Texas Instruments
SemiconductorsDSP SolutionsSearchFeedBackTI Home
Product InformationIn the NewsToolsLiteratureSupport

Digital Signal Processing
Blue Band
Digital Signal Processing Solutions Products from Texas Instruments: TI TMS320 DSP Solutions Device Features         Device Specifications

Development Tools         Technical Documentation

Benchmark         Third Party

Training         Executive Summary

New C549 & C5410         Home

TMS320C54x Features

Key Features:
  • 40/50/66/80/100-MIPS performance
  • 2.5-, 3.3-, and 5-volt versions available
    TMS320C54x Block Diagram
  • Integrated Viterbi operation
  • Three power-down modes
  • Integrated RAM/ROM
  • Buffered serial ports
  • Multi-channel Buffered Serial
        Port
  • 6-channel DMA controller
  • Host port interface
  • Ultra-thin packaging

    Optimized CPU architecture

    The high performance levels of the TMS320C54x DSPs are made possible by an innovative architecture designed to meet the needs of a variety of applications reduces Viterbi “butterfly update” down to only four instruction cycles for GSM channel decoding. This frees MIPS so the device CPU can perform other telephony system tasks.

    In addition to the Viterbi accelerator, other key features include:

    • Four internal buses and dual address generators enable multiple operand operations and reduce memory bottlenecks.
    • A 40-bit adder and two 40-bit accumulators support crucial parallel instructions that execute in only one instruction cycle.
    • A second 40-bit adder available at output of the multiplier allows unpipelined MAC operation as well as dual addition and multiplication in parallel.
    • Single-cycle normalization and exponential encoding support floating-point arithmetic that is useful in voice coding.
    • A 17 x 17 multiplier allows 16-bit signed or unsigned multiplication, with rounding and saturation control—all in one instruction cycle.
    • New single-cycle instructions efficiently execute common DSP tasks like a symmetrical FIR filter.
    • A 40-bit arithmetic logic unit (ALU) features a dual 16-bit configuration capability to enable dual, one-cycle operations.
    • Eight auxiliary registers and a software stack enable the industry’s most advanced fixed-point DSP C compiler.

    Return to top


    Integrated Viterbi operation

    Another example of the innovative architecture of the C54x is the Viterbi accelerator that reduces Viterbi “butterfly update” down to only four instruction cycles for GSM channel decoding. This frees MIPS so the device CPU can perform other telephony system tasks.

    Return to top


    Power-down modes for ultra-low power

    TMS320C54x DSPs feature three power-down modes which are critical for power-sensitive systems. They are executed by the IDLE1, IDLE2, IDLE3 instructions. In these modes, the ’C54x DSP enters a dormant state and dissipates considerably less power than in normal operation. The core CPU shutdown sequence is activated when the IDLE1 instruction is executed.

    However, the system clock remains applied to the peripheral modules. The IDLE2 instruction shuts down the core CPU and peripheral modules, while IDLE3 is used for a complete shutdown of the processor.

    Millwatts per function - the true measure of power dissipation

    Processing capability in terms of MIPS is important but, due to differences in DSP architecture, MIPS data on one processor may not be as powerful as MIPS data on another processor.

    When a DSP executes a task more efficiently, designers are able to use remaining MIPS to implement tasks normally handled by off-chip ASICs or microcontrollers. This integration not only yields space savings, but power savings as well. Also, when a DSP task can be done more efficiently, the DSP can spend more time in power-down, or IDLE, mode.

    Thus, milliwatts per function—a measurement that comprehends the MIPS differences between architectures—is the most accurate and definitive way to measure power dissipation.

    The TMS320C54x DSPs feature three IDLE modes, low voltage operation down to 2.5 V at full performance, and an advanced architecture to efficiently implement wireless communication system tasks. For example, the TMS320C54x DSPs take only an average of 12.7 MIPS to implement full-rate GSM, 26.2 MIPS to implement half-rate GSM, only 2.3 MIPS for the Full-Rate GSM Speech Codec, and only 12.8 MIPS for IS-54/136 VSELP.

    Putting this into perspective of milli-watts per function—IS-54/136 VSELP requires only 31.1 mW, and the GSM Speech Codec only 5.6 mW of dissipated power. Clearly, the efficiency of the TMS320C54x architecture enables TI to provide the lowest milliwatts per function capability for the wireless communications market.

    Return to top


    Buffered serial ports

    An auto-buffered serial port (BSP) provides read-write capability to a 2K-word data buffer, significantly lowering the processor’s overhead. The BSP can buffer up to 2K data samples before the CPU processes the stored data. The maximum throughput of the BSP is 50 Mbits/sec at a 20-ns instruction cycle. The BSP can operate at full-speed even while the CPU is in IDLE mode.

    Return to top


    Multi-channel Buffered Serial Port

    The Multi-Channel Buffered Serial Port (MCBSP) is a superset of the BSP. The MCBSP enables full-duplex, bi-directional communication and direct interface to T1/E1 framers or H.100 telephony standards. It supports multi-channel transmit and receive up to 128 channels and data sizes from 8 bits to 32 bits.

    Return to top


    6-channel DMA controller

    The six-channel DMA controller offers added flexibility and enhances performance. The DMA allows the peripherals to operate independently of the CPU and can facilitate program overlays, maximizing the overall efficiency of the device.

    Return to top


    Host port interface

    A host port interface (HPI) provides a glueless interface to standard microprocessors as well as other TI devices. The HPI has the ability to maintain its high level of functionality as the interface between the host CPU and the ’C54x, even while the internal processor is idle.

    Return to top


    Ultra-thin packaging

    The ’C54x generation allows systems to reduce chip count to lower system cost or increase functionality with fewer physical components. The ultra-small physical size of the ’C54x microStar ball grid array (BGA) packaging can also help increase the performance per square inch for MIPS intensive or space constrained applications.

    Return to top


  • SemiconductorsDSP SolutionsSearchFeedBackTI Home
    © Copyright 1998 Texas Instruments Incorporated. All rights reserved.
    Trademarks, Important Notice!