Tensor Processing Units are specialized hardware devices built to train and apply Machine Learning models at high speed through high-bandwidth memory and massive instruction parallelism. In this short paper, we investigate how relational operations can be translated to those devices. We present mapping of relational operators to TPU-supported TensorFlow operations and experimental results comparing with GPU and CPU implementations. Results show that while raw speeds are enticing, TPUs are unlikely to improve relational query processing for now due to a variety of issues.

International Workshop on Data Management on New Hardware, DaMoN 2019, Held with ACM SIGMOD/PODS 2019
Centrum Wiskunde & Informatica, Amsterdam, The Netherlands

Holanda, P.T, & Mühleisen, H.F. (2019). Relational queries with a tensor processing unit. In Proceedings of the 15th International Workshop on Data Management on New Hardware. doi:10.1145/3329785.3329932