Click here, go fullscreen and enjoy the 360 panorama from Curiosity :)
Last Updated on Wednesday, 15 August 2012 06:13
NASA Curiosity Rover Successfully Landed on Mars
PASADENA, Calif. -- NASA's most advanced Mars rover Curiosity has landed on the Red Planet. The one-ton rover, hanging by ropes from a rocket backpack, touched down onto Mars Sunday to end a 36-week flight and begin a two-year investigation.
The Mars Science Laboratory (MSL) spacecraft that carried Curiosity succeeded in every step of the most complex landing ever attempted on Mars, including the final severing of the bridle cords and flyaway maneuver of the rocket backpack.
"Today, the wheels of Curiosity have begun to blaze the trail for human footprints on Mars. Curiosity, the most sophisticated rover ever built, is now on the surface of the Red Planet, where it will seek to answer age-old questions about whether life ever existed on Mars -- or if the planet can sustain life in the future," said NASA Administrator Charles Bolden. "This is an amazing achievement, made possible by a team of scientists and engineers from around the world and led by the extraordinary men and women of NASA and our Jet Propulsion Laboratory. President Obama has laid out a bold vision for sending humans to Mars in the mid-2030's, and today's landing marks a significant step toward achieving this goal."
Curiosity landed at 10:32 p.m. Aug. 5, PDT, (1:32 a.m. EDT Aug. 6) near the foot of a mountain three miles tall and 96 miles in diameter inside Gale Crater. During a nearly two-year prime mission, the rover will investigate whether the region ever offered conditions favorable for microbial life.
7 Minutes of Terror: Curiosity Rover's Risky Mars Landing
Last Updated on Thursday, 02 August 2012 01:53
Multi-GPU back-propagation through time
We have implemented a multi-GPU version of the back-propagation through time algorithm for the training of MTRNN (multiple time-scales recurrent neural network). We have also tested the code on 4xGPU setups and the preliminary result show excellent scaling. From the graph you can see that when we use 2 GPUs we get almost perfect 2x speedup over single GPU; when we use 3 GPUs we get nearly 3x speed, etc.
We are now starting to develop scalable genetic algorithms for large genotypes making good use of GPU-based fitness evaluations.
All these developments will be added to Aquila 2.0. New developers are welcome, contact me if you are interested.
GPUs in Cognitive Robotics
My first few weeks at NVIDIA have been a great experience. I have improved couple of modules in Aquila and run a benchmark comparison of their performance on GTX590 and Dual-socket Intel Xeon X5550. This allowed me better see the real benefit of using GPUs over CPUs for our biologically inspired models used for our experiments. I have also learned few optimisation techniques that I applied to these modules, which resulted in additional 2.5x speedup over previous GPU implementations.
I am starting to work on new stuff from today, which involves developing scalable genetic algorithms and what is more exciting is what I plan to use them for. Later on that in few weeks :)
Last Updated on Friday, 13 July 2012 23:55
NVIDIA GTC Express on Aquila - update
The NVIDIA GTX Express on Aquila went really well. I have put the slides on SilideShare together with the videos.
It seems they spread really well on the internet, enjoy!
Last Updated on Friday, 22 June 2012 01:44
NVIDIA GTC Express Webminar on Aquila and Cognitive Robotics
Accelerating cognitive robotics research with GPUs
Aquila, its history and modules
There will also be lots of pictures and videos. The talk will also be available offline for later view. If you would like to attend you can register here.
Last Updated on Tuesday, 19 June 2012 01:31
Moving to California to do Research at NVIDIA
It has been a while since I last updated the website with something interesting. The single reason was that I was waiting for the right time to publish this little article with some of the latest news from my life.
I remember when I started doing research for my PhD, I had no interest at all in working for companies. As time went by, I did some work with CUDA, Aquila and iCub that got published on NVIDIA website. I said to myself, "Ok, it would be really inspiring to work for NVIDIA one day." I was really fascinated by the computing revolution NVIDIA triggered several years ago when they designed a GPU processor that has now become a disruptive technology that is changing the way we think about computing. This paradigmatic change in computing caused a revolution in so many different domains and NVIDIA's CUDA Community Showcase clearly demonstrates that. Amazing speed-ups (when using GPU vs CPU) were reported from different institutions worldwide that used GPUs for Computational Structural Mechanics, Bio-Informatics and Life Sciences, Medical Imaging, Weather and Space, Data Mining and Analytics, Imaging and Computer Vision, Computational Finance, Computational Fluid Dynamics, Electromagnetics and Electrodynamics, Molecular Dynamics and also Robotics.
One morning, several months ago, I checked my inbox and found an offer from NVIDIA saying that they would be interested in having me involved in their research team. I was extremely excited and accepted the offer, which was followed by three separate and nearly one hour long interviews by different NVIDIA researchers.
A few days after the interviews, I received my contract and instructions for applying for VISA. NVIDIA made getting a VISA as easy as possible but despite of that, there were lots of things I had to do in the last moment. For example, the US embassy in London had no appointment for me that would be soon enough. This meant I had to travel to US consulate in Northern Ireland instead. I thought I would never get the VISA on time because many people experience long delays in processing their applications and I had only about one week until I had to flight to California!!! I arrived to the consulate with all the documents I had, even those I did not need. I made an additional effort and printed many other documents I thought could be useful. Basically, I did not want to risk anything. The interview at the consulate took only about 5 minutes. The lady there asked me about my research and at told me that I have made it a long way from being a fish processor to where I am at the moment. Then she told me that I should get the VISA by the end of the week (2days), which was amazing!
Now I am in Santa Clara, California and I love it here. The weather is amazing, the apartment is awesome, too. It's called Old Orchard consisting of many apartments that look like little cottages. I feel like living in a forest because there are so many trees and plants here. Also, it's a great place to chill out since there are two swimming pools here, jacuzzi and gym as well. I can easily travel to San jose or San Francisco by train. NVIDIA is just one mile from here. I brought my aggressive rollerblades with me and plan to skate to NVIDIA and back. I am really excited and cannot wait to start working! I have a really really interesting project that I am going to be working on.
Thanks for reading and keep an eye on here for the follow-up stories :)