https://hgpu.org/?p=23723
It's all about data movement: Optimising FPGA data access to boost performance