|
|
Executive Summary
Device Features
'C67x Architecture
Device Specifications
Development Tools
Technical Documentation
 TMS320C67x Third Parties
TMS320C67x Training
Schedules for Europe
and North America
TMS320C67x Assembly Benchmarks
TMS320C6000 Home
|
|
TMS320C67x
Digital Signal Processors
Now, designers of floating-point DSP applications can use the industry-leading performance of the TMS320C6000 DSP platform to create tomorrow's visionary applications. The first floating-point DSP to break the 1 GFLOP barrier is the TMS320C6701. And breakthrough performance isn't the only advantage of the 'C6701. Using TI's revolutionary VelociTI™ architecture, the 'C6701 offers code compatibility and pin-for-pin compatibility with the 1600 MIPS 'C6201 fixed-point DSP. This means a single set of development tools and board-level device interchangeability that provides significant development, resource and manufacturing cost savings.
The TMS320C6201 -- the 'C6701's fixed-point counterpart -- is in production today and is already being used to revolutionize next-generation DSP applications. Now, the 'C6701 brings this same performance to floating-point designs. The device will enable faster speed, precision and dynamic range in next-generation applications like wireless local loop and beam-forming base stations, virtual reality 3-D graphics, speech recognition, audio, radar, industrial control, and imaging systems such as fingerprint recognition, CAT scan, ultrasound and magnetic resonance imaging (MRI) among many others.
TI delivers this leap-frog performance with the revolutionary VelociTI™ advanced Very Long Instruction Word (VLIW) DSP architecture -- a combination of 21st century hardware and development tools that all 'C6000 devices have in common. This means 'C6000 devices share more than high performance. Their common architecture also provides architectural, code and pinout compatibility for simple interchangeability of fixed and floating point members. For instance, designers can develop code for the TMS320C6201 fixed-point device, and then port the same instructions to the 'C6701 floating-point family. And because the devices are pin-for-pin compatible, they are interchangeable on the production board.
|
|
