T
o achieve amore accurate and complete picture of the subsurface,
oil and gas companies are leveraging acquisition technologies
such as full/wide‑azimuth data to improve illumination and
signal‑to‑noise ratio, reservoir simulation andmodelling of complex
reservoirs. This helps understand and predict reservoir behaviour, and
reservoirmonitoring and 4D seismic help gain insight into pressure changes,
fluidmovement over time andmore. The common denominator among
these technologies is this: they create large amounts of data needed for
informed decisionmaking and forecasting.
This need to image, interpret, model and simulate increasingly complex
fields has seenmany companies discovering limitations in their current
technology investments – especiallywith regard to hardware and software
architecture and infrastructure – and has sent them looking for solutions
that can address their challenges.
This articlewill outline some of the issues facing the oil and gas industry
with regard to technology demands created by increasing data sizes.
Additionally, it will offer a look at approaches companies can use to leverage
current software and hardware technologies tomitigate these challenges
and provide solutions to the industry that will prepare it for the next cycle of
innovation.
Newemergingmodels
Today, innovation is beginning to emerge in cloudmodels, colocating data,
computation and visualisation, and remote visualisation technology – not
tomention using virtualisation technologies to serve the needs of many
concurrent users and diverse applications. Yet these innovations are still far
fromrealising their full potential, as geophysicists continue to rely on old
technologies and fail to capitalise on the capabilities available.
The industry is now looking formodels that will allowexploration
and production technology to scale up to high‑performance computing
standards, and it is likely the technologies that achieve thiswill lay the
foundations for the next 10 to 20 years of innovation. Thankfully, these
models aremoving closer to becoming reality, largely due to advancements
in graphics acceleration technology such as NVIDIA graphics processing
units (GPUs).
Ask any application support group – today, howmany in‑house and
commercial software packages take full advantage of currently available
GPU computational power? And howmany companies have invested
in hardware powerful enough to support them? Can hardware and
software handle terabyte data sets and power multi‑GPU computations
on the fly?
Linda Gaasø, Hue, Chris McCoy,
Lenovo Thinkstation, Jim Madeiros,
Magma, and Ty McKercher, Nvidia,
explain how modern computing
technology can enable geoscience
innovation in the E&P industry.
PUTTING GPUs
TOUSE IN
GEO
Figure 1.
Interpretation of complex geologic
environments requires innovative displays.
22 |
