| Company
|
This DSP family is: (1)
|
It is available as:
|
Fixed (2)
|
Word size
|
Accumulator
|
Floating (3)
|
Word size
|
Accumulator
|
Architecture
|
CISC/RISC?
|
No. of instructions
|
Implementation type
|
Number of separate functional units
|
No. of Gen't purpose calculation units
|
Program Memory
|
Data Memory
|
Min. Single -Tap FIR Filter Cycle Time
|
Upward Code Compatible w/ What Other Part Family
|
C Complier Available
|
Assembly Language Optimizations Tools Available
|
Architectural Support for C Stack Frames
|
JTAG Support for Emulation
|
Full JTAG Test Support
|
Low Cost Emulation Available
|
Full Featured Emulator Available
|
Peripherals Available
|
Mixed-Signal Versions Available
|
Part Numbers in the Family
|
URL for Datasheets
|
URL for Application Notes
|
| 3DSP
|
MCU-based
|
Core
|
X
|
32
|
48
|
|
|
|
Mod. Harvard
|
RISC
|
50
|
SuperScalar SIMD
|
four 16-bit multiplier, four 32-bit accumulates, eight 8-bit multiplier, four loads, four memory address calculations, eight 8-bit adds, four 16-bit adds, two 32-bit adds, one 32-bit shift
|
75
|
Up to four giga 32-bit words, user can implement with ROM, RAM, FLASH, cache capability is provided in user selectable blocks.
|
Up to 4 GB user can implement with ROM, RAM, FLASH.
|
6.4 ns. Note: this uses only one eighth or the processing capability of a single
cycle. For example, if eight such operations were required each would average only 0.8 ns.
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
No
|
SP-5
|
http://www.3dsp.com/lit/lit-gateway.html
|
http://www.3dsp.com/lit/lit-gateway.html
|
| 3DSP
|
MCU-based
|
Core
|
X
|
32
|
48
|
|
|
|
Mod. Harvard
|
RISC
|
50
|
Scalar SIMD
|
two 16-bit multiplier, two 32-bit accumulates, four 8-bit multiplier, two loads, two memory address calculations, four 8-bit adds, two 16-bit adds, one 32-bit adds, one 32-bit shift
|
75
|
Up to four giga 32-bit words, user can implement with ROM, RAM, FLASH, cache capability is provided in user selectable blocks.
|
Up to 4 GB user can implement with ROM, RAM, FLASH.
|
6.4 ns. Note: this uses only one eighth or the processing capability of a single cycle. For example, if four
such operations were required each would average only 1.6 ns.
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
No
|
SP-3
|
http://www.3dsp.com/lit/lit-gateway.html
|
http://www.3dsp.com/lit/lit-gateway.html
|
| AMSI
|
|
Core
|
X
|
8 to 32
|
Up to 64
|
|
|
|
Mod. Harvard
|
RISC
|
20
|
SuperScalar
|
one multiply-accumulate, one ALU, two address calculation
|
One (ALU)
|
32 bit words; Internal, ROM
Memory
|
8- to 32-bit words; internal, RAM memory
|
10 ns for 16-bit implementation
|
No
|
No
|
No
|
No
|
No
|
No
|
No
|
No
|
Yes
|
Yes
|
GEP02, GEP03, GEP04, GEP05, GEP10
|
|
|
| Analog Devices
|
MCU-based
|
Standard
|
X
|
16 bit
|
40
|
|
|
|
Mod. Harvard
|
RISC
|
70
|
Scalar
|
one ALU/MAC/shifter, two memory reads, two address calculation, one instruction fetch
|
ALU, MAC, Shifter, 2x Data Address Generators.
