Energy-Efficient Gaussian Processes Using Low-Precision Arithmetic
Abstract
The widespread use of artificial intelligence requires finding energy-efficient paradigms for the field. We propose to reduce the energy consumption of Gaussian process regression using lowprecision floating-point representations. We explore how low-precision representations impact the results of Gaussian process regression and how data set properties, implementation approach, model performance, and energy consumption interact. Our findings show that a well-conditioned kernel matrix allows reducing the energy consumption by up to 89.01% for 98.08% of arithmetic operations with little to no impact on model performance. Our findings are relevant whenever one needs to invert a symmetric full-rank matrix.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
Builds on1
Related papers
- Scalable Gaussian Processes with Latent Kronecker StructureJihao Andreas Lin, Sebastian Ament, Maximilian Balandat, David Eriksson et al.ICML 2025
- KernelMatmul: Scaling Gaussian Processes to Large Time SeriesTilman Hoffbauer, Holger H. Hoos, Jakob BossekAAAI 2025
- Privacy-Preserving Gaussian Process Regression - A Modular Approach to the Application of Homomorphic EncryptionPeter Fenner, Edward Pyzer-KnappAAAI 2020 · 25 citations
- Sampling from Gaussian Process Posteriors using Stochastic Gradient DescentJihao Andreas Lin, Javier Antorán, Shreyas Padhy, David Janz et al.NeurIPS 2023 · 34 citations
- Log-Linear-Time Gaussian Processes Using Binary Tree KernelsMichael K. Cohen, Samuel Daulton, Michael A. OsborneNeurIPS 2022 · 6 citations
