version of SCI-A was “sent out probably more than
once to people on the mailing list.” (Hynix Trial Tr. 1297:6-10.)
With respect to SCI-B, Micron contends that testimony regarding
SCI-A shows standard operation practices and thereby shows that all
SCI documents, including SCI-B, more likely than not were disseminated.
But Micron’s evidence does not support this contention. (See App. Br. 19-
20.) The record indicates that one witness, Moussouris, an author listed on
SCI-A, relied partly on the distribution mailing list which appeared to have
been attached to the SCI-A document at some point in time after the date
appearing on the document. Moussouris testifies to the effect that the
mailing list (“I notice that the date on this . . . . [on] this mailing list at the
end is November 23rd”) corroborates his understanding first that SCI-A
“may have been” (Hynix Trial Tr. 1239:16-17) mailed, and later that he
was “quite certain it was” mailed (id. at 1240:10) sometime after the listed
August 1988 date. The other (unnamed) author discussed supra also
testifies about the mailing list and responds to the effect that the
August 22, 1989 date corresponds to about the time the witness participated
in writing the document. (See Hynix Trial Tr. 1239:6-22.)
But no such corroborating mailing list is asserted with respect to the
SCI-B document as the Examiner finds. While Micron argues that the

26 As the Examiner finds, the Kristiansen article, i.e., Kristiansen et al., n.1,
Scalable Coherent Interface, European Conf. Proc. (May 10, 1989), cites
Gustavson et al., The Scalable Coherent Interface Project (SuperBus), Draft
(Aug. 22, 1988); i.e., Kristiansen et al. cites the SCI-A document, see note
10 supra.
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

31
evidence shows that the SCI committee was meeting “two or three times a
month,” the citation to the Hynix trial transcript does not support the
contention. (See App. Br. 19 (citing Hynix Trial Transcript at 1240:7-10).)
Micron apparently relies on the testimony that “‘periodically there
were meetings and there were mailings done at the time of the meetings.’”
(See App. Br. 19 (quoting Hynix Trial Tr. at 1239:24-25).) But this
statement, and other statements appearing in the proffered 10 page Hynix
transcript record (supra note 23), do not specifically state how often
meetings were held or which documents were mailed. No adequate
explanation appears as to the standard procedure behind the date
nomenclatures appearing on the SCI documents, e.g., the SCI-A document
(“SCI-22Aug88-doc1”) or the SCI-B document (“SCI-5May89-doc77”).
The testimony shows that the date does not necessarily correspond to a
publication date because, according to Moussouris, the SCI-A document was
probably mailed sometime after the listed August 1988 date. (See Hynix
Trial Tr. 1239:18-19 (“there may have been a delay of some couple of
months).) Micron does not establish what the date nomenclature typically
signifies or otherwise show that the date signifies a mailing.
In an attempt at corroboration, Micron argues that another article by
Gustavson, an SCI member, corroborates the SCI-B dissemination. (See
App. Br. 20.) But that article only shows that members of the project may
have been aware of using dual edges of a clock and a phase locked loop in
November 1988, but it does not show that any such awareness stemmed
from reading the SCI-B document as Micron contends. (See App.
Br. 20.) In other words, the Gustavson article, presented orally at a
November 9-11, 1988 conference according to Micron (see id.), shows a
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

32
date prior to the May 1989 date associated with SCI-B. As a matter of
simple timing and logic, the earlier (November 1988) Gustavson article
cannot not show what the SCI members knew as a result of a later
(May 1989) SCI-B document. As such, the evidence does not corroborate
the dissemination of SCI-B.
Based on the foregoing discussion, Micron’s reliance on Constant v.
Advanced Micro-Devices, Inc., 848 F.2d 1560, 1568-69 (Fed. Cir. 1988) is
misplaced. (See App. Br. 19-20.) In Constant, the court cited “extensive
and uncontroverted evidence of business practice.” The evidence presented
here is much more limited. Constant, 848 F.2d at 1569, relies on In re Hall,
781 F.2d 897 (CAFC 1986) for the proposition that evidence of routine
business practice can show dissemination. But in Hall, the “affidavits give a
rather general library procedure as to indexing, cataloging, and shelving
theses.” Id. at 899. Micron’s evidence falls short of showing the general
practice associated with the SCI dates, e.g., the “SCI-5May89-doc77”
nomenclature appearing on the SCI-B document. See also in re Lister, 583
F.3d 1307, 1317 (Fed. Cir. 2009) (copyright date on manuscript in
Copyright Office insufficient as prima facie evidence of public
dissemination where Copyright Office’s automated catalog only allowed
searching by author’s name or first word in title, and there was no evidence
of when the document was listed in a keyword searchable data base).
Rambus’s contention that “Micron did not establish that Moussouris is
a printed publication” lacks merit. Rambus does not appear to contest the
Examiner’s finding that at least one version of the document, relied upon by
the Examiner, was part of a published “Computer Letter” having “Vol. 54,
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

