a
Preliminary Technical Data
SHARC and the SHARC logo are registered trademarks of Analog Devices, Inc.
SHARC
®
Processor
Data Sheet Addendum
ADSP-21367/ADSP-21368/ADSP-21369
Rev. PrA
Information furnished by Analog Devices is believed to be accurate and reliable.
However, no responsibility is assumed by Analog Devices for its use, nor for any
infringements of patents or other rights of third parties that may result from its use.
Specifications subject to change without notice. No license is granted by implication
or otherwise under any patent or patent rights of Analog Devices. Trademarks and
registered trademarks are the property of their respective companies.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106 U.S.A.
Tel: 781.329.4700 www.analog.com
Fax: 781.461.3113 ©2007 Analog Devices, Inc. All rights reserved.
SUMMARY
High performance 32-bit/40-bit floating point processor
optimized for high performance audio processing
Single-instruction, multiple-data (SIMD) computational
architecture
On-chip memory—2M bit of on-chip SRAM and 6M bit of on-
chip mask programmable ROM
400 MHz maximum core clock frequency
1.3 V core V
DD
/3.3 V I/O
Code compatible with all other members of the SHARC
family
The ADSP-21367/ADSP-21368/ADSP-21369 are available
with a 400 MHz core instruction rate with unique audio-
centric peripherals such as the digital audio interface,
S/PDIF transceiver, serial ports, 8-channel asynchronous
sample rate converter, precision clock generators, and
more. For complete ordering information, see Ordering
Guide on Page 11.
At 400 MHz (2.5 ns) core instruction rate, the processors per-
form 2.4 GFLOPS/800 MMACS
Transfers between memory and core at a sustained
6.4G bytes/s bandwidth at 400 MHz core instruction rate
GENERAL DESCRIPTION
This data sheet addendum introduces the 400 MHz ADSP-21367/ADSP-21368/ADSP-21369 SHARC processors. This addendum pro-
vides the frequency benchmark, as well as ac and dc specifications that differ from the 333 MHz ADSP-21367/ADSP-21368/ADSP-21369
SHARC processors. All other specifications and timing data as well as package information for these devices can be found in the
ADSP-21367/ADSP-21368/ADSP-21369 SHARC Processor Data Sheet, Rev A. The products listed in the addendum are engineering
grade and have not been fully characterized. For complete ordering information, see the Ordering Guide on Page 11.
Rev. PrA | Page 2 of 12 | March 2007
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet Addendum Preliminary Technical Data
TABLE OF CONTENTS
Summary ................................................................1
General Description ..................................................1
Specifications ...........................................................3
Operating Conditions .............................................3
Electrical Characteristics ..........................................3
Timing Specifications .............................................4
Output Drive Currents .......................................... 10
Capacitive Loading ...............................................10
Ordering Guide ......................................................11
PERFORMANCE BENCHMARKS
The processors use two computational units to deliver a signifi-
cant performance increase over the previous SHARC processors
on a range of DSP algorithms. Fabricated in a state-of-the-art,
high speed, CMOS process, the ADSP-21367/ADSP-
21368/ADSP-21369 processors achieve an instruction cycle
time of up to 2.5 ns at 400 MHz. With its SIMD computational
hardware, the processors can perform 2.4 GFLOPS running at
400 MHz.
Table 1 shows performance benchmarks for these devices.
POWER SUPPLIES
The processors have separate power supply connections for the
internal (V
DDINT
), external (V
DDEXT
), and analog (A
VDD
/A
VSS
)
power supplies. The internal and analog supplies must meet the
1.3 V requirement for the 400 MHz device. The external supply
must meet the 3.3 V requirement. All external supply pins must
be connected to the same power supply.
Table 1. Processor Benchmarks (at 400 MHz)
Benchmark Algorithm Speed (at 400 MHz)
1024 Point Complex FFT (Radix 4, with
reversal)
23.2 μs
FIR Filter (per tap)
1
1
Assumes two files in multichannel SIMD mode.
