QDIPS: Deep Sets Network for FPGA investigated for high speed inference on ATLAS

We adapted DIPS (Deep Impact Parameter Sets), a deep sets neural network flavour tagging algorithm previously used in ATLAS offline low-level flavour tagging and online b-jet trigger preselections, for use on FPGA with the aim to assess its performance and resource costs. A deep sets network architecture has useful applications in finding correlations in unordered and variable length data input. Its use on FPGA would open up accelerated machine learning in areas where the input has no fixed length or order. We compare an aggressively downscaled for-FPGA DIPS algorithm performance to the CPU-based full precision performance, and explore the associated FPGA resource costs.