|
Up to 48K x 24, Internal,
Shared, RAM
|
Up to 56K x 16, internal
|
13 ns
|
Code Compatible w/ all 21xx family members
|
Yes
|
No
|
Yes
|
No
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
ADSP-2101, ADSP 2103, ADSP 2104, ADSP 2105, ADSP 2109, ADSP 2111, ADSP 2115, ADSP 2161, ADSP-2162, ADSP-2163,
ADSP-2164, ADSP-2171, ADSP-2173, ADSP-2181, ADSP-2183, ADSP-2184, ADSP-2184L, ADSP-2185, ADSP-2185L, ADSP-2185M, ADSP-2186, ADSP-2186L, ADSP-2186
|
http://www.analog.com/dsp
|
http://www.analog.com/dsp
|
| Analog Devices
|
MCU-based
|
Standard
|
X
|
32
|
40
|
X
|
32 & 40
|
40
|
Mod. Harvard
|
CISC
|
34
|
Scalar
|
ADSP-2106x: one ALU, one multiplier, two address calculations with circular buffing, one cache, two DMA address calculation, three transfers from/to memory, one timer operations, four serial ports accesses, six link port access, four external port DMA ADSP-2116x: two ALU, two Multiplier, four address calculations with circular buffing, one cache, two DMA address calculation, three transfers
from/to memory, one timer operation, four serial ports accesses, six link port access, four external port DMA
|
ADSP-2106x: one ALU, one multiplier, two address calculations with circular buffing, two DMA address calculation. ADSP-2116x: two ALU, two multiplier, four address calculations with circular buffing, two DMA address calculation.
|
Number of memory locations,
external: 4 gigawords, Internal: 128 32-bit words, word width (bits) :32-bit, program memory: 512 kbit to 4 Mbits, Shared, RAM
|
|
15 ns
|
2116x is upward code compatible with 2106x
|
Yes
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
No
|
ADSP-21020, ADSP-21060, ADSP-21062, ADSP-21061,ADSP-21061L, ADSP-21062L, ADSP-21065L, ADSP-21160M
|
http://www.analog.com/dsp
|
http://www.analog.com/dsp
|
|
ARM
|
MPU-based
|
Core
|
X
|
|
32/64
|
|
|
|
Harvard
|
RISC
|
Similar to ARM instruc-tion set
architecture. VSTE
|
Scalar
|
Multiply and accumulate in parallel
|
1
|
Internal; ROM, RAM, EPROM, EEPROM
|
Internal, external, shared; ROM,
RAM, EPROM, EEPROM
|
5.7 ns
|
ARM architecture V5TE. This is binary backwards compatible with all previous ARM architectures.
|
Yes
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
ARM9E, ARM946E-S, ARM966E-S
|
http://www.arm.com/Pro+Peripherals/Cores/ARM9E
|
http://www.arm.com/Documentation/AppNotes/
|
|
BOPS
|
MCU-based
|
Core
|
X
|
8,16,32
|
40,64
|
X
|
32
|
32
|
ManArray
|
RISC
|
32-bit
|
iVLIW - Indirect VLIW in a manifold array
|
1x1 5-32 bit VLIW instructions: 16-bit fixed - two ALU, two multiplier,
four address, four load, four store, four DMA ; 32-bit float - one ALU, one multiplier, two address, one load, one store, two DMA 2x2; 16-bit Fixed - eight ALU, eight multiplier, eight address, eight load, eight store, eight DMA, eight intercore; 32-bit float - four ALU, four multiplier, eight address, four load, four store, eight DMA, four intercore 4x4: 16-bit fixed - 32 ALU, 32 Multiplier, 32 Address, 32 load, 32 store, 32 DMA, 32 Intercore; 32-bit float - 16 ALU, 16 multiplier, 16 address, 16 load, 16
store, 16 DMA, 16 intercore
|
|
User Configurable VLIW/ PE and SP 160,32 Word Width. Internal; RAM
|
User configurable 32/PE word width
|
200 MHz, 2x2 Array, 16-bit data =
0.625 ns/tap
|