1.25 ns
IIR Filter (per biquad)
1
5.0 ns
Matrix Multiply (pipelined)
[3×3] × [3×1]
[4×4] × [4×1]
11.25 ns
20.0 ns
Divide (y/×) 8.75 ns
Inverse Square Root 13.5 ns
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet AddendumPreliminary Technical Data
Rev. PrA | Page 3 of 12 | March 2007
SPECIFICATIONS
OPERATING CONDITIONS
ELECTRICAL CHARACTERISTICS
Parameter
1
1
Specifications subject to change without notice.
Description Min Max Unit
V
DDINT
Internal (Core) Supply Voltage 1.25 1.35 V
A
VDD
Analog (PLL) Supply Voltage 1.25 1.35 V
V
DDEXT
External (I/O) Supply Voltage 3.13 3.47 V
V
IH
2
2
Applies to input and bidirectional pins: DATAx, ACK, RPBA, BRx, IDx, FLAGx, DAI_Px, DPI_Px, BOOT_CFGx, CLK_CFGx, RESET, TCK, TMS, TDI, TRST.
High Level Input Voltage @ V
DDEXT
= max 2.0 V
DDEXT
+ 0.5 V
V
IL
2
Low Level Input Voltage @ V
DDEXT
= min –0.5 +0.8 V
V
IH
_
CLKIN
3
3
Applies to input pin CLKIN.
High Level Input Voltage @ V
DDEXT
= max 1.74 V
DDEXT
+ 0.5 V
V
IL
_
CLKIN
3
Low Level Input Voltage @ V
DDEXT
= min –0.5 +1.1 V
T
J
Junction Temperature, 256-Ball SBGA @ T
AMBIENT
0°C to +70°C 0 +105 °C
Parameter
1
Description Test Conditions Min Typ Max Unit
V
OH
2
High Level Output Voltage @ V
DDEXT
= min, I
OH
= –1.0 mA
3
2.4 V
V
OL
2
Low Level Output Voltage @ V
DDEXT
= min, I
OL
= 1.0 mA
3
0.4 V
I
IH
4, 5
High Level Input Current @ V
DDEXT
= max, V
IN
= V
DDEXT
max 10 μA
I
IL
4,
6, 7
Low Level Input Current @ V
DDEXT
= max, V
IN
= 0 V 10 μA
I
IHPD
6
High Level Input Current Pull-down @ V
DDEXT
= max, V
IN
= 0 V 250 μA
I
ILPU
5
Low Level Input Current Pull-up @ V
DDEXT
= max, V
IN
= 0 V 200 μA
I
OZH
8, 9
Three-State Leakage Current @ V
DDEXT
= max, V
IN
= V
DDEXT
max 10 μA
I
OZL
8,
10
Three-State Leakage Current @ V
DDEXT
= max, V
IN
= 0 V 10 μA
I
OZLPU
9
Three-State Leakage Current Pull-up @ V
DDEXT
= max, V
IN
= 0 V 200 μA
I
DD
-
INTYP
11
Supply Current (Internal) t
CCLK
= 2.5 ns, V
DDINT
= 1.3 V, 25°C 1.4 A
AI
DD
12
Supply Current (Analog) A
VDD
= max 10 mA
C
IN
13, 14
Input Capacitance f
IN
= 1 MHz, T
CASE
= 25°C, V
IN
= 1.3 V 4.7 pF
1
Specifications subject to change without notice.
2
Applies to output and bidirectional pins: ADDRx, DATAx, RD, WR, MSx, BRx, FLAGx, DAI_Px, DPI_Px, SDRAS, SDCAS, SDWE, SDCKE, SDA10, SDCLKx, EMU, TDO,
CLKOUT.
3
See Output Drive Currents on Page 10 for typical drive current capabilities.
4
Applies to input pins without internal pull-ups: BOOT_CFGx, CLK_CFGx, CLKIN, RESET, TCK.
5
Applies to input pins with internal pull-ups: ACK, RPBA, TMS, TDI, TRST.
6
Applies to input pins with internal pull-downs: IDx.
7
Applies to input pins with internal pull-ups disabled: ACK, RPBA.
8
Applies to three-statable pins without internal pull-ups: FLAGx, SDCLKx, TDO.