33
No. 25” listed thereon. (Compare Reb. Br. 10 and RAN 61, n.1, with Resp.
Br. 11.)
Moussouris and SCI-A Teachings
Moussouris describes how “standard interfaces provided by chip
vendors have improved in bandwidth incredibly slowly relative to the
underlying transistor improvements.” (Moussouris 2.) Moussouris explains
that the market created a bandwidth gap in “DRAM density” versus DRAM
bandwidth, and that “[t]here’s no technical reason for this.” (Id.)
Moussouris then describes “a new standard . . . based on extremely
fast signaling technology,” implemented by “an IEEE working group called
P1596,” and notes that “if you . . . applied it to a 40-pin, surface mount
RAM package, you would exactly close this [DRAM bandwidth] gap.” (Id.
at 2.) Moussouris refers to the technology as “Scalable Coherent Interface”-
i.e., SCI. (Id.)27 Moussouris further explains that “chip interfaces could be
just the P1596 style interfaces, and with 32 megabytes of memory on each
chip, there would be adequate bandwidth to connect together a few hundred
like this. So nothing very exotic is required for this to happen.”
(Moussouris 3.) Moussouris and lists a 500MHz speed and refers to “fast
RAMS.” (Moussouris 3, slides.)
SCI-A discloses packets for a typical SCI read operation from a
memory. (SCI-A at 7.) “While the memory is looking for data, SCI is freed
for other operations.” (Id.) SCI-A states that “[w]e are still working on
error handling and initializations” but “block parity checking is currently
assumed.” (Id. at 8, 14.) “SCI provides synchronization mechanisms . . . in

27 One set of presentation slides attached to the Moussouris article refers to
“Scalable Coherent Interface (IEEE P1596)” and 500Mhz speeds.
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

34
a multiprocessor system.” (Id.) Figures 14 and 15 show proposed node
schemes and a corresponding packet data format. Figure 14 and an earlier
discussion of “any real system,” indicates that a system clock and strobe
perform synchronized communication, even though the “the phase of this
clock with respect to data strobes will vary from place to place.” (See id. at
12.)
Obviousness
The Examiner reasons that the combination involving Moussouris,
SCI-A, APA, and Wicklund, Bowater or Olson, does not support the
rejection because “the Request . . . did not provide an obviousness type of
support for this combination.” (See RAN 64.)
Micron’s Inter Partes Request points out, inter alia, that Moussouris
discloses the use of SCI signaling technology on a RAM to improve the
DRAM bandwidth, and it relies on SCI-A to disclose using an external clock
system and synchronous communication. (See Micron Request 22.)
Micron’s Brief relies on this and similar rationale to support obviousness.
(See App. Br. 20.) The Examiner appears to agree to a certain extent: “The
Examiner agrees that one of ordinary skill in the art, after reading
Moussouris, would look to Gustavson (SCI-A), however the Examiner was
pointing out that SCI in Moussouris is not shown to be the same as SCI in
Gustavson.” (RAN 64.) As Micron points out, skilled artisans would have
looked to published SCI documents dealing with related technology based
on the Moussouris reference to SCI (and the P1596 working group) and
specifically, references to that scalable interface technology to improve
DRAMs. (See App. Br. 20-21; Reb. Br. 11.) Further, Moussouris provides
motivation for such DRAM interfaces in terms of increased speed – up to
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

