Case Study: Accenture Labs accelerates cybersecurity threat detection with NVIDIA GPUs and GPU-accelerated libraries

Share Your Science Visualizing 200M Cybersecurity Alerts Daily with GPUs

Accenture Labs, through principal data scientist Joshua Patterson, was working to analyze massive network graphs to detect cybersecurity threats and anomalies across an enormous volume of alerts. The challenge was handling very large-scale graph data fast enough to support security investigations, using NVIDIA technology to accelerate the work.

NVIDIA provided GPUs and GPU-accelerated libraries, including the DGX-1 deep learning supercomputer and the nvGRAPH library in CUDA 8, to help Accenture Labs process graphs with up to 4 billion nodes. NVIDIA’s accelerated computing approach improved analysis speed and was expected to cut months from development cycles, enabling faster detection of security threats.

Accenture Labs

Joshua Patterson

Principal Data Scientist