Rapid application development for HPC. New GPGPU technology.

Full waveform inversion trial leverages Python and C++. Tesla Maxwell shines in benchmark.

In a presentation at a recent EAGE workshop on high performance computing, Geoff Clark (Acceleware) demonstrated the use of the Python programming language for rapid development of research codes. Clark’s team used the SciPy extension to demonstrate its full waveform inversion algorithm on the EAGE’s Marmousi synthetic data set. A combination of Python and C++ codes provides researchers with the tools required to tailor full waveform inversion to specific applications and optimize codes for HPC clusters.

Acceleware has also been testing the latest Nvidia Tesla/Maxwell architecture which, while it promises significant improvement on single precision floating-point arithmetic, is less performant on double precision where the older K40/K80 will outperform the new M40/M60 cards. More from Acceleware.

Click here to comment on this article

Click here to view this article in context on a desktop

© Oil IT Journal - all rights reserved.