Embedded Glossary

x86 refers to a 16/32-bit processor architecture that was initially developed by Intel. The name is based on the first 16-bit CPU, _8086_, in this series from 1978. By design, x86 is assigned to the CISC architecture; originally with 16-bit-wide registers and a data bus as well as a 20-bit address bus and therefore a maximum of 1 MB of addressable memory. The first instruction set is still supported today. In the 1980s and 1990s, this architecture dominated the desktop segment and was manufactured under licence and further developed by other manufacturers (AMD). To this day, it represents a de facto standard. In the embedded field, it is used when compatibility with the desktop field, including the operating system used, and high processing performance are the priority. This is associated with high power consumption (65–105 W for the AMD Ryzen™ Embedded 7000 Series) and high thermal load. With CPUs not manufactured specifically for the embedded field, all interfaces must be connected as external hardware. x86 remains a de facto standard for embedded applications requiring desktop compatibility and high processing performance.

Xilinx is a supplier of programmable logic devices whose FPGAs and SoC families have been run under AMD's roof since its acquisition in 2022. An FPGA can be freely configured after manufacture and processes signal paths in parallel in hardware, rather than working through them step by step on a processor. Well known is the Zynq range, which combines programmable logic and an ARM processing core on a single chip. One example is the Zynq-7000 family with a core from the Cortex-A9 series. On the processing core, an Embedded Linux often runs, while time-critical data streams arise directly in the FPGA. This makes the architecture well suited to fast sensor connection and image processing. Where software and signal processing have to mesh closely, Xilinx brings its strengths of programmable hardware and an integrated processor fully to bear.