As oil and gas exploration is forced to deal with increasingly complex geologies, many of the established imaging methods (for example Kirchhoff depth) can struggle to produce accurate images. Reverse Time Migration (RTM) has become firmly established in recent years as the preferred imaging option for complex geologies. Its simplicity and superior imaging quality stem from the use of a full solution 2-way wave equation. This makes no approximations limiting the direction in which seismic energy can travel.

The high computational cost required to run RTM has traditionally restricted the use of the technology to companies with 1000’s of expensive, traditional processing cores available.

Acceleware’s AxRTM™ is the industry’s solution to fully support compute GPUs (in addition to traditional CPU clusters) to dramatically improve the economic cost of RTM and make it a viable solution for a wide range of companies, large or small.

Join Dr. Ray McGarry, Acceleware's Head of Seismic Research for an informative introduction to RTM.

SuperComputing 10, the last of Acceleware’s three big fall tradeshows took place last week in New Orleans. The show also coincided with our CUDA training class in Tampa, Florida, resulting in a number of Accelewarians who were able to enjoy being in a warm climate as the first big deep freeze settled over our home city of Calgary.

Beneficially for the business (but a bit disappointing for us personally) is that we had very little time outside to enjoy the sunshine as the show was very busy in New Orleans and our CUDA training classes are always a full day of learning.
Acceleware was located on the show floor close to one of our key partners, NVIDIA, where the traffic was brisk throughout all four days of the show. Visitors to our booth were able to try out the Acceleware performance calculator to generate numbers that reveal how much faster their codes could run after they have been ‘Accelerized,’ and earned a free Starbucks coupon for their effort. Also at the booth we hosted a strong contingent of existing customers and new prospects from the oil and gas industry as the location of the show in New Orleans helped to bring out the energy crowd. There were also many people gathering information about our Electromagnetic solvers and another big hit was information on the advanced CUDA/OpenCL training classes. With the proliferation of NVIDIA Tesla GPUs throughout the show floor this year, the demand to learn how to program GPUs efficiently continues to grow.

SEMCAD workshop hosted by Dymstec and sponsored by SPEAG and Acceleware, October 2010

Last week I traveled to Seoul, Korea to take part in various customer events organized by NVIDIA and Dymstec, a distributor of Acceleware technology in Korea. The trip was organized along with one of our partners in EM simulation, SPEAG from Zurich. It was my first actual visit to Korea, as many times before I have transited through the airports at Kimpo and Incheon, but never have been able to experience the non-airport version of Korea. I was highly impressed by what Seoul offers in terms of a dynamic and technically advanced city. For example, the GPS systems in cars are equipped with 3D fly-through views and receive multiple inputs of information from traffic to ETAs. Another interesting feature of the GPS systems is the ability to watch live TV though I was wondering about the sanity of that as several cars passed me the baseball game playing on the dash. From a general economic activity point of view, I must have seen at least 200 building cranes during the six day visit creating new forests of office towers and condo buildings.

I know that Acceleware has not posted any blog entries since the end of June…..let me try to explain why. In this part of the world, summer is short. This year our summer weather has created a new definition for short as May and June were cool and rainy, and July was not much better. However during our summer months we spend as much time outside as possible soaking up the long summer days and the blogging production falls off accordingly. Now that September is here, it is time to resume normal programming. Here is a quick round-up the Acceleware fall schedule of activities.

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.

The fall has always been the busy season in the high-tech industry for conferences and tradeshows and despite the recession, 2009 is continuing the tradition. Time to pack the bags, get on airplanes and get away from the office to spend high quality time meeting customers, partners and acquaintances from the industry.

Acceleware started the season last week at NVIDIA’s GPU Tech Conference. This conference broke away from the NVISION conference that incorporated all things NVIDIA including gamer parties, all night Rock Band marathons and sprouted into its own business focused event. With this more narrow focus, Acceleware even became a sponsor of the event, I think mainly for better access to drink tickets. According to NVIDIA, the conference sold out 2 weeks before opening and the atmosphere was definitely buzzing with lots of cool new technologies to show off.

Followers of the HPC world may have noticed the announcement this week that our fellow Parallel Computing company (and fellow Canucks BTW) RapidMind was purchased by Intel. Of course this generated a flurry of calls into the Acceleware World Headquarters by colleagues, partners and of course investors asking what does this mean for Acceleware?

First off, congratulations to RapidMind for joining an organization as great as Intel and we wish them success in the future. Over the years, we have been asked countless times if Acceleware competed with RapidMind, and the answer has always been “not really.” It is a slightly ‘hedgy’ answer due to similarities in the markets we collectively address and how both companies talk about parallel computing, use accelerators and often show up at the same conferences.