9
Applies to three-statable pins with internal pull-ups: ADDRx, DATAx, RD, WR, MSx, BRx, DAI_Px, DPI_Px, SDRAS, SDCAS, SDWE, SDCKE, SDA10, EMU.
10
Applies to three-statable pins with internal pull-ups disabled: ADDRx, DATAx, RD, WR, MSx, BRx, DAI_Px, DPI_Px, SDRAS, SDCAS, SDWE, SDCKE, SDA10
11
See Engineer-to-Engineer Note 299 for further information.
12
Characterized, but not tested.
13
Applies to all signal pins.
14
Guaranteed, but not tested.
Rev. PrA | Page 4 of 12 | March 2007
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet Addendum Preliminary Technical Data
TIMING SPECIFICATIONS
The processor’s internal clock (a multiple of CLKIN) provides
the clock signal for timing internal memory, processor core, and
serial ports. During reset, program the ratio between the proces-
sor’s internal clock frequency and external (CLKIN) clock
frequency with the CLK_CFG1–0 pins. To determine switching
frequencies for the serial ports, divide down the internal clock,
using the programmable divider control of each port (DIVx for
the serial ports).
The processor’s internal clock switches at higher frequencies
than the system input clock (CLKIN). To generate the internal
clock, the processor uses an internal phase-locked loop (PLL).
This PLL-based clocking minimizes the skew between the sys-
tem clock (CLKIN) signal and the processor’s internal clock.
Use the exact timing information given. Do not attempt to
derive parameters from the addition or subtraction of others.
While addition or subtraction would yield meaningful results
for an individual device, the values given in this data sheet
reflect statistical variations and worst cases. Consequently, it is
not meaningful to add parameters to derive longer times.
Switching Characteristics specify how the processor changes its
signals. Circuitry external to the processor must be designed for
compatibility with these signal characteristics. Switching char-
acteristics describe what the processor will do in a given
circumstance. Use switching characteristics to ensure that any
timing requirement of a device connected to the processor (such
as memory) is satisfied.
Timing Requirements apply to signals that are controlled by cir-
cuitry external to the processor, such as the data input for a read
operation. Timing requirements guarantee that the processor
operates correctly with other devices.
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet AddendumPreliminary Technical Data
Rev. PrA | Page 5 of 12 | March 2007
Power-Up Sequencing
The timing requirements for processor startup are given in
Table 2.
Table 2. Power-Up Sequencing Timing Requirements (Processor Startup)
Parameter Min Max Unit
Timing Requirements
t
RSTVDD
RESET Low Before V
DDINT
/V
DDEXT
On 0 ns
t
IVDDEVDD
V
DDINT
On Before V
DDEXT
–50 +200 ms
t
CLKVDD
1
CLKIN Valid After V
DDINT
/V
DDEXT
Valid 0 +200 ms
t
CLKRST
CLKIN Valid Before RESET Deasserted 10
2
μs
t
PLLRST
PLL Control Setup Before RESET Deasserted 20 μs
Switching Characteristic
t
CORERST
Core Reset Deasserted After RESET Deasserted 4096t
CK
+ 2 t
CCLK
3, 4
1
Valid V
DDINT
/V
DDEXT
assumes that the supplies are fully ramped to their 1.3 volt rails and 3.3 volt rails. Voltage ramp rates can vary from microseconds to hundreds of milliseconds
depending on the design of the power supply subsystem.
2
Assumes a stable CLKIN signal, after meeting worst-case startup timing of crystal oscillators. Refer to your crystal oscillator manufacturer’s data sheet for start-up time. Assume
a 25 ms maximum oscillator start-up time if using the XTAL pin and internal oscillator circuit in conjunction with an external crystal.
3
Applies after the power-up sequence is complete. Subsequent resets require RESET to be held low a minimum of four CLKIN cycles in order to properly initialize and propagate
default states at all I/O pins.
4
The 4096 cycle count depends on t
srst
specification. If setup time is not met, 1 additional CLKIN cycle may be added to the core reset time, resulting in 4097 cycles maximum.
