Case Study: Cornell University achieves rapid analysis of massive acoustic datasets with MathWorks' MATLAB HPC platform

Cornell Bioacoustics Scientists Develop a High-Performance Computing Platform for Analyzing Big Data

Cornell University’s Bioacoustics Research Program faced the challenge of detecting and classifying animal sounds in massive passive acoustic data sets—hundreds of terabytes and millions of events—where environmental noise and variability made algorithms prone to false positives/negatives and left most data unprocessed. To tackle this, the team worked with MathWorks tools, notably MATLAB along with Parallel Computing Toolbox and MATLAB Distributed Computing Server, to prototype and scale detection and classification workflows.

Using MathWorks’ MATLAB ecosystem, Cornell built an HPC platform that optimized and parallelized detection, classification, noise analysis, and acoustic modeling on a 64‑worker cluster and provided an interface for researchers to run experiments. The MathWorks‑based solution cut development from an estimated three years to under three months, reduced a 19‑week analysis to 8 hours, and enabled processing of a 100,000‑hour data set six times in two days—turning previously unanalyzed data into actionable ecological insights.

Cornell University

Christopher Clark

Senior Scientist and Director