White light interferometry is a time consuming operation even on modern architectures. To overcome the high power consumption and size of traditional desktop computers an embedded approach containing the hybrid architecture Zynq will be presented. This architecture contains a dual core ARM and programmable logic provided by an FPGA. FPGAs offer massively parallel logic gates and DSP-slices to parallelise certain tasks. Another important part is the internal memory BRAM. The presented approach aims to speedup calculation time of the ARM processor by utilization of this BRAM. It is well known that memory transfers consume a lot of time. To speed the transfers up, the bottlenecks have to be identified. In this paper it will be illustrated how to easily access an FPGA BRAM from a running operating system and the possible speedup will be analysed and estimated.