Figure 1. Power-Up Sequencing
CLKIN
RESET
tRSTVDD
RESETOUT
VDDEXT
VDDINT
tPLLRST
tCLKRST
tCLKVDD
tIVDDEVDD
CLK_CFG1-0
tCORERST
Rev. PrA | Page 6 of 12 | March 2007
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet Adddendum Preliminary Technical Data
Clock Input Clock Signals
The processors can use an external clock or a crystal. See the
CLKIN pin description. Programs can configure the processor
to use its internal clock generator by connecting the necessary
components to CLKIN and XTAL. Figure 3 shows the compo-
nent connections used for a crystal operating in fundamental
mode. Note that the clock rate is achieved using a 25 MHz crys-
tal and a PLL multiplier ratio 16:1 (CCLK:CLKIN achieves a
clock speed of 400 MHz). To achieve the full core clock rate,
programs need to configure the multiplier bits in the PMCTL
register.
Table 3. Clock Input
Parameter
400 MHz
UnitMin Max
Timing Requirements
t
CK
CLKIN Period 18
1
1
Applies only for CLK_CFG1–0 = 00 and default values for PLL control bits in
PMCTL.
100
2
2
Applies only for CLK_CFG1–0 = 10 and default values for PLL control bits in
PMCTL.
ns
t
CKL
CLKIN Width Low 8
1
45
2
ns
t
CKH
CLKIN Width High 8
1
45
2
ns
t
CKRF
CLKIN Rise/Fall (0.4 V to
2.0 V)
3ns
t
CCLK
3
3
Any changes to PLL control bits in the PMCTL register must meet core clock
timing specification t
CCLK
.
CCLK Period 2.5
1
10 ns
t
CKJ
4, 5
4
Actual input jitter should be combined with ac specifications for accurate timing
analysis.
5
Jitter specification is maximum peak-to-peak time interval error (TIE) jitter.
CLKIN Jitter Tolerance –250 +250 ps
Figure 2. Clock Input
CLKIN
tCK
tCKH tCKL Figure 3. 400 MHz Operation (Fundamental Mode Crystal)
C1
22pF Y1
R1
1M(TYPICAL) XTAL
CLKIN
C2
22pF
24.576MHz
R2
47(TYPICAL)
ADSP-2136x
R2 SHOULD BE CHOSEN TO LIMIT CRYSTAL DRIVE POWER.
REFER TO CRYSTAL MANUFACTURER’S SPECIFICATIONS
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet AddendumPreliminary Technical Data
Rev. PrA | Page 7 of 12 | March 2007
SDRAM Interface Timing (133 MHz SDCLK)
The 133 MHz access speed is for a single processor. When mul-
tiple ADSP-21368 processors are connected in a shared memory
system, the access speed is 100 MHz.
Table 4. SDRAM Interface Timing
1
Parameter Min Max Unit
Timing Requirements
t
SSDAT
DATA Setup Before SDCLK 0.78 ns
t
HSDAT
DATA Hold After SDCLK 1.23 ns
Switching Characteristics
t
SDCLK
SDCLK Period 7.5 ns
t
SDCLKH
SDCLK Width High 3.65 ns
t
SDCLKL
SDCLK Width Low 3.65 ns
t
DCAD
Command, ADDR, Data Delay After SDCLK
2
4.8 ns
t
HCAD
Command, ADDR, Data Hold After SDCLK
2
1.2 ns
t
DSDAT
Data Disable After SDCLK 5.3 ns
t
ENSDAT
Data Enable After SDCLK 1.2 ns
1
For F
CCLK
= 400 MHz (SDCLK ratio = 1:2.5).
2
Command pins include: SDCAS, SDRAS, SDWE, MSx, SDA10, SDCKE.
Figure 4. SDRAM Interface Timing
tHCAD
tHCAD
tDSDAT
tSSDAT
tDCAD
tENSDAT
tHSDAT
tSDCLKL
tSDCLKH
tSDCLK
SDCLK
DATA (IN)
DATA(OUT)
CMND ADDR
(OUT)
tDCAD
Rev. PrA | Page 8 of 12 | March 2007
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet Adddendum Preliminary Technical Data
SDRAM Interface Enable/Disable Timing (133 MHz SDCLK)
Pin to Pin Direct Routing (DAI and DPI)
For direct pin connections only (for example, DAI_PB01_I to
DAI_PB02_O).