35
“five hundred million transfers per second” – i.e., 500MHz. (Moussouris
at 2.)
Rambus maintains that SCI-A and Moussouris are non-enabling as
lacking sufficient detail and as being general and forward-looking “works-
in-progress.” (See Resp. Br. 11-12) But these references are presumptively
enabling for what they teach and such references can be considered under
obviousness independent of whether they are enabling or describe an
operable device. See In re Antor Media Corporation, _F.3d__, WL
3055928, slip op. at *5, 6, 7-9 (Fed. Cir. July, 2012) (also stating that any
“forward-looking language” does not address undue experimentation
involved in the enablement query, addressing enablement in terms of the
claim scope, and quoting Beckman Instruments, Inc. v. LKB Produkter AB,
892 F.2d 1547, 1551 (Fed.Cir.1989) (“Even if a reference discloses an
inoperative device, it is prior art for all that it teaches.”)); see also In re
Paulsen, 30 F.3d 1475, 1480-81 (Fed. Cir. 1994) (enabling references need
not describe computer circuit details where highly skilled artisans could
have provided such circuitry).
As Micron contends, Rambus does not set forth which one of the
claim elements are not enabled. Rambus has not demonstrated that any
undue experimentation would have been required to modify a known
DRAM to render it synchronous – according to the combined teachings of
Moussouris, APA, and SCI-A. (See Reb. Br. 10-11.) As Micron also points
out, Moussouris indicates that using a synchronous interface system in a
DRAM would improve the bandwidth and speed. Bennett constitutes
evidence that skilled artisans knew how to provide synchronous single chip
memory devices. The record here, including expert opinion evidence, the
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

36
references, and other secondary evidence, shows that the skill level involved
was quite advanced and the ‘446 patent itself is somewhat conceptual in
detail as Micron notes. (See Reb. Br. 10.)
As indicated supra, Moussouris describes what occurs in the
SCI system, and the SCI-A document pertains to part of that description.
Moussouris proposes a solution which involves putting SCI standard
interfaces on standard memory chips, which includes DRAMs, to improve
the bandwidth thereof as necessary to keep pace with the DRAM memory
density. The SCI-A document shows that SCI standard interfaces operated
synchronously. Because, according to Moussouris, “nothing very exotic is
required for this to happen” and the DRAM bandwidth to density gap was
not due to technical reasons, Rambus’s generic enablement contentions,
lacking in specific detail as to specific claim elements, and lacking any
showing of undue experimentation, are not supported.
Rambus’s similarly general contention that Micron failed to identify
how the “combination meets all the claim limitations and makes the
combination obvious” lacks sufficient specificity. (See Resp. Br. 12.) The
record reflects at the least, that Micron’s Request, including detailed claim
charts, the Appeal Brief, the Rebuttal Brief, and the Examiner’s findings,
constitute a prima facie case of obviousness. For example, contrary to
Rambus’s assertions, Micron explains how Bennett, SCI-A, Wicklund,
Bowater, or Olson, support the obviousness of a precharge signal in an
operation code. SCI-A discloses operation codes and Wicklund’s system
determines whether to switch between page-mode and non-page mode,
which amounts to a decision to precharge or not. (See Request 16-18, 22-
34.) As is the case with the Bennett-based rejection, Micron also
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

37
persuasively relies on admissions in the ‘446 patent to show that precharging
was known and followed normal read requests, thereby bolstering
obviousness. (See, e.g., Micron Request 24.)
Micron’s responses countering Rambus’s contentions are persuasive.
The Examiner’s main reason for not maintaining the rejection appears to be
that Micron did not propose a proper obviousness rejection and only
attempted to “incorporate [the] SCI-A teaching into Moussouris.” (See RAN
63.) Based on the foregoing discussion, Micron shows error in the
Examiner’s decision not to maintain Micron’s proposed obviousness
rejection of claims 1-2 based on Moussouris as combined with the above-
listed prior art. Micron does not show error in the Examiner’s decision not
to maintain the obviousness rejection of claims 3-4 based on the added
teachings of SCI-B.
V. Priority (relative to the JEDEC and Park references)
Micron asserts that since claims 1-4 do not recite a multiplexed bus
and that claim 4 recites a delay lock loop (DLL), the claims are not
originally supported back to the (first-filed) ‘898 application having a filing
date of April, 1990. The first contention is that the claims are too broad; i.e.,
too broad absent a recitation to a multiplexed bus which the ‘446 patent touts
as important. The second contention is that the original ‘898 application
does not adequately describe a DLL. (See App. Br. 21-34.)
Based on these contentions, Micron asserts that the claims are not
entitled to a filing date prior to the ‘446 patent’s application filing date of
December 21, 2001, and are therefore anticipated by the JEDEC and Park
references which antedate the 2000 date. (See App. Br. 21, 34.) Rambus’s
procedural attack, asserting that Micron does not have standing to argue
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

