The IS61NSCS series
the SigmaRAM pinout standard for synchronous SRAMs.
The implementations are 18,874,368-bit (18Mb) SRAMs.
These are the first in a family of wide, very low voltage
I/O SRAMs
implement economical high performance networking
systems.
ISSI’s
emulate other synchronous SRAMs, such as Burst RAMs,
NBT RAMs, Late Write, or Double Data Rate (DDR) SRAMs.
The logical differences between the protocols employed by
these RAMs hinge mainly on various combinations of
address bursting, output data registering and write cueing.
ΣΣ
ΣRAMs allow a user to implement the interface protocol best
ΣΣ
suited to the task at hand.
This specific product is Common I/O, SDR, Double Late
Write & Pipelined Read (same as Pipelined NBT) and in
the family is identified as 1x1Dp.
ADVANCE INFORMATION
JUNE 2001
Bottom View
209-Bump, 14 mm x 22 mm BGA
1 mm Bump Pitch, 11 x 19 Bump Array
ΣΣ
ΣRAMs are built in compliance with
ΣΣ
CMOS
designed to operate at the speeds needed to
ΣΣ
ΣRAMs are offered in a number of configurations that
Integrated Silicon Solution, Inc. — 1-800-379-4774
ADVANCE INFORMATION Rev. 00A
06/19/01
1
IS61NSCS25672
IS61NSCS51236ISSI
Functional Description
Because SigmaRAM is a synchronous device, address,
data Inputs, and read/write control inputs are captured on
the rising edge of the input clock. Write cycles are
internally self-timed and initiated by the rising edge of the
clock input. This feature eliminates complex off-chip write
pulse generation required by asynchronous SRAMs and
simplifies input signal timing.
IS61NSCS25672 PINOUT
256K x 72 Common I/O—Top View
1234567891011
Single data rate ΣRAMs incorporate a rising-edge-triggered
output register. For read cycles, ΣRAM’s output data is
temporarily stored by the edge-triggered output register
during the access cycle and then released to the output
drivers at the next rising edge of clock.
IS61NSCS series
high performance CMOS technology and are packaged in
a 209-bump BGA.
Integrated Silicon Solution, Inc. — 1-800-379-4774
ADVANCE INFORMATION Rev. 00A
06/19/01
IS61NSCS25672
IS61NSCS51236ISSI
IS61NSCS51236 PINOUT
512K x 36 Common I/O—Top View
1234567891011
ANCNCAE2AADVAE3ADQbDQb
(16M)
BNC NC BcNCAWABbNCDQbDQb
(x36)
CNC NC NC BdNCE1NCNCBaDQbDQb
(128M)
DNCNCGNDNCNCMCLNCNCGNDDQbDQb
ENCDQPc VCCQ V CCQVCCVCCVCC V CCQ V CCQNCDQPb
FDQcDQcGNDGNDGNDZQGNDGNDGNDNCNC
GDQcDQc VCCQ V CCQVCCEP2VCC V CCQ V CCQNCNC
HDQcDQcGNDGNDGNDEP3GNDGNDGNDNCNC
JDQcDQc VCCQ V CCQVCCM4VCC V CCQ V CCQNCNC
KCQ2CQ2CLKNCGNDMCLGNDNCNCCQ1CQ1
Integrated Silicon Solution, Inc. — 1-800-379-4774
ADVANCE INFORMATION Rev. 00A
06/19/01
5
IS61NSCS25672
IS61NSCS51236ISSI
BACKGROUND
The central characteristics of the ISSI ΣRAMs are that
they are extremely fast and consume very little power.
Because both operating and interface power is low,
ΣRAMs can be implemented in a wide (x72) configuration,
providing very high single package bandwidth (in excess
of 20 Gb/s in ordinary pipelined configuration) and very low
random access latency (5 ns). The use of very low voltage
circuits
in the core and 1.8V or 1.5V interface voltages allow
the speed, power and density performance of ΣRAMs.
