BCH codes are widely used where bit errors are scattered randomly within the codeword. The Creonic Ultra-fast BCH Decoder is capable of processing an entire BCH codeword per clock cycle in a pipelined way. Therefore, tt achieves outstanding data rates.
The design can be parameterized at design-time to support different codeword sizes and code rates. Latency can be adjusted by insertion or removal of pipeline register stages.
UltraFast BCH Decoder
Overview
Key Features
- Coded throughput of up to 163.5 Gbit/s at 320 MHz (FPGA, 511 codeword bits)
- Coded throughput of up to 122.4 Gbit/s at 480 MHz (FPGA, 255 codeword bits)
- Latency in the range of 4 to 14 clock cycles (466 payload bits, 511 codeword bits))
Benefits
- Throughput of one codeword per clock cycle
- Fully pipelined design
- Design-time configuration of BCH code size and code rate
- Design-time configuration of primitive polynomial
- Design-time configuration of latency for optimal resource utilization
- Bit error rate measurement
- Low-power and low-complexity design
- Available for ASIC and FPGAs (AMD Xilinx, Intel, Microchip)
Applications
- Die-to-die communication
- Free-space optical (FSO) communication
- Further applications with the need for tremendous data rates
Deliverables
- VHDL or Verilog source code or synthesized netlist
- HDL simulation models e.g. for Aldec's Riviera-PRO
- VHDL testbench
- bit-accurate Matlab, C or C++ simulation model
- comprehensive documentation