NVIDIA Discuss the Rise of Deep Learning
Andy Ruffell / 3 years ago
GTC 2015: For the most part, GTC has already been a huge event for NVIDIA with the unveiling of the new TITAN X graphics card to the market but while this is great for the average consumer, NVIDIA have a lot more to offer than just their GeForce segment and one aspect of this is deep learning, and actually is a segment from within their company that crosses over into GeForce and other more industrial sectors that NVIDIA have to offer.
Jen-Hsun Huang (CEO and co-founder of NVIDIA) took to the stage at NVIDIA GTC 2015 to talk through the theme of the event this year and deep learning was at the core.
You can see part 1 of the video here:
As you can see from the video, Jen-Hsun went through the various milestones of the company last year including GeForce, Shield, Tegra, enterprise graphics and of course what NVIDIA are doing from within the car industry.
A big highlight was what the new TITAN X GPU can do in comparison to other offerings on the market. A comparison of a 16-core Xeon CPU, original TITAN, TITAN Black and the newest TITAN X were shown training AlexNet and while the TITAN and TITAN Black managed to improve on what the Xeon CPU could do by cutting down the time it takes to train AlexNet, the new TITAN X dramatically improved even on that. This sees the AlexNet training take just 3 days opposed to the 43 days that the Xeon takes showing that NVIDIA are focussing on more than just gaming.