38
rejections raised by requester Samsung (Resp. Br. 11-12), has been
addressed in repeated petition decisions, including one in this proceeding,
which decided against Rambus as Micron and Rambus point out. (See Reb.
Br. 11-12; Resp. Br. 13, n. 10; Pet. Dec. (June 17, 2011).) Prior decisions of
the Board have relied upon one or more of these petition decisions. See,
e.g., BPAI 2012-000168. Essentially, Micron’s appeal is an appeal of the
Examiner’s decision and Micron has standing to argue rejections adverse to
its position where the Examiner has addressed those rejections in the RAN.
Rambus’s assertion that the Examiner improperly determined a
written description instead of a priority issue lacks merit. (See App. Br. 13.)
The Examiner determined priority and the Office has authority to determine
such issues.
A. The Multiplex Bus
Rambus relies, inter alia, on Rambus Inc. v. Infineon Technologies
AG, 318 F.3d 1081, 1094-95 (Fed. Cir. 2003) to show that the claims do
not require a multiplex bus and are therefore originally supported. (Resp.
Br. 13.) The Examiner agrees with Rambus that the claims are originally
supported. (See RAN 9-21 (incorporating by reference previous office
actions).)
The Board’s related decisions, BPAI App. Nos. 2012-000142,
2012-000168, and 2012-000169, address the same or similar issue between
the same parties. Our decision, analysis and findings there are adopted and
incorporated by reference herein.
As discussed in our prior decisions, Infineon held that the ordinary
term “bus” could not be read restrictively as a “multiplexed bus” even
though the patentee described the “present invention” in terms of a
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

39
“multiplexed bus” in isolated portions of the specification because “the
remainder of the specification and the prosecution history shows that
Rambus did not clearly disclaim or disavow such claim scope in this case.”
Infineon, 318 F.3d at 1094-95.
Micron maintains that the claims violate the written description
requirement since the ‘898 patent application touts the importance of a
multiplexed bus and distinguishes prior art generic bus inventions. (See
App. Br. 31-32.) Micron supports the theory, which in essence, amounts to
a scope of enablement attack on the claims, by relying, inter alia, on
LizardTech, Inc. v. Earth Resource Mapping, Inc., 424 F.3d 1336, 1344
(Fed. Cir. 2005) (generic seamless DWT claim too broad absent an updated
sums limitation). (App. Br. 23-24, 34) Micron relies on a “key factor” in
LizardTech as embodied in an analogy there to an inventor who describes a
fuel-efficient engine in such detail that it would not necessarily support “a
broad claim to every possible type of fuel-efficient engine.” (App. Br. 24
(quoting LizardTech, 424 F.3d at 1346).) As another example, Micron
reasons that the claims here are analogous to the “‘spikeless’” valve claims
addressed in ICU Medical. (App. Br. 24 (quoting ICU Medical, Inc. v.
Alaris Medical Systems, Inc., 558 F.3d 1368, 1377 (Fed. Cir. 2009)).
But the gravamen of these arguments is that the ‘446 patent inventors
were not in possession of broad claims not requiring a multiplexed bus
scheme including memory devices with interfaces for such multiplexing.
The ‘446 patent refers to “a preferred implementation” as using only bus
connections of the invention –signifying more generic bus connections or no
bus connections. (See ‘446 patent, col. 5, ll. 62-68.) Moreover, Infineon’s
claim construction analysis, at the minimum, implies that skilled artisans
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