BOPS ManArray code compatible across 1x1, 1x2, 2x2, 2x4, 4x4,
4x4x2, 4x4x4
|
Yes
|
Yes
|
No
|
Yes
|
Yes
|
No
|
Yes
|
Yes
|
Yes
|
2010FX, 2010FL, 2010XL, 2010DL
2020FX, 2020FL, 2020XL, 2020DL
2040FX, 2040FL, 2040XL, 2040DL
|
http://www.bopsnet.com/publications.shtml
|
http://www.bopsnet.com/publications.shtml
|
| Catalina
|
n/a
|
Standard/Core
|
X
|
32
|
36
|
|
|
|
Dataflow
|
n/a
|
18
|
Vector Processor
|
25
|
n/a
|
n/a
|
n/a
|
0.02 ns per tap (complex FIR implemented in frequency domain)
|
Pathfinder-1 is the first product in the family
|
n/a
|
n/a
|
n/a
|
No
|
No
|
Yes
|
Yes
|
Yes
|
No
|
Path1-80
|
http://www.catalinaresearch.com
|
http://www.catalinaresearch.com
|
| Hitachi
|
MCU-based
|
Standard
|
X
|
16 bit
|
16 bit
|
|
|
|
Mod. Harvard
|
RISC
|
154
|
Scalar
|
one ALU, one multiplier, two address
|
1
|
Number of memory locations: 48 K; Word width (bits) 8; internal, ROM
|
Memory locations: 4K word width (bits):16; internal, RAM
|
1
|
SH-1,SH-2
|
Yes
|
Yes
|
No
|
Yes
|
No
|
Yes
|
Yes
|
Yes: 4 DMA,5 Serial ports, 3 16-timers, 1 Watch Dog timer,
|
No
|
HD6437410F60
|
http://semiconductor.hitachi.com/superh
|
http://semiconductor.hitachi.com/superh
|
| Hitachi
|
MCU-based
|
Standard
|
X
|
16
|
16
|
|
|
|
Mod. Harvard
|
RISC
|
154
|
Scalar
|
one ALU, one multiplier, two address
|
1
|
Memory locations: 256K word width: 8; internal, Flash
|
Memory locations: 4K word width (bits):16; internal, RAM
|
1
|
SH-1,SH-2
|
Yes
|
Yes
|
No
|
No
|
No
|
No
|
Yes
|
Yes:4 DMA,3 Serial Ports 9 16-Timers, 1 Watch Dog Timer
|
Yes
|
HD64F7065F60
|
http://semiconductor.hitachi.com/superh
|
http://semiconductor.hitachi.com/superh
|
| Hitachi
|
MPU-based
|
Standard
|
X
|
16
|
16
|
|
|
|
Mod. Harvard
|
RISC
|
154
|
Scalar
|
one ALU, one multiplier, two address
|
1
|
Memory locations: 256M word width: 8; external, SDRAM
|
Memory locations: 8K word width (bits):16; internal, RAM
|
1
|
SH-1,SH-2
|
Yes
|
Yes
|
No
|
Yes
|
No
|
Yes
|
Yes
|
Yes:2 DMA,6 Serial Ports 1 16-Timer, 1 Watch Dog Timer
|
No
|
HD6417612F60
|
http://semiconductor.hitachi.com/superh
|
http://semiconductor.hitachi.com/superh
|
| Hitachi
|
MPU-based
|
Standard
|
X
|
16
|
16
|
|
|
|
Mod. Harvard
|
RISC
|
154
|
Scalar
|
one ALU, one multiplier, two address
|
1
|
Memory locations: 256M word width: 8; external,
SDRAM
|
Memory locations: 4K word width (bits):16; internal, RAM
|
1
|
SH-1,SH-2
|
Yes
|
Yes
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes:2 DMA,3 Serial Ports 1 16-Timer, 1 Watch Dog Timer 1/100-Mbps Ethernet
|
No
|
HD6417615F60
|
http://semiconductor.hitachi.com/superh
|
http://semiconductor.hitachi.com/superh
|
| Hitachi
|
MPU-based
|
Standard
|
X
|
16
|
16
|
|
|
|
Mod. Harvard/ Reconfigurable
|
RISC
|
156
|
Scalar
|
1 ALU, 1 Multiplier, 2 Address
|
1
|
Memory locations: 488M word width: 8; internal & external, SDRAM
|
Memory locations: 488M word width: 8; internal & external, SDRAM
|
1
|
SH-1,SH-2, SH-3 and SH-DSP
|
Yes
|
Yes
|
No
|
Yes
|
No
|
Yes
|
Yes
|
Yes:4 DMA,3 Serial Ports 3 32-Timers, 1 Watch Dog Timer
|
Yes
|
HD6417729F133
|
http://semiconductor.hitachi.com/superh
|
http://semiconductor.hitachi.com/superh
|
|
Hitachi
|
MPU-based
|
Standard
|
|
|
|
X
|
32/64
|
32/64
|
Von Neuman
|
RISC
|
154
|
SuperScalar
|
three ALU, four multiplier, two address