Although the
to support a number of different common read and write
protocol options, not all SigmaRAM implementations will
support all possible
provide a quick comparison between read and write
protocols options available in the context of the SigmaRAM
Sigma
RAM
family pinouts
protocols. The following timing diagrams
have been designed
COMMON I/O SigmaRAM FAMILY MODE COMPARISON—LATE WRITE VS. DOUBLE LATE WRITE
standard. This data sheet covers the single data rate
DDR)
, Double Late Write, Pipelined Read SigmaRAM.
The character of the applications for fast synchronous
SRAMs in networking systems are extremely diverse.
ΣRAMs have been developed to address the diverse
needs of the networking market in a manner that can be
supported with a unified development and manufacturing
infrastructure. ΣRAMs address each of the bus protocol
options commonly found in networking systems. This
allows the ΣRAM to find application in radical shrinks and
speed-ups of existing networking chip sets that were
designed for use with older SRAMs, like the NBT or Nt,
Late Write, or Double Data Rate SRAMs, as well as with
new chip sets and ASIC’s that employ the Echo Clocks
and realize the full potential of the ΣRAMs.
(non-
®
Double Late Write—Pipelined Read (
CK
Address
Control
DQ
CQ
Late Write—Pipelined Read (
CK
Address
A B C D E F
R W R W R W
QA DB QC DD QE
ΣΣ
Σ1x1Lp). For reference only.
ΣΣ
A B C D E F
ΣΣ
Σ1x1Dp). For reference only.
ΣΣ
6
Control
DQ
CQ
R X W R X W
QA DC QD DF
Integrated Silicon Solution, Inc. — 1-800-379-4774
ADVANCE INFORMATION Rev. 00A
06/19/01
IS61NSCS25672
IS61NSCS51236ISSI
®
Double Data Rate Write—Double Data Rate Read (
CK
Address
Control
DQ
CQ
A B C D E F
R X W R X W
QA0 QA1 QD0 QD1
ΣΣ
Σ1x2Lp). For reference only.
ΣΣ
DC0
Mode Selection Truth Table Standard
NameM2M3M4FunctionAnalogous to...In This Data Sheet?
Σ1x2Lp011
Σ1x1Dp101
Double Late Write, Pipelined ReadPipelined NBT SRAM
DDR
Double Data Rate SRAM
No
Yes
DF0DC1
Σ1x1Lp110
Notes:
All address, data and control inputs (with the exception of EP2, EP3, and the mode pins, M2–M4) are synchronized to rising clock
edges. Read and write operations must be initiated with the Advance/Load pin (ADV) held low, in order to load the new address.
Device activation is accomplished by asserting all three of the Chip Enable inputs (E1, E2, and E3). Deassertion of any one of the
Enable inputs will deactivate the device. It should be noted that ONLY deactivation of the RAM via E2 and/or E3 deactivates the
Echo Clocks, CQ1–CQn.
READ OPERATIONS
Pipelined Read
Read operation is initiated when the following conditions
are satisfied at the rising edge of clock: All three chip
enables
signal
The address presented to the address inputs is latched into
the address register and presented to the memory core
and control logic. The control logic determines that a read
access is in progress and allows the requested data to
(E1, E2, and E3)
(W)
is deasserted high, and ADV is asserted low.
are active, the write enable input
Late Write, Pipelined ReadPipelined Late Write SRAM
WRITE OPERATIONS
Write operation occurs when the following conditions are
satisfied at the rising edge of clock: All three chip enables
(E1, E2, and E3) are active and the write enable input
signal (W) is asserted low.
Double Late Write
Double Late Write means that Data In is required on the
third rising edge of clock. Double Late Write is used to
implement Pipeline mode NBT SRAMs.
No
propagate to the input of the output register. At the next
rising edge of clock the read data is allowed to propagate
through the output register and onto the output pins.
