During a seismic survey, each shot sends a wave propagating through the Earth, while receivers on the surface listen for reflections as that wave bounces off of geologic layers. Real-world geology can be extremely complex, and because of the different wave propagation velocities the of the different layers, the wave never expands in simple circles, like ripples in a pond. Instead it is scattered off of high-velocity contrasts, refracts around slower regions, is focused into beams.

The purpose of Reverse Time Migration (RTM) is to take those incredibly complex wavefields, as recorded at the surface, and form an image of the underlying geological structure.

The images shown here are Illumination Maps, which show how much energy from a single shot reached each point in the subsurface.

Click to enlarge

There were several information pieces that caught my attention over the last few weeks that seemed to be worthy of sharing. As one of the few non-technical people here at Acceleware, what I appreciated about all of these snipets was how clearly they affirmed the value of the technologies that we are working on. Two of these pieces have a connection with NVIDIA but the third is Intel, so that provides a good balance.

The first one occurred May 18th with IBM and NVIDIA announcing that Big Blue would start incorporating GPU technology in their iDataPlex Servers. It is another great endorsement for using GPUs as part of the processing engines in modern data centers. Check out the video with Scott Denham who gives a very concise overview of the multiple benefits of GPUs.

Acceleware's CTO Ryan Schneider introducing our Oil and Gas solutions and it's impact on seismic imaging and reservoir simulations.

A major feature was recently incorporated into AxRTM allowing for the propagation of seismic waves in anisotropic media. In seismic jargon, the anisotropy is widely known as tilted transverse isotropy (TTI) whereby the axis of symmetry for wave propagation can have arbitrary tilt which usually corresponds with the dip angle of geological substructures.

In the presence of anisotropy, isotropic migration will incorrectly image the position of seismic structures below dipping anisotropic bedding. TTI is an effective method for correctly imaging the position of geological substructures located below dipping anisotropic overburdens. Imaging with anisotropic modeling capabilities reduces the risk in oil-and-gas exploration drilling decisions. Incorporating anisotropic modeling into vanilla isotropic reverse time migration means that each migration now requires not only the velocity model, but also the Thomsen anisotropy parameters, dip angle and azimuth angle (in the case of 3D migrations).

The BP 2007 TTI model was migrated using both isotropic and TTI anisotropic propagation. Image (a), produced using isotropic AxRTM, shows a lateral shift in the position of the vertical column below the salt. Image (b), produced using TTI anisotropic AxRTM correctly positions the vertical column below the salt. Compare the images with the velocity model shown in image (c).

In conclusion, reverse time migration with TTI anisotropy correctly images the lateral position of vertical columns below dipping anisotropic overburdens. The ability to model with anisotropy provides another tool to confirm the correctness of a given velocity model with more confidence.

Image: (a)

Starting to work in a new team of high-performance computing developers and researchers is like going skydiving. To go skydiving, you first would like to go to skydiving school so that you can at least survive your first skydive. This is where those university years and CUDA courses taught by experienced people become very handy; by no means I am saying that you cannot learn CUDA, skydiving, or anything else on your own but I am saying that with proper training, new abilities can be learned safely and quickly. Once you have all of your training, you go out to jump off as many flying apparatus as you can find; keeping in mind that all you have is training and very little experience. With time, practice, and lots of patience you master your skills; regardless in the air or in front of your computer. The experience that you gather does not make you invulnerable to all the problems that can occur during a skydive or while developing software, but your experience teaches you how you can deal effectively with the many problems that can occur; in skydiving - line over malfunctions, line twists, horseshoe malfunctions, pilot chutes in tow, and in developing high performance software memory leaks, logic errors, race conditions, and problems parallelizing serial algorithms.

Working with the KTM and RTM team at Acceleware has been a great journey over the past few months, just like being at 13000 feet above the ground in a Twin Otter watching as the door slides open and all the noise and wind from the propellers invade the cabin of the plane. You walk to the door and look down to identify the dropzone among all the tiny things on the ground. A sign next to the door catches your eye reading "No Easy Way Down"; and you jump.

Remember to take advantage of all the years of experience that Acceleware has acquired by registering for our training courses, making use of our professional services, and integrating with our products for FDTD, RTM, Medical Imaging, and Matrix.

