| Electronic Components Datasheet Search |
|
ADSP-SC570 Datasheet(PDF) 9 Page - Analog Devices |
|
|
|||||||||||||||||||||||||||||
ADSP-SC570 Datasheet(HTML) 9 Page - Analog Devices |
|
9 / 142 page ![]() Rev. B | Page 9 of 142 | June 2018 ADSP-SC570/SC571/SC572/SC573/ADSP-21571/21573 Single-Instruction, Multiple Data (SIMD) Computational Engine The SHARC+ core 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 data registers and each contain an arithmetic logic unit (ALU), mul- tiplier, shifter, and register file. PEx is always active and PEy is enabled by setting the PEYEN mode bit in the mode control register (MODE1). SIMD mode allows the processors to execute the same instruc- tion in both processing elements, but each processing element operates on different data. This architecture efficiently executes math intensive DSP algorithms. In addition to all the features of previous generation SHARC cores, the SHARC+ core also pro- vides a new and simpler way to execute an instruction only on the PEy data register. SIMD mode also affects the way data transfers between memory and the processing elements because to sustain computational operation in the processing elements requires twice the data bandwidth. Therefore, entering SIMD mode doubles the band- width between memory and the processing elements. When using the DAGs to transfer data in SIMD mode, two data values transfer with each memory or register file access. Independent Parallel Computation Units Within each processing element is a set of pipelined computa- tional units. The computational units consist of a multiplier, arithmetic/logic unit (ALU), and shifter. These units are arranged in parallel, maximizing computational throughput. These computational units support IEEE 32-bit single-precision floating-point, 40-bit extended-precision floating-point, IEEE 64-bit double-precision floating-point, and 32-bit fixed-point data formats. A multifunction instruction set supports parallel execution of the ALU and multiplier operations. In SIMD mode, the parallel ALU and multiplier operations occur in both processing ele- ments per core. All processing operations take one cycle to complete. For all floating-point operations, the processor takes two cycles to complete in case of data dependency. Double-precision float- ing-point data take two to six cycles to complete. The processor stalls for the appropriate number of cycles for an interlocked pipeline plus data dependency check. Core Timer Each SHARC+ processor core also has a timer. This extra timer is clocked by the internal processor clock and is typically used as a system tick clock for generating periodic operating system interrupts. Data Register File Each processing element contains a general-purpose data regis- ter file. The register files transfer data between the computation units and the data buses, and store intermediate results. These 10-port, 32-register register files (16 primary, 16 secondary), combined with the enhanced Harvard architecture of the pro- cessor, allow unconstrained data flow between computation units and internal memory. The registers in the PEx data regis- ter file are referred to as R0–R15 and in the PEy data register file as S0–S15. Context Switch Many of the registers of the processor have secondary registers that can activate during interrupt servicing for a fast context switch. The data, DAG, and multiplier result registers have sec- ondary registers. The primary registers are active at reset, while control bits in MODE1 activate the secondary registers. Universal Registers General-purpose tasks use the universal registers. The four USTAT registers allow easy bit manipulations (set, clear, toggle, test, XOR) for all control and status peripheral registers. The data bus exchange register (PX) permits data to pass between the 64-bit PM data bus and the 64-bit DM data bus or between the 40-bit register file and the PM or DM data bus. These registers contain hardware to handle the data width difference. Data Address Generators (DAG) With Zero-Overhead Hardware Circular Buffer Support For indirect addressing and implementing circular data buffers in hardware, the ADSP-SC57x/ADSP-2157x processor uses the two data address generators (DAGs). Circular buffers allow effi- cient programming of delay lines and other data structures required in digital signal processing, and are commonly used in digital filters and fast Fourier transforms (FFT). The two DAGs of the processors contain sufficient registers to allow the cre- ation of up to 32 circular buffers (16 primary register sets and 16 secondary sets). The DAGs automatically handle address pointer wraparound, reduce overhead, increase performance, and simplify implementation. Circular buffers can start and end at any memory location. Flexible Instruction Set Architecture (ISA) The flexible instruction set architecture (ISA), a 48-bit instruc- tion word, accommodates various parallel operations for concise programming. For example, the processors can condi- tionally execute a multiply, an add, and a subtract in both processing elements while branching and fetching up to four 32-bit values from memory—all in a single instruction. Addi- tionally, the double-precision floating-point instruction set is an addition to the SHARC+ core. |
|
Link URL |
| Does ALLDATASHEET help your business so far? [ DONATE ] |
About Alldatasheet | Advertisement | Contact us | Privacy Policy | Link to Datasheet | Link Exchange | Manufacturer List All Rights Reserved©Alldatasheet.com |
| Russian : Alldatasheetru.com | Korean : Alldatasheet.co.kr | Spanish : Alldatasheet.es | French : Alldatasheet.fr | Italian : Alldatasheetit.com Portuguese : Alldatasheetpt.com | Polish : Alldatasheet.pl | Vietnamese : Alldatasheet.vn Indian : Alldatasheet.in | Mexican : Alldatasheet.com.mx | British : Alldatasheet.co.uk | New Zealand : Alldatasheet.co.nz |
|
Family Site : ic2ic.com |
icmetro.com |