40
were in possession of generic bus claims. See Infineon, 318 F.3d at 1094-95
(noting that “multiplexing is not a requirement in all of Rambus’s claims”
and that the PTO issued a restriction to a multiplexing group and a latency
group, that “the PTO demonstrated an understanding of “bus” that is not
limited to a multiplexing bus”).
Original claims 73 and 91 in the first-filed ‘898 application recite a
generic bus, showing the possession of generic bus claims. Also, as Rambus
points out, “the original disclosure describes many inventions.” (Resp.
Br. 13.) Based on the original disclosure, skilled artisans would have
understood that other important touted features in the ‘898 disclosure,
including clocking schemes and writing blocks of data, could have been
practiced on generic buses without a multiplexing interface. Accord
Infineon, 318 F.3d at 1095 (“a multiplexing bus is only one of many
inventions disclosed in the ‘898 application”); cf. Crown Packaging Tech.
Inc. v. Ball Metal Beverage Container Corp., 635 F.3d 1373, 1382-84 (Fed.
Cir. 2011) (district court erred in finding lack of written description in
generic claims where the application discloses separate solutions to related
problems).
The claims at issue require synchronous reading of data. Skilled
artisans would have recognized that these elements could have been
practiced on known buses, whether multiplexed or not. The lack of
multiplexing would have been much simpler than a multiplexing scheme.
Cf. Hynix Semiconductor Inc. v. Rambus Inc., 645 F.3d 1336, 1351-53
(2011) (holding that evidence supported jury verdict of written description
for similar Rambus claims where “the supposed genus consists of only two
species, a multiplexed bus and a non-multiplexed bus”).
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

41
Also, Micron’s spikeless valve analogy is not entirely apt here
because valves were recited in the ICU Medical claims, but in the
‘446 patent claims, neither a bus nor an interface for attaching to a bus is
recited, so ICU Medical does not dictate that the claims must support any
type of bus or bus interface. But even if the claims implicitly require such a
bus scheme, Rambus, the Examiner, the Infineon claim construction, and the
Hynix Semiconductor written description analysis, show that on this record,
the inventors originally possessed inventions directed to a generic bus
scheme.
B.) The DLL
Micron asserts that the Examiner improperly construed the delay lock
loop (DLL) limitation in claim 4. Micron considers the avowed support for
claim 4, i.e., Figure 12 of the ‘446 patent, to show a clock averaging circuit,
and not a circuit that adjusts the delay of an internal clock with respect to an
external clock. Micron also maintains that the internal clock is “not locked
to either external clock signal.” (App. Br. 26.) Micron explains that that
Figure 12 shows a clock that “synchronizes data output to a midpoint
between two external clock signals” but does not “synchronize the output of
. . . data with the external clock signal” as claim 4 requires. (App. Br. 27.)
Micron’s argument does not show a lack of written description.
Synchronizing an internal clock with the average of two external clock
signals amounts to synchronizing not only with either of the two clock
signals which define the average, but also with the clock generator which
generates the two signals. (See Resp. Br. 13-14 (Rambus stating that
Micron’s contentions do not show a distinction).) That is, Figure 12
signifies an early clock and a late clock, but there is only one clock
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

42
generator. For example, Figures 8A and 8B show one clock generator 50
creating an early clock1 and a late clock2, the latter due to delays on the
bus. Thus, synchronizing to the average of the two clock signals in the
‘446 patent amounts synchronizing to the external clock generator 50.
Micron also asserts that Rambus argued a distinction between a phase
locked loop (PLL) and a delay lock loop (DLL) in other proceedings and
that any argued distinction should be binding. (App. Br. 28.) However, the
Board previously essentially bound Rambus and determined that Rambus’s
experts had essentially equated PLL and DLL circuits in determining the
meaning of DLL (see supra the DLL discussion involving Bazes and
Lofgren). In other words, Rambus’s experts essentially generically grouped
DLL’s and PLL’s.
The Examiner also cites and quotes several experts who generally
support the position that Figure 12 reasonably supports a DLL. (See
RAN 11-12.) The Examiner also relies on prior art patents to Petrich and
Bazes. (See RAN 14-18.) While Micron maintains both that the external
clock and internal clock in a DLL must be locked together and that “the
internal clock is aligned with the time of the external clock” (App. Br. 26),
the expert testimony, for example Mr. Karp, and Mr. Murphy, indicate that
delay with respect to the external clock and feedback is required, but not
necessarily strict alignment of the internal clock with an external clock. The
Examiner also points to a previous claim construction order showing a
consistent definition. (See RAN 14.)
The record supports the Examiner. Claim 4 recites a delay lock loop
and Figure 12, in light of Figures 8A and 8B, supports the concept of a DLL
as recognized by some of the experts relied upon by the Examiner. And
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