Integrated Silicon Solution, Inc. — 1-800-379-4774
ADVANCE INFORMATION Rev. 00A
06/19/01
7
IS61NSCS25672
IS61NSCS51236ISSI
Single Data Rate Pipelined Read
CLK
®
Address
E1
W
DQ
CQ
A XX C D E F
Read Deselect Read Read Read
Double Late Write with Pipelined Read
QA QC QD
CLK
Address
E1
W
DQ
CQ
8
A B C D E F
QA DB QC DD
Read Write Read Write Read Write
Integrated Silicon Solution, Inc. — 1-800-379-4774
ADVANCE INFORMATION Rev. 00A
06/19/01
IS61NSCS25672
IS61NSCS51236ISSI
SPECIAL FUNCTIONS
®
Slow Down Mode
The SD pin allows the user to activate a delay element in
the on-chip clock chain that is routed to the data and echo
Clock output drivers. Activating Slow Down mode by
pulling the SD pin low introduces extra delay in every
Burst Order
The burst address counter wraps around to its initial state
after four addresses (the loaded address and three more)
have been accessed. SigmaRAMs always count in linear
burst order.
synchronous output driver specification. Address, control
and data input specifications are not affected by Slow
Down Mode. See “Slow Down Mode Clock to Data Out and
Clock to Echo Clock Timing” table for specifics.
Linear Burst Order
Burst Cycles
ΣΣ
ΣRAMs provide an on-chip burst address generator that
ΣΣ
can be utilized, if desired, to further simplify burst read or
write implementations. The ADV control pin, when driven
high, commands the
ΣΣ
ΣRAM to advance the internal ad-
ΣΣ
dress counter and use the counter generated address to
read or write the
cycle in a burst cycle series is loaded into the
1. The burst counter wraps to initial state on the 5th rising edge
of clock.
Sigma Pipelined Burst Reads with Counter Wrap-around
A[1:0]A[1:0] A[1:0] A[1:0]
CLK
External
Address
Internal
Address
E1
W
ADV
DQ
CQ
A2 XX XX XX XX XX
A2 A3 A0 A1 A2 A3
Counter Wraps
QA2 QA3 QA0 QA1
Read Continue Continue Continue Continue
Integrated Silicon Solution, Inc. — 1-800-379-4774
ADVANCE INFORMATION Rev. 00A
06/19/01
9
IS61NSCS25672
IS61NSCS51236ISSI
Echo Clock
ΣΣ
ΣRAMs feature Echo Clocks, CQ1,CQ2, CQ1, and CQ2
ΣΣ
that track the performance of the output drivers. The Echo
Clocks are delayed copies of the main RAM clock, CLK.
Echo Clocks are designed to track changes in output
driver delays due to variance in die temperature and
supply voltage. The Echo Clocks are designed to fire with
the rest of the data output drivers. Sigma RAMs provide
both in-phase, or true, Echo Clock outputs (CQ1 and
CQ2) and inverted Echo Clock outputs (CQ1 and CQ2).
It should be noted that deselection of the RAM via E2 and
E3 also deselects the Echo Clock output drivers. The
deselection of Echo Clock drivers is always pipelined to
Echo Clock Control in Two Banks of Sigma Pipelined SRAMs
the same degree as output data. Deselection of the RAM
via E1 does not deactivate the Echo Clocks.
In some applications it may be appropriate to pause
between banks; to deselect both RAMs with E1 before
resuming read operations. An E1 deselect at a bank
switch will allow at least one clock to be issued from the
new bank before the first read cycle in the bank. Although
the following drawing illustrates a E1 read pause upon
switching from Bank 1 to Bank 2, a write to Bank 2 would
have the same effect, causing the RAM in Bank 2 to issue
at least one clock before it is needed.
®
CLK
Address
E1
E2 Bank 1
E2 Bank 2
DQ Bank 1
DQ Bank 2
CQ Bank 1
CQ Bank 2
A B C D E F
QA QC
QB QD
CQ1+ CQ2
Read Read Read Read Read Read
Note:
E1 does not deselect the Echo Clock Outputs. Echo Clock outputs are synchronously deselected by E2 or E3 being sampled false.
10
Integrated Silicon Solution, Inc. — 1-800-379-4774
ADVANCE INFORMATION Rev. 00A
06/19/01
Loading...
+ 22 hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.