The Linux Journal Magazine is currently running an article in their November issue featuring an introduction to Parallel Programming using the NVIDIA CUDA technology.
The article was written by Algorithmia’s director Alejandro Segovia for RealityFrontier and kindly sponsored by CoroWare Inc.
The Linux Journal introduces the article in their latest issue:
“If all that seems a bit too fluffy for your technology taste, you might just love Alejandro Segovia’s in-depth piece on parallel computing with NVIDIA’s CUDA technology. Video cards are great for gaming, but it’s amazing how powerful they can be when you use them for straight up mathematical processing. Alejandro shows how to take advantage of the little powerhouse sitting inside your computer case.”
As Raphael states in his post at the RealityFrontier website, a big thank you to CoroWare for the opportunity to write the article and to the Linux Journal for publishing it!
Did you know Algorithmia: Visualization and Acceleration has a Twitter account?
Not only can you stay current with our latest announcements, but also learn what’s going on in the world of Parallel Programming with GPGPU and the latest advancements in Graphic application Development.
To start following us, just visit http://twitter.com/algorithmiadev/ or click the “Follow me” box floating at the right of this page.
Let’s keep in touch!
Our industry is currently undergoing a radical shift toward parallel processing and mobile computing. RealityFrontier, an affiliate of Algorithmia and its partners, was created as a direct response to our changing partners’ and customers’ needs in this new world.
Dedicated to delivering services and sales-enabling solutions in three converging fields of expertise: OSS platforms, connected systems and telepresence devices, RealityFrontier aims to build tools for product designers and R&D centers at the crossroad between telepresence, augmented reality and mobile robotics.
At Algorithmia, we provide know-how, technology and consultancy and development services for RealityFrontier, our first subsidiary in the United States.
Come meet us at the NVIDIA GPU Technology Conference in San Jose, California, and find out what we have been working on and how we can help you deliver cost-effective parallel solutions to your customers, today.
Edit: NVIDIA GTC 2010 is over. It was an incredible show. Thanks to all that came by our stand and we’ll see you next year!
Algorithmia was proud to participate in the iOS Hackathon (part of the OpenDay) organized by Globant LLC, one of the biggest outsourcing companies in the South Cone.
During the hackathon, we developed a custom iPad application that leverages the Video Card of the device to render the Julia-Fatou and Mandelbrot sets. The following image depicts the Demo running on Apple’s iPad Simulator:
Presenting at the iOS Hackathon
The implementation was performed using OpenGL ES Frament Shaders to leverage the iPad’s GPU as a highly parallel coprocessor, capable of evaluating the fractals at the same time for every pixel displayed on the device’s screen.
After developing our Demo, we were invited to explain the concept and technologies used to the Audience using an actual iPad device.
You can see some pictures taken at the Event below.
iFractals app running on Apple's iPad Simulator.
The Julia-Fatou Set, as rendered on an actual iPad
The Mandelbrot set, as rendered on an actual iPad.
We thank Globant for the opportunity to present our Demo as part of their conference.
Algorithmia was proud to be invited to the Parallel Processing Symposium at the Mathematics and Statistics Institute of one of the most important universities in our Country.
At the symposium we conducted a short presentation on our experience designing and developing applications that leverage Video Cards as highly parallel coprocessors. Typical applications suitable for implementing on the Video Card include: image and media processing, signal processing, physics simulation, computational finance and computational biology.
GPGPU, named after “General Purpose Graphics Processing Unit”, is a new parallel programming paradigm that allows developers to speedup their algorithms dramatically. Heavily backed by NVIDIA, GPGPU (and CUDA in particular) is gaining momentum because of two reasons: first because of the incredible speedups it allows a developer to obtain when comparing a parallel algorithm to its sequential counterpart; second: because all it requires to work is inexpensive commodity hardware. In our experience we have developed several GPGPU applications, some of which achieved up to 80 times speedup, depending on the algorithm being implemented.
At Algorithmia we provide GPGPU development services on NVIDIA CUDA, the pioneer technology on this field, as well as on OpenCL, the Khronos Group maintained standard for general purpose programming on the GPU.
Let your customers benefit from the power of GPGPU, contact us today.
