WCSS0418V1P Datasheet, PDF (1/17 Pages) Weida Semiconductor, Inc. – 256K x 18 Synchronous-Pipelined Cache RAM

Y7C1327
WCSS0418V1P
256K x 18 Synchronous-Pipelined Cache RAM
Features
• Supports 100-MHz bus for Pentium and PowerPC™
operations with zero wait states
• Fully registered inputs and outputs for pipelined
operation
• 256K by 18 common I/O architecture
• 3.3V core power supply
• 2.5V / 3.3V I/O operation
• Fast clock-to-output times
— 3.5 ns (for 166-MHz device)
— 4.0 ns (for 133-MHz device)
— 5.5 ns (for 100-MHz device)
• User-selectable burst counter supporting Intel
Pentium interleaved or linear burst sequences
• Separate processor and controller address strobes
• Synchronous self-timed writes
• Asynchronous Output Enable
• JEDEC-standard 100 TQFP pinout
• “ZZ” Sleep Mode option and Stop Clock option
Functional Description
The WCSS0418V1P is a 3.3V, 256K by 18 synchronous-pipe-
lined cache SRAM designed to support zero wait state sec-
ondary cache with minimal glue logic.
The WCSS0418V1P I/O pins can operate at either the 2.5V or
the 3.3V level. The I/O pins are 3.3V tolerant when V_D-
DQ^=2.5V.
All synchronous inputs pass through input registers controlled
by the rising edge of the clock. All data outputs pass through
output registers controlled by the rising edge of the clock. Max-
imum access delay from the clock rise is 3.5 ns (166-MHz
device).
The WCSS0418V1P supports either the interleaved burst se-
quence used by the Intel Pentium processor or a linear burst
sequence used by processors such as the PowerPC. The
burst sequence is selected through the MODE pin. Accesses
can be initiated by asserting either the Processor Address
Strobe (ADSP) or the Controller Address Strobe (ADSC) at
clock rise. Address advancement through the burst sequence
is controlled by the ADV input. A 2-bit on-chip wraparound
burst counter captures the first address in a burst sequence
and automatically increments the address for the rest of the
burst access.
Byte write operations are qualified with the four Byte Write
Select (BW_[1:0]) inputs. A Global Write Enable (GW) overrides
all byte write inputs and writes data to all four bytes. All writes
are conducted with on-chip synchronous self-timed write cir-
cuitry.
Three synchronous Chip Selects (CE₁, CE₂, CE₃) and an
asynchronous Output Enable (OE) provide for easy bank se-
lection and output three-state control. In order to provide prop-
er data during depth expansion, OE is masked during the first
clock of a read cycle when emerging from a deselected state.
Logic Block Diagram
CLK
ADV
ADSC
ADSP
A_[17:0]
18
GW
BWE
BW ₁
BW₀
MODE
(A_[1;0]) 2
BURST ^Q0
CE COUNTER
CLR
Q1
Q
ADDRESS
16
CE
D
REGISTER
16
D DQ[15:8], DP[1] Q
BYTEWRITE
REGISTERS
D DQ[7:0], DP[0] Q
BYTEWRITE
REGISTERS
18
256KX18
MEMORY
ARRAY
CE1
CE2
CE3
OE
ZZ
D
CE
ENABLE CE
REGISTER
Q
D _{ENABLE DELAY} Q
REGISTER
SLEEP
CONTROL
Intel and Pentium are registered trademarks of Intel Corporation.
PowerPC is a trademark of IBM Corporation.
Document #: 38-05247
18
18
OUTPUT
REGISTERS
CLK
INPUT
REGISTERS
CLK
DQ_[15:0]
DP_[1:0]
Revised February 6, 2001