Table 5. SDRAM Interface Enable/Disable Timing
1
1
For F
CCLK
= 400 MHz (SDCLK ratio = 1:2.5).
Parameter Min Max Unit
Switching Characteristics
t
DSDC
Command Disable After CLKIN Rise 2 × t
PCLK
+ 1 ns
t
ENSDC
Command Enable After CLKIN Rise 4.0 ns
t
DSDCC
SDCLK Disable After CLKIN Rise 8.5 ns
t
ENSDCC
SDCLK Enable After CLKIN Rise 3.8 ns
t
DSDCA
Address Disable After CLKIN Rise 9.2 ns
t
ENSDCA
Address Enable After CLKIN Rise 2 × t
PCLK
– 4 4 × t
PCLK
ns
Figure 5. SDRAM Interface Enable/Disable Timing
CLKIN
COMMAND
SDCLK
ADDR
tDSDC
tDSDCC
tDSDCA
tENSDC
tENSDCA
COMMAND
SDCLK
ADDR
tENSDCC
Table 6. DAI Pin to Pin Routing
Parameter Min Max Unit
Timing Requirement
t
DPIO
Delay DAI Pin Input Valid to DAI Output Valid 1.5 12 ns
Figure 6. DAI Pin to Pin Direct Routing
DAI_Pn
DPI_Pn
tDPIO
DAI_pm
DPI_Pm
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet AddendumPreliminary Technical Data
Rev. PrA | Page 9 of 12 | March 2007
Memory Read – Bus Master to Memory Read
Use these specifications for asynchronous interfacing to memo-
ries. These specifications apply when the processors are the bus
master accessing external memory space in asynchronous access
mode. Note that timing for ACK, DATA, RD, WR, and strobe
timing parameters only apply to asynchronous access mode.
Table 7. Memory Read—Bus Master
Parameter Min Max Unit
Timing Requirements
t
DAD
Address, Selects Delay to Data Valid
1, 2
W+t
SDCLK
–5.12 ns
t
DRLD
RD Low to Data Valid
1
W– 2.9 ns
t
SDS
Data Setup to RD High 2.2 ns
t
HDRH
Data Hold from RD High
3,
4
0ns
t
DAAK
ACK Delay from Address, Selects
2, 5
t
SDCLK
9.5+ W ns
t
DSAK
ACK Delay from RD Low
4
W – 7.0 ns
Switching Characteristics
t
DRHA
Address Selects Hold After RD High RH + 0.18 ns
t
DARL
Address Selects to RD Low
2
t
SDCLK
–3.3 ns
t
RW
RD Pulse Width W – 1.2 ns
t
RWR
RD High to WR, RD Low HI +t
SDCLK
– 0.8 ns
W = (number of wait states specified in AMICTLx register) × t
SDCLK
.
HI =RHC + IC (RHC = number of read hold cycles specified in AMICTLx register) × t
SDCLK
IC = (number of idle cycles specified in AMICTLx register) × t
SDCLK
.
H = (number of hold cycles specified in AMICTLx register) × t
SDCLK
.
1
Data delay/setup: system must meet t
DAD
, t
DRLD
, or t
SDS.
2
The falling edge of MSx is referenced.
3
Note that timing for ACK, DATA, RD, WR, and strobe timing parameters only apply to asynchronous access mode.
4
Data hold: User must meet t
HDA
or t
HDRH
in asynchronous access mode.
5
ACK Delay/Setup: User must meet t
DAAK
, or t
DSAK
, for deassertion of ACK (low). For asynchronous assertion of ACK (high) user must meet t
DAAK
or t
DSAK
.