43
even if the term “delay lock loop” does not comport strictly with a definition
presented by Micron, Figure 12 supports the general concept of a delay lock
loop – especially where its meaning includes other characteristics, e.g.,
phase lock loop characteristics. See Application of Anderson, 471 F.2d
1237, 1244 (CCPA 1973) (“The question . . . is not whether ‘carrying’ was a
word used in the specification as filed but whether there is support in the
specification for employment of the term in a claim; is the concept of
carrying present in the original disclosure?”) Skilled artisans would have
recognized that the inventors of the ‘446 patent originally possessed a delay
lock loop in light of Figures 8A, 8B and 12 and accompanying discussions
in the ‘446 patent.
Based on the foregoing discussion, Micron has not shown error in the
Examiner’s finding that claims 1-4 have original written description support
as necessary to antedate Park or JEDEC as prior art references.
Secondary Considerations
Rambus contends that substantial secondary evidence supports
unobviousness. But the evidence fails to establish a nexus because any
success likely flows from a variety of several unclaimed features touted here
or in other Rambus proceedings or patents. Such unclaimed, but disclosed
features, include eight data lines, small DRAM sizes with minimal bus
loading, multiplexed bus architecture and device interfaces, packetized
control, unique device identifiers, time access and arbitration schemes, a
500 MHz data rate, and memory devices having all the functionality of prior
art circuit boards. (See ‘446 patent, col. 3, ll. 33-53; col. 4, ll. 5-9, 44-53;
col. 7, ll. 11-23; col. 9, ll. 56-59; col. 12, ll. 45-58; col. 14, ll. 48-50.) See
also Rambus Inc. v. Infineon Technologies AG, 318 F.3d 1081, 1095 (Fed.
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

44
Cir. 2003) (“The present invention is designed to provide a high speed,
multiplexed bus’” (quoting Rambus’s related ‘918 patent, col. 5, ll. 36-46),
but “the prosecution history shows that a multiplexing bus is only one of
many inventions disclosed in the ‘898 application.”)28 Rambus agrees that
“the original disclosure describes many inventions.” (Resp. Br. 13.)
As Micron points out, Rambus presents the same evidence it provided
“in every one of the 10 pending inter partes reexaminations.” (Reb. Br. 16.)
Different claims in the different proceedings have varying scope. As Micron
also points out, Rambus’s proffered evidence is short on objectivity and
relies on interested witnesses. (See id.)
Rambus’s evidence does not demonstrate that any success was due
solely to the claimed features, or to claimed features that were not already
known in the prior art. For example, Rambus’s contention “that the
Farmwald family, which includes the ‘446 patent, has numerous licensees”
is a vague statement which lacks a specific nexus to the claims. (See Resp.
Br. 15.) Rambus does not provide a copy of any licenses (even if one does
pertain to the ‘446 patent) or provide evidence showing what other
unclaimed features any of the licenses involve. Also, it is well known and
settled law that competitors often take licenses for commercial or other
reasons having nothing to do with unobviousness.
Single chip synchronous memory devices were known as Bennett
discloses, as were DRAMs, precharging and DLL circuits, as the prior art

28 See Infineon, 318 F.3d at 1084-86 (finding that the written descriptions of
each of four related Rambus patents, 5,954,804; 5,953,263; 6,034,918; and
6,032,214, are substantially identical to the written description of the
07/510,898 application to which they, and the ‘446 patent here, all claim
continuity).
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

45
discussed supra discloses. For example, “the most popular form of
read/write memory is the semiconductor DRAM.” (Wicklund, col. 1, ll. 36-
39.) Hence, the record suggests that at least part of any commercial success
would have been due to “the most popular” memory chip, a DRAM chip, in
general (or to Bennett’s known synchronous chip in general). Cf. In re
DBC, 545 F.3d 1373, 1384 (Fed. Cir. 2008) (Board’s conclusion of
nonobviousness supported, the Board finding, inter alia, that “evidence in
the record suggested that the success of XanGo™ juice may be due to other
factors-for example, the increasing popularity of the mangosteen fruit in
general” ). The Federal Circuit further reasoned in DBC that
. . . DBC has done little more than submit evidence of sales.
However substantial those sales, that evidence does not reveal
in any way that the driving force behind those sales was the
claimed combination of mangosteen fruit, mangosteen rind
extract, and fruit or vegetable juice. Nor is there any evidence
that sales of XanGo™ juice were not merely attributable to the
increasing popularity of mangosteen fruit or the effectiveness of
the marketing efforts employed.

