The NEUCHIPS RecAccelTM is world's first recommendation engine for Deep Learning Recommendation Model (DLRM). It can perform 500,000 inferences per second. It is equipped with an ultra-high capacity, high bandwidth memory subsystem for embedding table lookup, and a massively parallel compute FPGA for neural network inference. Running open-source PyTorch DLRM, RecAccelTM outperforms server class CPU and inference GPU by 28X and 65X, respectively.