|
1
|
Memory locations: 8K word width: 8; internal and external, CACHE
|
Memory locations: 16K word width (bits):8; internal, CACHE
|
1
|
SH-1,SH-2
|
Yes
|
Yes
|
No
|
Yes
|
No
|
Yes
|
Yes
|
Yes:4 DMA,2 Serial Ports 3 32-Timers, 1 Watch Dog Timer
|
No
|
HD6417750BP200
|
http://semiconductor.hitachi.com/superh
|
http://semiconductor.hitachi.com/superh
|
|
Hitachi
|
MPU-based
|
Standard
|
|
|
|
X
|
32/64
|
32/64
|
Von Neuman
|
RISC
|
154
|
SuperScalar
|
three ALU, four multiplier, two address
|
1
|
Memory locations: 8K word width: 8; internal and external, CACHE
|
Memory locations: 16K word width (bits):8; internal, CACHE
|
1
|
SH-1,SH-2
|
Yes
|
Yes
|
No
|
Yes
|
Yes
|
Yes
|
No
|
Yes:4 DMA,2 Serial Ports 5 32-Timers, 1 Watch Dog Timer, PCI
|
No
|
HD6417750BP200
|
http://semiconductor.hitachi.com/superh
|
http://semiconductor.hitachi.com/superh
|
| Infineon
|
MPU-based
|
Core
|
X
|
16
|
40
|
|
|
|
Mod. Harvard/ Reconfigurable
|
CISC
|
90
|
SuperScalar
|
18
|
8
|
Memory locations: 16K word width: 24; internal and external; ROM,RAM,EPROM, EEPROM
|
Memory locations: 64M word width (bits):16; internal and external; ROM, RAM, EPROM
|
3 ns
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
No
|
Yes
|
Yes
|
Yes
|
No
|
CARMEL
|
http://www.carmeldsp.com
|
http://www.carmeldsp.com
|
| Infineon
|
MCU-based
|
Standard/Core
|
X
|
16
|
48
|
|
|
|
Von Neuman
|
RISC
|
409
|
SuperScalar
|
three ALU, two multiplier, one address
|
3
|
Memory locations: 16K word width: 24; internal; ROM,RAM
|
Memory locations: 8K word width (bits):64; internal; RAM
|
5 ns
|
n/a
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
No
|
TC10GP
|
http://www.infineon.com/tricore
|
http://www.infineon.com/tricore
|
| LSI Logic
|
|
Standard/Core
|
X
|
16
|
40
|
|
|
|
Mod. Harvard
|
RISC
|
16-bit instruction set
|
SuperScalar
|
three ALU, two multiplier, two address
|
|
Memory locations: 128K/1M word width: 16; internal, shared; ROM,RAM
|
Memory locations: 128K/1M word
width: 16; internal, shared; ROM,RAM
|
Real: N/2 cycles. Therefore 2.5 ns per tap @ 200 MHz.
|
Assembler level compatibility across all processors.
|
Yes
|
No
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
No
|
LSI401Z
LSI402Z
Cores:
ZSP400
|
http://www.zsp.com
|
http://www.zsp.com
|
| Lucent
|
|
Standard
|
X
|
16
|
36
|
|
|
|
Mod. Harvard
|
RISC
|
80
|
Scalar
|
one ALU, one multiplier, two address, two post modify
|
|
Memory locations: 64K word width: 16; internal and external; ROM,RAM, EPROM, EEPROM
|
Memory locations: 64K word width: 16; internal and external; RAM,
|
8.3 ns
|
DSP16000 via translator
|
No
|
No
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
DSP1609, DSP1690, DSP1610, DSP1611, DSP1616, DSP1617, DSP1618, DSP1620, DSP1627, DSP1629, POMP16, DSP1695, DSP1673, DSP1643
|
http://www.lucent.com/micro/dsp/wiredoc.html
|
http://www.lucent.com/micro/dsp/wiredoc.html
|
| Lucent
|
|
Standard/Core
|
X
|
16
|
40
|
|
|
|
Mod. Harvard
|
RISC
|
180
|
Scalar
|
two ALU, two multiplier, two address, two post modify
|
|
Memory locations: 1M word width: 16; internal and external, shared; ROM,RAM, EPROM, EEPROM
|
Memory locations: 16M word width: 16; internal, external, shared; RAM
|
Two taps can be calculated every clock cycle (due to dual-MAC architecture).