Figure 7. Memory Read—Bus Master
ACK
DATA
tDARL tRW
tDAD
tDAAK
tHDRH
tRWR
tDRLD
tDRHA
tDSAK
tSDS
ADDRESS
MSx
RD
WR
Rev. PrA | Page 10 of 12 | March 2007
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet Adddendum Preliminary Technical Data
OUTPUT DRIVE CURRENTS
Figure 8 shows typical I-V characteristics for the output drivers
of the ADSP-21367/ADSP-21368/ADSP-21369. The curves rep-
resent the current drive capability of the output drivers as a
function of output voltage.
CAPACITIVE LOADING
Output delays and holds are based on standard capacitive loads:
30 pF on all pins. Figure 11 shows graphically how output delays
and holds vary with load capacitance. The graphs of Figure 9,
Figure 10, and Figure 11 may not be linear outside the ranges
shown for Typical Output Delay vs. Load Capacitance and
Typical Output Rise Time (20% to 80%, V = Min) vs. Load
Capacitance.
Figure 8. Typical Drive at Junction temperature
Figure 9. Typical Output Rise/Fall Time (20% to 80%,
V
DDEXT
= Max)
CLKOUT (CLKOUT DRIVER), V
DDEXT
(MAX)
= 3.65V, TEMPERATURE = 85°C
LOAD CAPACITANCE (pF)
RISE AND FALL TIME ns (10% to 90%)
10
9
8
7
6
5
4
3
2
1
0
0 50 100 150 200 250
TBD
CLKOUT (CLKOUT DRIVER), V
DDEXT
(MAX)
= 3.65V, TEMPERATURE = 85°C
LOAD CAPACITANCE (pF)
RISE AND FALL TIME ns (10% to 90%)
10
9
8
7
6
5
4
3
2
1
0
0 50 100 150 200 250
TBD
Figure 10. Typical Output Rise/Fall Time (20% to 80%,
V
DDEXT
= Min)
Figure 11. Typical Output Delay or Hold vs. Load Capacitance
(at Junction Temperature)
CLKOUT (CLKOUT DRIVER), V
DDEXT
(MAX)
= 3.65V, TEMPERATURE = 85°C
LOAD CAPACITANCE (pF)
RISE AND FALL TIME ns (10% to 90%)
10
9
8
7
6
5
4
3
2
1
0
0 50 100 150 200 250
TBD
CLKOUT (CLKOUT DRIVER), V
DDEXT
(MAX)
= 3.65V, TEMPERATURE = 85°C
LOAD CAPACITANCE (pF)
RISE AND FALL TIME ns (10% to 90%)
10
9
8
7
6
5
4
3
2
1
0
0 50 100 150 200 250
TBD
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet AddendumPreliminary Technical Data
Rev. PrA | Page 11 of 12 | March 2007
ORDERING GUIDE
Part Number
Temperature
Range
Instruction
Rate
On-Chip
SRAM ROM
Operating
Voltage
Internal/External
Package
Description
Package
Option
ADSP-21367KBP-3A
1
1
Available with a wide variety of audio algorithm combinations sold as part of a chipset and bundled with necessary software. For a complete list, visit our website at
www.analog.com/SHARC.
C to +70°C 400 MHz 2M bit 6M bit 1.3 V/3.3 V 256-Ball SBGA BP-256
ADSP-21367KBPZ-3A
2
2
Z = RoHS Compliant Part.
C to +70°C 400 MHz 2M bit 6M bit 1.3 V/3.3 V 256-Ball SBGA BP-256
ADSP-21368KBP-3A C to +70°C 400 MHz 2M bit 6M bit 1.3 V/3.3 V 256-Ball SBGA BP-256
ADSP-21368KBPZ-3A
2
C to +70°C 400 MHz 2M bit 6M bit 1.3 V/3.3 V 256-Ball SBGA BP-256
ADSP-21369KBP-3A C to +70°C 400 MHz 2M bit 6M bit 1.3 V/3.3 V 256-Ball SBGA BP-256
ADSP-21369KBPZ-3A
2
C to +70°C 400 MHz 2M bit 6M bit 1.3 V/3.3 V 256-Ball SBGA BP-256
Rev. PrA | Page 12 of 12 | March 2007
ADSP-21367/ADSP-21368/ADSP-21369 Data Sheet Addendum Preliminary Technical Data
©2007 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
PR06770-0-4/07(PrA)