Id. at 1384. See also Ormco Corp. v. Align Technology, Inc., 463 F.3d
1299, 1312, 1313 (Fed. Cir. 2006) (“[I]f the feature that creates the
commercial success was known in the prior art, the success is not pertinent.”
Reasoning that success that is due “‘partially’ to claimed features” and to
unclaimed features and/or other features already in the art lacks the requisite
nexus to show unobviousness.) (citations omitted).  
While Rambus argues that the claims solve a memory bottleneck
problem and obtain high-speed performance (Resp. Br. at 14), the claims
read on slow memory devices, since the claims do not recite any speed (as a
functional limitation) and do not recite other sufficient and necessary
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

46
circuitry to obtain such speed.29 Moreover, Dr. Farmwald testified that
“even up into the early part of the ‘90s, it [speed] wasn’t going to be a
problem.” (See Farmwald Trial Dep. 276 (attached as Resp. Br. Evidence
Ex. E-5).) In other words, Dr. Farmwald may have solved a problem
predicted to occur in the future, but not a long-standing problem.
Rambus’s allegations of recognition and praise for the “‘high
bandwidth memory-interface technology’” and bandwidth advances “‘as a
result of the ideas [Dr. Horowitz] pioneered’” (Resp. Br. 15) point to a lack
of nexus as to any success or praise. The devices claimed here have no
bandwidth limitation, let alone limitations directed to the myriad other
“ideas [Dr. Horowitz] pioneered” – whatever they may have been.
Based on the foregoing discussion, the record suggests that the
proffered evidence is not commensurate with the claim scope and lacks a
nexus thereto. Rambus has not demonstrated that any success is not due to
popular DRAMs in general, synchronous memory chips in general, or, to a
whole host of unclaimed features, including the unclaimed but touted
multiplexed bus interface, high bandwidth and/or speed, and other
unclaimed circuit features, such as the identification feature, arbitration
control features, low capacitance and power, etc. The record indicates that

29 In another reexamination proceeding, similar to the proceeding here
(Resp. Br. 15), Rambus alleges “disbelief” and pervasive skepticism “‘over a
500 megabit per second DRAM data rate’” and “about many of the specific
features of the technology” as showing “‘strong evidence of
nonobviousness.’” (See Rambus Resp. Br. 19 (citations omitted) in the
BPAI 2012-000142 reexamination proceeding.) Assuming for the sake of
argument that uncorroborated statements by the inventors show skepticism
by others, as noted, claims 1-4 do not require the 500 MHz speed touted or
“many of the other specific features” – whatever they may be.
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

47
such features (and other pioneering ideas) would have been required to
obtain the touted high speed from a single DRAM. See Infineon, 318 F.3d at
1095 (quoted supra, mentioning Rambus’s high speed multiplexed system).
After careful consideration of the record, the obviousness of
combining known precharge, DRAM memory, and DLL, with Bennett’s
synchronous memory device, and combining a known synchronizing
SCI interface and precharge circuit with Moussouris’s DRAM, outweighs
the proffer of secondary considerations.

CONCLUSION
Micron has demonstrated that the following rejections are warranted
based on this record: claims 1 and 2 as obvious based on Bennett, APA, and
either Wicklund or Bowater; claim 3 as obvious based on Bennett, Wicklund
or Bowater, and either Inagaki or Novak; claim 4 as obvious based on
Bennett, either Wicklund or Bowater, and either Bazes or Lofgren; and
claims 1 and 2 and as obvious based on Moussouris, APA, SCI-A, and either
Wicklund or Bowater.
Micron has not demonstrated that the Examiner erred in deciding not
to reject claims 3 and 4 based on the added teachings of SCI-B since Micron
has not demonstrated that SCI-B is prior art. Micron also has not
demonstrated that the Examiner erred in deciding not to reject claims 1-4 as
anticipated by JEDEC or as obvious based on Park and JEDEC since the
references do not antedate the effective filing date of claims 1-4.

Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

48
DECISION
The Examiner’s decision not to reject claims 1-4 is reversed.30
This decision contains a new ground of rejection pursuant to
37 C.F.R. § 41.77(b), which provides that “[a]ny decision which includes a
new ground of rejection pursuant to this paragraph shall not be considered
final for judicial review.” Correspondingly, no portion of the decision is
final for purposes of judicial review. A requester may also request rehearing
under 37 C.F.R. § 41.79, if appropriate; however, the Board may elect to
defer issuing any decision on such request for rehearing until such time that
a final decision on appeal has been issued by the Board.
For further guidance on new grounds of rejection, see 37 C.F.R.
§ 41.77(b)-(g). The decision may become final after it has returned to the
Board. 37 C.F.R. § 41.77(f).
37 C.F.R. § 41.77(b) also provides that the Patent Owner, WITHIN
ONE MONTH FROM THE DATE OF THE DECISION, must exercise one
of the following two options with respect to the new grounds of rejection to
avoid termination of the appeal as to the rejected claims:
(1) Reopen prosecution. The owner may file a response requesting
reopening of prosecution before the examiner. Such a response must be
either an amendment of the claims so rejected or new evidence relating to
the claims so rejected, or both.
(2) Request rehearing. The owner may request that the proceeding be
reheard under § 41.79 by the Board upon the same record.

30 See 37 C.F.R. § 41.77(b) (denominating a reversal of a refusal to reject as
a new ground of rejection).
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

49
Any request to reopen prosecution before the examiner under
37 C.F.R. § 41.77(b)(1) shall be limited in scope to the “claims so rejected.”
Accordingly, a request to reopen prosecution is limited to issues raised by
the new ground(s) of rejection entered by the Board. A request to reopen
prosecution that includes issues other than those raised by the new ground(s)
is unlikely to be granted. Furthermore, should the patent owner seek to
substitute claims, there is a presumption that only one substitute claim would
be needed to replace a cancelled claim.
A requester may file comments in reply to a patent owner response.
37 C.F.R. § 41.77(c). Requester comments under 37 C.F.R. § 41.77(c) shall
be limited in scope to the issues raised by the Board’s opinion reflecting its
decision to reject the claims and the patent owner’s response under
paragraph 37 C.F.R. § 41.77(b)(1). A newly proposed rejection is not
permitted as a matter of right. A newly proposed rejection may be
appropriate if it is presented to address an amendment and/or new evidence
properly submitted by the patent owner, and is presented with a brief
explanation as to why the newly proposed rejection is now necessary and
why it could not have been presented earlier.
Compliance with the page limits pursuant to 37 C.F.R. § 1.943(b), for
all patent owner responses and requester comments, is required.
The examiner, after the Board’s entry of a patent owner response and
requester comments, will issue a determination under 37 C.F.R. § 41.77(d)
as to whether the Board’s rejection is maintained or has been overcome.
The proceeding will then be returned to the Board together with any
comments and reply submitted by the owner and/or requester under
Appeal 2012-001639
Reexamination Control Nos. 95/001,109 & 95/001,155
Patent 6,546,446 B2

50
37 C.F.R. § 41.77(e) for reconsideration and issuance of a new decision by
the Board as provided by 37 C.F.R. § 41.77(f).
Extensions of time for taking action under 37 C.F.R. § 41.77(b) are
governed by 37 C.F.R. § 41.77(g). See also 37 C.F.R. § 41.77 regarding
extensions of time for requesting rehearing.
REVERSED

Finnegan, Henderson, Farabow,
Garrett & Dunner, LLP
901 New York Avenue, NW
Washington, DC 20001-4413

Third Party Requesters:

Haynes & Boone, LLP
IP Section
2323 Victory Avenue
Suite 700
Dallas, TX 75219

Novak Druce & Quigg, LLP
1000 Lousiana Street
53rd Floor
Houston, TX 77002