A clock cycle is 6.6 ns resulting in a throughput of 1 tap/3.3 ns.
|
|
Yes
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
DSP16210, DSP16410, DSP16270, DSP16370,
|
http://www.lucent.com/micro/dsp16000/dsp16000_doc.html
|
http://www.lucent.com/micro/dsp16000/dsp16000_doc.html
|
| Phillips
|
MCU- and MPU-based
|
Core
|
X
|
16
|
16/40
|
|
|
|
Dual Harvard
|
CISC
|
n/a
|
Modified VLIW through application specific instructions
|
four ALU, two multiplier, dedicated shift, barrelshift/normaliser, two address calculation, saturation/scaling/rounding
|
|
X data memory: up to 64 Kword (SRAM, ROM)
Y data memory: up to 64 Kword (SRAM, ROM)
X and Y data memory may be
shared
Memory is internal, external, shared; RAM,ROM
|
X data memory: up to 64 Kword (SRAM, ROM)
Y data memory: up to 64 Kword (SRAM, ROM)
X and Y data memory may be shared
|
Two taps per cycle @ 90 MHz
|
|
No
|
No
|
No
|
Yes
|
Yes
|
|
Yes
|
Yes
|
Yes
|
|
|
|
| Philips
|
MCU- and MPU-based
|
Core
|
X
|
24
|
24/56
|
|
|
|
Dual Harvard
|
CISC
|
n/a
|
Modified VLIW through application specific instructions
|
four ALU, two multiplier, dedicated shift, barrelshift/normaliser, two address calculation, saturation/scaling/rounding
|
|
X data memory: up to 4096K word (SRAM, ROM)
Y data memory: up to 4096K word (SRAM, ROM)
X and Y data memory may be shared
Memory is internal, external, and shared; RAM,ROM
|
X data memory: up to 4096 Kword (SRAM, ROM)
Y data memory: up to 4096 Kword (SRAM, ROM)
X and Y data memory may be shared
|
Two taps per cycle @ 150 MHz
|
|
Yes
|
No
|
Yes
|
Yes
|
Yes
|
|
Yes
|
Yes
|
Yes
|
|
|
|
| Philips
|
MCU- and MPU-based
|
Standard
|
X
|
8/16/32
|
64
|
|
32
|
64
|
Mod Harvard
|
RISC
|
210
|
VLIW
|
5
|
27
|
Off-chip SDRAM up to 64MB, RAM
|
External SDRAM
|
7 ns
|
TriMedia and Nexperia DVP
|
Yes
|
Yes
|
Yes
|
No
|
Yes
|
No
|
No
|
Yes
|
No
|
TM-1100 TM-1300
|
http://www.trimedia.philips.com
|
http://www.trimedia.philips.com
|
| Star*Core
|
MPU-based
|
Core
|
X
|
16
|
40
|
|
|
|
Mod Harvard
|
RISC
|
185
|
Variable length execution set (VLES): a multi-issue architecture that allows up to 6 executable 16-bit RISC instructions, and 2 prefix instructions to be issued simulaneously w/no alignment or NOP padding
|
four DALU (MAC, ALU or BFU), two address
|
4 DALUs
|
Core memory, so no memory included. Lucent & Motorola will add memory and peripherals to the final SOC products they release to market
|
Core memory, so no memory included. Lucent & Motorola will add memory and peripherals to the final SOC products they release to market
|
|
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes - Emulation, peripherals/memory & potentially mixed-signal and
multi-core versions will be made available by Lucent & Motorola as they develop the final SOC products
|
Yes - Emulation, peripherals/memory & potentially mixed-signal and multi-core versions will be made available by Lucent & Motorola as they develop the final SOC products
|
Yes - Emulation, peripherals/memory & potentially mixed-signal and multi-core versions will be made available by
Lucent & Motorola as they develop the final SOC products
|
DSP Core: SC140
|
http://www.starcore-dsp.com
|
http://www.starcore-dsp.com
|
| Texas Instruments
|
More DSP-based
|
Standard
|
X
|
16
|
32
|
|
|
|
Mod Harvard
|
CISC
|
87
|
|
32-bit CALU, three scaling shifters, 16-bit x 16-bit multiplier with 32-bit product capability, 32-bit accumulator
|
Same as #8
|
256 words x 16 bits, internal, shared, RAM
|
288 words x 16 bits, Internal RAM
|
33 ns
|
|
Yes
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
TMS320LF2407, TMS320LF2406,
TMS320LF2402, TMS320LC2406, TMS320LC2404, TMS320LC2402
|
http://www.ti.com/sc/dmc
|
http://www.ti.com/sc/dmc
|
| Texas Instruments
|
DSP-based
|
Standard
|
X
|
16
|
32
|
|
|
|
Mod Harvard
|
CISC
|
87
|
|
32-bit CALU, three scaling shifters, 16-bit x 16-bit multiplier with 32-bit product capability, 32-bit
accumulator
|
Same as #8
|
256 words x 16 bits, internal, shared, RAM
|
|
50 ns
|
No
|
Yes
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
TMS320F243, TMS320F241, TMS320F240,
TMS320C242, TMS320C240
|
http://www.ti.com/sc/dmc
|
http://www.ti.com/sc/dmc
|
| Texas Instruments
|
DSP-based
|
Standard
|
X
|
32
|
40
|
X
|
32
|
32
|
Mod Harvard
|
RISC
|
50
|
VLIW
|
Eight orthogonal functional units including two multipliers and six ALUs
|
Same as #8
|
Dependent on device type. Please contact vendor for details.
