New GPU Starter DevKit Provides Fast, Affordable and Easy On-ramp to Accelerated ComputingJune 4, 2014
NVIDIA GPU accelerators are used to tackle some of the toughest computational challenges. Challenges that include identifying new types of renewable energy. Or advanced machine learning tasks like object recognition.
Now, you can get your hands on this technology to help advance your research and application development work, with our GPU Starter DevKit with OpenACC.
It’s a fast and easy on-ramp to help you accelerate the performance of your C/C++/Fortran applications with GPUs. This “personal supercomputer” is a great starting point for developers looking to try GPU computing. No need to buy our build your own supercomputing cluster.
Powered by the NVIDIA Tesla K20 GPU accelerator and the PGI OpenACC compiler, the devkit makes it easy to begin accelerating your code using the directives-based OpenACC parallel programming standard.
If you’ve used OpenMP before, you’ll see that OpenACC is similar. First, identify the areas of the code to accelerate. Then provide “hints” to the compiler without modifying or adapting the underlying code. OpenACC does the hard work. This lets you achieve significant speed-ups – 2x, 5x, even 7x faster –with minimal effort.
OpenACC doesn’t just offer a straightforward path to parallel programming. It also helps you get powerful acceleration for large legacy codes. And it gives you the portability to run your code on various accelerators.
The Starter DevKit has everything you need for a fast and easy on-ramp to GPU computing , for under $5,000. That’s a significant discount off the cost of the workstation. Visit the GPU Starter DevKit website to learn more.
And, be sure to let us know about your experiences using OpenACC to accelerate your applications.