Improv moves Jazz DSP core into high-volume 16-bit markets
Improv moves Jazz DSP core into high-volume 16-bit markets
By Semiconductor Business News
June 5, 2001 (12:39 p.m. EST)
URL: http://www.eetimes.com/story/OEG20010605S0032
BEVERLY, Mass. -- Taking aim at a broader range of DSP applications, Improv Systems Inc. here this week rolled out its Jazz 2 digital signal processor technology, using a streamlined 16-bit derivative of its initial 32-bit Jazz 1 architecture launched last year. "The 32-bit Jazz 1 is for high performance--like a 'DSP dragster'--but the Jazz 2 is being aimed at 80% of the market to serve everything from hearing aids to MP3 players to residential gateways," explained Steven Brightfield, vice president of marketing at Improv. Like the more powerful 32-bit Jazz 1 processor, the new DSP architecture is aimed at system-on-chip designs and offers a high degree of scalability for multiprocessor ICs, said Brightfield. The architecture uses a non-blocking soft switch, called a "DataCom" block, which eliminates bottlenecks in accessing register files while expanding the overall performance, he said. But to tailor the new Jazz 2 for higher-volume applications, Improv created a new architecture with a 16-bit data path and streamlined the DSP register transfer language (RTL) code to lower power dissipation as well as increase gate efficiency, Brightfield said. "The low-end implementation is 50,000 gates," he noted. The company also has added low-power static and dynamic features to the Jazz 2 core, and it improved code density by 50% compared to existing DSPs, said Brightfield. One novel technique used in the Jazz 2 is a compiler than generates a binary image of code, which is then decompressed on-the-fly by the processor when software is fetched from memory. The decompression function requires just a few thousand gates on the processor, but it enables the compiler to dramatically reduce code size and lower memory requirements, Brightfield said. Improv is also supporting system-on-chip designs that combine the Jazz 2 DSP with reduced instruction set computing (RISC) processors from ARM Ltd. and MIPS Technologies Inc. ARM cores are supported wi th AMBA/AHB bus interfaces while the MIPS PI bus supports MIPS cores. The Jazz 2 Performer Tool Suite 2.0 hardware/software design environment incorporates an automated processor configuration tool--called Jazz Composer 2. This tool will produce customized Jazz 2 processors within an hour, according to the company. The core and enhancements to the technology will allow designers to target 0.13-micron and below process technologies for clock speed of 200 MHz. "We have done test chips in 0.2- and 0.15-micron process technologies. We working with partners to create test chips for 0.13-micron processes by the end of this year," Brightfield said. He said the first Jazz 2-based products are expected to be introduced this year as well. The Beverly company is also offering a set of pre-designed versions of Jazz 2 processors, ranging from a single-MAC (multiply-accumulator) to quad-MAC chips. The Jazz 2 processor and tool suites are now available. Design seats for standard tool suites begin at less t han $5,000 each. Improv also collects licensing and unit-royalty fees for ICs based on its DSP technology.
Related News
- LSI Logic Introduces ZSP200 DSP Core For High-Volume, Cost-Sensitive IC Markets
- CMX Systems announces unique, embeddable SNMP functionality for CMX-MicroNet TCP/IP stack for 8-bit, 16-bit and DSP processors
- Oki licenses 16-bit general-purpose core from DSP Group
- STMicro doubles speed of 16-bit MCU, adds DSP functions
- Brite Launches High-Precision 16 bit SAR ADC
Breaking News
- Ubitium Debuts First Universal RISC-V Processor to Enable AI at No Additional Cost, as It Raises $3.7M
- TSMC drives A16, 3D process technology
- Frontgrade Gaisler Unveils GR716B, a New Standard in Space-Grade Microcontrollers
- Blueshift Memory launches BlueFive processor, accelerating computation by up to 50 times and saving up to 65% energy
- Eliyan Ports Industry's Highest Performing PHY to Samsung Foundry SF4X Process Node, Achieving up to 40 Gbps Bandwidth at Unprecedented Power Levels with UCIe-Compliant Chiplet Interconnect Technology
Most Popular
- Cadence Unveils Arm-Based System Chiplet
- CXL Fabless Startup Panmnesia Secures Over $60M in Series A Funding, Aiming to Lead the CXL Switch Silicon Chip and CXL IP
- Esperanto Technologies and NEC Cooperate on Initiative to Advance Next Generation RISC-V Chips and Software Solutions for HPC
- Eliyan Ports Industry's Highest Performing PHY to Samsung Foundry SF4X Process Node, Achieving up to 40 Gbps Bandwidth at Unprecedented Power Levels with UCIe-Compliant Chiplet Interconnect Technology
- Arteris Selected by GigaDevice for Development in Next-Generation Automotive SoC With Enhanced FuSa Standards
E-mail This Article | Printer-Friendly Page |