All the excitement from SuperComputing10 may be behind us, but much of what we learned remains top of mind. We have enjoyed our beignets, strolled the French Quarter, toured the Garden District and maybe even tried a couple of Hurricane Punches. Now, it’s back to solving tomorrow’s problems which seem to be more within our grasp now than before the conference.
As we started exploring in the post with Microsoft’s Director of Technical Computing Bill Hamilton, the strides that Microsoft has made in their standard development tools is impressive. Visual Studio 2010 is specifically designed to aid developers writing, testing, and debugging parallel code. Working closely with partners from Nvidia, this also includes GPU development using Nvidia’s parallel Nsight for Visual Studio 2010. With the paradigm shift to multi-core computing, cluster farms reaching 1000s of nodes and the emergence of cloud computing, the need for this evolution is obvious. For example, a new Task Parallel Library in .NET allows you to create for loops that run in parallel by simply replacing the for construct with Parallel.For. They’ve also included a similar parallel patterns library for C++ developers and parallel profiling and debugging tools that give a visual representation of what each is thread is doing, where blocking or other inefficiencies may be occurring, so the developer can easily see how to optimize their code.
I believe this is a significant breakthrough and that the application of this technology is bound to be pervasive. It finds use in multi-core engines and in cloud applications. While it is argued that HPC and cloud are at opposite ends of the spectrum, the reality is that parallelism is at the heart of both operations. Both require function segmentation with complete definition to facilitate scalability. With the advent of cloud computing and its ”pay as you use” business paradigm, the availability of HPC clusters is becoming a reality for all of the HPC community, not just the elite few. This has the potential of bringing another 55 million users online within the next few years.
Bill Hamilton referenced Bill Gates’ desire to put a computer on every desk when he said that it is the intent of Microsoft Technical Computing to put supercomputing capability into the hands of every scientist, engineer, or analyst. The early adopters of cloud HPC will be the embarrassingly parallel functions that do not rely on high speed interconnectivity among individual nodes according to Hamilton. As users become more accustomed to cloud HPC and the technology continues to evolve and improve, the need for MPI applications in the cloud will follow suit. Modeling will become a more pervasive and normal activity in the development and manufacturing process.
With this vision, I can easily see HPC becoming a larger part of mainstream corporate America much sooner than you might think. With HPC cloud, response times from simulations are measured in hours rather than days. Having the ability to model, test and refine development and subsequent manufacturing processes using HPC in a cost effective manner in the cloud will certainly produce better products in a quicker timeframe. This offers a true competitive advantage.
Another thought provoking topic is as HPC becomes more pervasive, we can anticipate more technical breakthroughs resulting in a plethora of exciting new products. The future is certainly much more promising as we uncover other enriching opportunities and applications for High Performance Computing. Isn’t it always the case with each edition of the SuperComputing Conference that dreams shape into reality and previously unknown vistas form on the horizon! Thank you , New Orleans. Seattle and SC11 are less than a year away!
{ 0 comments }