Thank you, blue skies, and safe programming.

Roberto S. Palacio Jaramillo (53 jumps and counting).

January 31: So I ended up in Zhuozhou City about 120km from Beijing this week as a guest of the Geophysical Research Institute (GRI). I have trouble following the connection chart, and it changes, but to make a long, convoluted story seem short: it seems like the path is China National Petroleum Company (CNPC), which is then connected to PetroChina, which is then or also connected to the Bureau of Geophysical Processing (BGP), which is then the parent of GRI. GRI is the seismic processing workhorse and I think they do most of the seismic processing for the Chinese oilfields/companies.

Seismic migration images the subsurface by propagating a wave into the Earth and listening for the reflections that return to the surface. Like any waves, the major properties of seismic waves are speed, frequency and wavelength. Seismic wave speed is largely determined by the type of rock it is traveling through, as well as how much pressure that rock is under. Wave speeds in rock typically vary from 3000 m/s to 6000 m/s, which is up to 20 times the speed of sound in air. The frequency of the wave is determined by a few factors, but mostly by the type of energy that is injected at the surface and how well that energy travels through the rock (the dispersive nature of the rock). At the end of the day, usable seismic frequencies recorded at the surface are usually between 20Hz and 60Hz.

The seismic wavelength is determined by the other two properties: wavelength = speed / frequency. And it is wavelength that determines the resolution possible to achieve in seismic imaging. The smallest subsurface Earth feature that is visible to the seismic wave is approximately 1/4 of the wavelength. Typical seismic wavelengths therefore vary from

to

Putting this into perspective, a large mountain face is on the order of 1000m high and is made up of features from 100s of meters in size down to centimeter-sized details. The imaging resolution determines which features are visible and which are lost. Of course, the complex layering and folding of the visible geology above ground are similar to the complex structures that seismic imaging is meant to reveal in the subsurface.

The Oil and Gas team just delivered on a project with Saudi Aramco, and it's my job to acknowledge (immortalize) the team's success by blogging about it. The project objective was to redevelop the customer's production Kirchhoff Time Migration codes (KTM, the "workhorse" algorithm in the seismic processing industry) to exploit heterogonous multi-core hardware, or in plain language, computer clusters equipped with GPUs and modern CPUs. Note that I didn't say port.

While some may be tempted to apply the "port" label to what we do in our Professional Services projects, it couldn't be more misrepresentative of our development process. While it's well known that algorithms must be substantially re-worked to be effective on parallel hardware (especially true on the GPU), what some organizations may not realize is that when end-to-end production software is migrated, a lot more than the core computational kernels have to be reconsidered. In particular, load balancing becomes more significant (drastically faster hardware elements can lead to drastically worse imbalances) and elaborate (nodes, cores , GPUs and network/disk/bus IO all play together simultaneously), Amdahl's Law rears its performance-damning head (small contributors to overall runtime suddenly become major bottlenecks) and maximally exploiting all available compute hardware becomes interesting (does the CPU have spare cycles that we can harness to take some of the load of the GPU, tangential from the usual notion of load balancing between like-resources?).

I have just gotten my first filtered seismic image using prediction error filters (PEFs), so I thought I should share this experience with our Acceleware blog readers. Coming from electromagnetic field modeling, I have quite a bit of experience with wave propagation through different types of materials. However, image filtering is a whole new (and interesting) world to me.

As most of the people doing seismic imaging know, reverse time migration (RTM) images obtained using a simple cross-correlation condition contain low frequency artifacts created by the unwanted cross-correlation of head waves and backscattered waves. These artifacts are clearly visible as the dark areas near the top of the BP model example shown in figure below.

The Acceleware SEG 2009 booth team is on their way back to Calgary. The message below reached us from Darren who manned the booth and also held a presentation about Industrial-Scale Reverse Time Migration on GPU Hardware at SEG:

There is a tremendous amount of interest in RTM at the SEG in Houston this year. Acceleware presented two talks on this topic, both were very well attended and generating a lot of interest, with questions during the talk and many people coming by the Acceleware booth afterwards to talk to us about RTM.

We are looking forward to their full report posted on our blog here once they are back!

The Acceleware Booth at SEG 2009 in Houston