|
Dependent on device type. Please contact vendor for details.
|
Half a clock cycle per tap
|
All eight devices within the 'C6000 platform are completely object code.
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
Yes
|
No
|
TMS320C6201, TMS320C6202, TMS320C6203, TMS320C6204, TMS320C6205,
TMS320C6211, TMS320C6701, TMS320C6711
|
|
|
| Texas Instruments
|
DSP-based
|
Standard
|
X
|
32
|
32
|
X
|
40
|
40
|
Mod Harvard
|
RISC
|
100+
|
Scalar
|
two address, two ALUs, one branch
|
Same as #8
|
Word width: 32 bits, internal, shared, RAM
|
|
One clock
cycle per tap
|
All C3x devices compatible with eachother. Not code compatible with other DSPs
|
Yes
|
No
|
Yes
|
Yes
|
Yes for C33, No for C31 and C32
|
Yes
|
Yes
|
Yes
|
No
|
TMS320C31, TMS320C32, TMS320C33
|
|
|
| VLSI
|
MPU-based
|
Core
|
X
|
16
|
40
|
|
|
|
Mod Harvard
|
RISC
|
35
|
Scalar
|
ALU/Multiplier + two address + one loop
|
1
|
Up to 4G x 32, internal, external, shared, ROM , RAM
|
Up to 4G x 16, internal, external, shared, ROM, RAM
|
1 cycle, 10 ns (in 0.25-ým technology)
|
VS_DSP1
|
Yes
|
Yes
|
No
|
No
|
No
|
Yes
|
No
|
Yes
|
Yes
|
VS_DSP1, VS_DSP2
|
http://vlsi.fi/ip/
|
|
| VLSI
|
MPU-based
|
Core
|
X
|
24
|
56
|
|
|
|
Mod Harvard
|
RISC
|
68
|
Scalar
|
ALU/Multiplier +one Address + one loop
|
1
|
Up to 64k x 24, internal and external; ROM, RAM
|
Up to 2 x 64k x 24, Internal, External, ROM, RAM
|
1 cycle, 15 ns
(0.35-ým technology)
|
|
Yes (from 3rd party)
|
Yes (from 3rd party)
|
No
|
No
|
Yes
|
Yes
|
No
|
Yes
|
Yes
|
VS56000
|
vlsi.fi/ip/
|
vlsi.fi/ip/
|
| ZiLOG
|
MCU-based
|
Standard, Core
|
X
|
16
|
24
|
|
|
|
Mod Harvard
|
RISC
|
30 (Clarkspur)
|
Scalar
|
one ALU, one multiplier, one address
|
|
0-8K x (0-8K)x16, internal and external; ROM
|
512 x 256 x 16+256 x 16,
internal, RAM
|
50 ns
|
|
Yes
|
No
|
No
|
No
|
No
|
Yes
|
Yes
|
Yes
|
Yes
|
Z89321, Z89371, Z89323, Z89373, Z89223, Z89273
|
http://zilog.com/products/dspspecs.html
|
http://zilog.com/products/dspapp.html
|