Frontier Design cores handle up to 6,250 channels
Frontier Design cores handle up to 6,250 channels
By Patrick Mannion, EE Times
November 1, 2000 (12:28 p.m. EST)
URL: http://www.eetimes.com/story/OEG20001101S0048
SANTA CLARA, Calif. Frontier Design has introduced four adaptive differential pulse code modulation (ADPCM) intellectual property cores that can handle up to 256 channels in a single Xilinx Virtex FPGA, or up to 6,250 channels in an ASIC implementation. The cores are designed to capitalize on the increasing use of ADPCM for voice-over-Internet Protocol, digital telephony, intercoms, telephone answering machines, and mass storage. The four cores comprise two single- and two multi-channel cores that are available in either VHDL (IEEE1076-1987) or Verilog (IEEE 1364-1993) for optimized implementation in Xilinx Virtex XCV400E, XCV200E or XCV150 FPGAs, as well as for standard cell-based ASIC implementations. The ADPCM core is also available in a bit-true C-language implementation that can be parameterized for different applications. The highly parallelized core was designed using the company's A/RT Builder C-to-HDL translation tool, which gene rated the single-cycle implementation. The cores are fully programmable for 16-, 24-, 32-, or 40-kbit/second operation, with programmable A-law, mu-law or linear coding of input or output. Both cores are fully compliant with ITU-recommended G.726 test vectors. A 2.048-MHz clock on the FPGA implementation minimizes power operation. The 512-channel Virtex implementation requires 2,306 slices and 36 block RAMs, while the 256-channel Virtex implementation requires 2,020 slices and 18 block RAMs. The 128-channel Virtex implementation requires 1,800 slices, but only nine block RAMs. The ASIC version needs a clock frequency of 8 kHz per channel, so 128 channels can be encoded or decoded with a 1.024-MHz clock. Each channel of ADPCM encoding and decoding requires 280 bits of dual-port memory. The ASIC implementation of the 256-channel core requires about 20,000 gates or 78 gates per channel, operating at 2.048 MHz Using a 0.18-micron process, a clock speed of about 12.5 MHz is attainable, and allows 1,562 c hannels to be implemented in the 20,000-gate core. As a result, all 672 channels required for two DS3 lines can be implemented in 20,000 ASIC gates. With additional pipelining, the ADPCM ASIC core can handle the full 6,250 channels, which allows seven DS3 lines to be encoded/decoded in the core. Four additional pipeline stages would increase the gate count to 32,000. In addition, each encoding or decoding channel needs just a single clock cycle per sample. First-use pricing ranges from $20,000 for the single-channel Virtex FPGA, to $40,000 for the multichannel ASIC.
Related News
- Frontier Design introduces 6,250 channel ADPCM IP core for ASICs and 512 channel ADPC core for Virtex FPGAS
- Aniah raises €6 million to speed up the deployment of its verification and design support software for semiconductors
- Xiaomi picks up 6% stake in chip designer VeriSilicon
- First Quarter Semiconductor Sales Up 6 Percent Compared to Last Year
- Intilop presents at Server Design Summit, how their ultra high performance 10G full TCP Offload Technology increases the network bandwidth by up to 6 times
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 |