Timezone: »

Augur: Data-Parallel Probabilistic Modeling
Jean-Baptiste Tristan · Daniel Huang · Joseph Tassarotti · Adam Pocock · Stephen Green · Guy L Steele

Thu Dec 11 11:00 AM -- 03:00 PM (PST) @ Level 2, room 210D #None

Implementing inference procedures for each new probabilistic model is time-consuming and error-prone. Probabilistic programming addresses this problem by allowing a user to specify the model and then automatically generating the inference procedure. To make this practical it is important to generate high performance inference code. In turn, on modern architectures, high performance requires parallel execution. In this paper we present Augur, a probabilistic modeling language and compiler for Bayesian networks designed to make effective use of data-parallel architectures such as GPUs. We show that the compiler can generate data-parallel inference code scalable to thousands of GPU cores by making use of the conditional independence relationships in the Bayesian network.

Author Information

Jean-Baptiste Tristan (Oracle Labs)
Daniel Huang (Harvard University)
Joseph Tassarotti (Carnegie Mellon University)
Adam Pocock (Oracle Labs)
Stephen Green (Oracle Labs)
Guy L Steele (Oracle Labs)

More from the Same Authors