Seismic Imaging Gets a GPU Boost From Viridien, NVIDIA

A new partnership aims to speed seismic imaging using accelerated computing, giving subsurface teams sharper insights into complex geology

10 Mar 2026

Viridien and Nvidia are collaborating to accelerate seismic imaging by adapting Viridien’s subsurface imaging software to Nvidia’s accelerated computing platforms, in an effort to speed the interpretation of complex geological data.

The partnership, announced on February 18, focuses on optimising Viridien’s algorithms for Nvidia hardware, including the use of tensor cores and mixed-precision computing. The companies say these techniques can improve computing performance while maintaining imaging accuracy.

Faster seismic processing is increasingly important for energy companies seeking to interpret large volumes of data more quickly. Clearer images of subsurface structures can reduce uncertainty during early exploration stages, helping teams evaluate prospects, plan well locations and guide development strategies.

Viridien has recently highlighted advances in seismic imaging in the US Gulf of Mexico, where it has applied full-waveform inversion, a computational method used to build detailed models of subsurface geology, to improve the clarity of complex subsalt formations.

The companies say combining such algorithms with high-performance computing infrastructure could shorten processing times for large seismic surveys. Advances in graphics processing units (GPUs) have increasingly drawn interest from geoscience groups seeking alternatives to traditional CPU-based computing.

The partnership also reflects a broader shift in the energy sector, where high-performance computing is becoming more central to subsurface workflows as datasets grow in size and complexity.

However, the commercial impact remains uncertain. The companies did not disclose customer agreements, revenue expectations or benchmark performance improvements tied to the collaboration, leaving the scale of potential gains still to be demonstrated in operational projects.