
RFEL makes FPGA performance available to ARM
"Making Zynq-7000 SoC design consultancy part of our service offering," explained Dr. Alex Kuhrt, RFEL’s CEO, "allows us to offer support to engineers and project managers who wish to take advantage of our world-class signal processing FPGA-based portfolio, but who do not wish to engage in FPGA development at the VHDL or Verilog ‘code-face’. In particular, we want to offer a low-risk, effective path for software development companies, or embedded product developers with little or no FPGA experience to take advantage of everything that FPGAs can offer in terms of system performance, total power and resource efficiency, but without leaving the microprocessor design domain that they are familiar with."
Zynq-7000 devices are essentially an ARM® Cortex®-A9 MPCore™ processor with NEON and double precision floating point extensions, and a rich peripheral set; Standard fare for those already working with advanced ARM processors, but what makes this device so attractive is the low-latency, high bandwidth, Xilinx 28nm FPGA programmable logic fabric that is included and directly connected to the processing core by multiple AMBA AXI4 interfaces, DMA channels and shared memory interfaces.
Adopting Zynq-7000 All Programmable SoC is a straightforward option for all members of a product development team. The hardware designer can add Zynq devices to their PCB design as easily as with any other conventional ARM device, since the processor-centric design facilitates fast turn-on, system bring-up and mastering. The software designer can target the platform in comfort, using standard tools from ARM and other well-known 3rd party developers. A selection of operating systems is supported with Zynq-7000 platform specific distributions, and the marriage of FPGA programmable logic and processor in single chip system solutions will reduce costs and development risks. Of course, the capability should equally be of interest to FPGA developers who are limited in their exposure to ARM development. Zynq adds floating point co-processor capability, allowing a system to perform complex non-real-time calculations and apply the results to the real-time FPGA data stream.
