adsp-21478 Analog Devices, Inc., adsp-21478 Datasheet - Page 4

• Manufacturer Part Number: adsp-21478
• Description: Sharc Processor
• Manufacturer: Analog Devices, Inc.

ADSP-21478/ADSP-21479
As shown in the block diagram
two computational units to deliver a significant 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 processor achieves an instruction cycle time of 3.75
ns at 266 MHz. With its SIMD computational hardware, the
processors can perform 1.596 GFLOPS running at 266 MHz.
FAMILY CORE ARCHITECTURE
The ADSP-2147x is code compatible at the assembly level with
the ADSP-2146x, ADSP-2137x, ADSP-2136x, ADSP-2126x,
ADSP-21160, and ADSP-21161, and with the first generation
ADSP-2106x SHARC processors. The ADSP-2147x shares
architectural features with the ADSP-2126x, ADSP-2136x,
ADSP-2137x, ADSP-2146x and ADSP-2116x SIMD SHARC
processors, as shown in
sections.
SIMD Computational Engine
The ADSP-2147x contains two computational processing ele-
ments that operate as a single-instruction, multiple-data
(SIMD) engine. The processing elements are referred to as PEX
and PEY and each contains an ALU, multiplier, shifter, and reg-
ister file. PEX is always active, and PEY may be enabled by
setting the PEYEN mode bit in the MODE1 register. When this
mode is enabled, the same instruction is executed in both pro-
cessing elements, but each processing element operates on
different data. This architecture is efficient at executing math
intensive DSP algorithms.
Entering SIMD mode also has an effect on the way data is trans-
ferred between memory and the processing elements. When in
SIMD mode, twice the data bandwidth is required to sustain
computational operation in the processing elements. Because of
this requirement, entering SIMD mode also doubles the band-
width between memory and the processing elements. When
using the DAGs to transfer data in SIMD mode, two data values
are transferred with each access of memory or the register file.
SIMD mode is supported from external SDRAM but is not sup-
ported in the AMI.
Independent, Parallel Computation Units
Within each processing element is a set of computational units.
The computational units consist of an arithmetic/logic unit
(ALU), multiplier, and shifter. These units perform all opera-
tions in a single cycle. The three units within each processing
element are arranged in parallel, maximizing computational
throughput. Single multifunction instructions execute parallel
ALU and multiplier operations. In SIMD mode, the parallel
ALU and multiplier operations occur in both processing ele-
ments. These computation units support IEEE 32-bit single-
precision floating-point, 40-bit extended precision floating-
point, and 32-bit fixed-point data formats.
• Digital peripheral interface that includes two timers, a 2-
wire interface, one UART, two serial peripheral interfaces
(SPI), 2 precision clock generators (PCG), three pulse
width modulation (PWM) units, and a flexible signal rout-
ing unit (DPI SRU).
Figure 2
on Page
and detailed in the following
5, the processor uses
Rev. PrB | Page 4 of 70 | March 2010
Timer
The processor contains a core timer that can generate periodic
software interrupts. The core timer can be configured to use
FLAG3 as a timer expired signal.
Data Register File
A general-purpose data register file is contained in each pro-
cessing element. The register files transfer data between the
computation units and the data buses, and store intermediate
results. These 10-port, 32-register (16 primary, 16 secondary)
register files, combined with the processor’s enhanced Harvard
architecture, allow unconstrained data flow between computa-
tion units and internal memory. The registers in PEX are
referred to as R0-R15 and in PEY as S0-S15.
Context Switch
Many of the processor’s registers have secondary registers that
can be activated during interrupt servicing for a fast context
switch. The data registers in the register file, the DAG registers,
and the multiplier result registers all have secondary registers.
The primary registers are active at reset, while the secondary
registers are activated by control bits in a mode control register.
Universal Registers
These registers can be used for general-purpose tasks. The
USTAT (4) registers allow easy bit manipulations (Set, Clear,
Toggle, Test, XOR) for all system registers (control/status) of
the core.
The data bus exchange register (PX) permits data to be passed
between the 64-bit PM data bus and the 64-bit DM data bus, or
between the 40-bit register file and the PM data bus. These reg-
isters contain hardware to handle the data width difference.
Single-Cycle Fetch of Instruction and Four Operands
The ADSP-2147x features an enhanced Harvard architecture in
which the data memory (DM) bus transfers data and the pro-
gram memory (PM) bus transfers both instructions and data
(see
buses and on-chip instruction cache, the processor can simulta-
neously fetch four operands (two over each data bus) and one
instruction (from the cache), all in a single cycle.
Instruction Cache
The ADSP-2147x includes an on-chip instruction cache that
enables three-bus operation for fetching an instruction and four
data values. The cache is selective—only the instructions whose
fetches conflict with PM bus data accesses are cached. This
cache allows full speed execution of core, looped operations
such as digital filter multiply-accumulates, and FFT butterfly
processing.
Data Address Generators With Zero-Overhead Hardware
Circular Buffer Support
The ADSP-2147x’s two data address generators (DAGs) are
used for indirect addressing and implementing circular data
buffers in hardware. Circular buffers allow efficient program-
ming of delay lines and other data structures required in digital
Figure
Preliminary Technical Data
2). With its separate program and data memory