What is the GTC?
The Graphics Technology Conference (GTC) began in 2009 as a way of bringing together researchers and developers of graphics technology which had previously been confined to individual companies and university computer departments. NVIDIA, who hosts the now 6 year old conference, also found it a convenient way to promote and share their newest technology. Year by year, the conference has developed along with incredibly strong growth in games and professional graphics technology for industry and science.
Approximately 6,000 people attended the 2015 GTC located in San Jose, the heart of silicone valley. Modelled after the larger SIGGRAPH conference, Nvidia has reworked and improved the experience, while focusing on technology and ideas, rather than entertainment and VFX. 500 + sessions and panels covering everything from Astrophysics to web-accelleration are presented along with several keynote addresses, press conferences and, most importantly, informal meetings and parties with like minded people. In fact, one of the most enjoyable aspects of the entire conference were the unplanned conversations one can have at the conference.
Every conference is defined by its keynote address. In the case of the GTC, there were three major keynotes: one by NVIDIA founder Jen-Hsun Huang, one a day later by Google tech-rockstar, Jeff Dean, and a third one by Andrew Ng. Unfortunately, I was unable to attend Mr. Ng's keynote address, so I'll focus on the first two I mentioned. Both Mr. Huang's and Mr. Dean's speeches were impressive displays of personal expression, ideas and, most importantly, passion. These important thinkers/leaders are passionate about ideas and they conveyed that passion very clearly.
The two major threads in each keynote (and in the conference as a whole), were “Deep Learning” and “Autonomous Cars”. Of course, new products were presented by NVIDIA founder Jen-Hsun (more on this below), but these two ideas resonated with the speakers and made their presentations very exciting.
NVIDIA has made all three keynotes available online in their entirety. You can watch Mr. Huang’s keynote here and Mr. Dean’s keynote here, and Mr. Ng's keynote here. I highly recommend them as they will surprise and perhaps inspire you.
The idea of Deep Learning is a new one to me, although computer graphics professionals have been aware of it for a decade. Previously called “Machine Learning”, the new appellation is more poetic and perhaps more precise. Essentially, the term refers to a machine’s ability to learn from past experience. It also is particularly focused on a machine’s ability to organize and define very, very large data sets.
Jeff Dean, from Google, spoke extensively on Deep Learning using image recognition as he focus. In an amazing development, machines have for the first time surpassed a human in the ability to successfully identify and define a large assortment of images. It’s obvious why Google would be interesting in this development, but effective image recognition is also an essential element in developing autonomous cars.
“Predicting cancer is about predicting the growth of cancer cells. Deep learning can analyse the patterns by itself and do an even better job.” -Nvidia CEO, Jen-Hsun Huang
At the end of Mr. Huang’s keynote address he invited Elon Musk, the CEO of Tesla motors, onstage to dialogue about autonomous cars. Tesla has led the industry in this area and Mr. Musk had much to say about how deep learning and advances in graphics technology have accellerated the advent of driverless, or “autonomous”, automobiles. His notion of the future car being a “computer on wheels” is intriguing and much closer than anyone (including this reporter) has imagined.
You can watch their entire conversationon NVIDIA’s Youtube channel.
Of course, NVIDIA uses the GTC conference as a way to promote their products and technology, but what is surprising is how balanced the conference is in terms of sheer creative ideas related to every conceivable type of computer graphics technology vs NVIDIA's need to sell you something. Even at a cocktail party, people are debating and sharing ideas on graphics cards, programming and the practical use of current graphics technology. This speaks well for NVIDIA's corporate culture as it shows they are just as interested in good ideas as everyone else. It also shows they company has class.
I spent a good hour with two NVIDIA technology professionals, Gail Laguna and Sean Kilbride, who carefully took me through the new products and new technology the company has been developing for the last year. I was particularly impressed with NVIDIA’s roadmap for their physically-based renderer, Iray. The wider availability and technical improvements of this excellent renderer will bring iray to a much wider audience in the future.
NVIDIA also announced their newest high-end gaming card, the Titan X, and a new professional graphics card, the Quadro M6000, both of which are the best cards you can buy in their fields. Along with the Quadro and Iray, NVIDIA laid out a roadmap leading to the PASCAL GPUarchitecture, which should come in 2016. At some point these cards are going to approach the speed of light. It all makes for happy gamers, developers, game makers, scientists and researchers.
NVIDIA has done a remarkable job of creating and hosting the 2015 GTC conference. San Jose is a beautiful city and the McEnery Conference center is an airy, well-designed building that was very easy to get around in. All events I attended were on time and carefully managed. I was also impressed with how NVIDIA dealt with the 180+ press and analysts present. I found myself deep in conversation with press people and attendees practically every day.
The spirit of the GTC with its breathless presenters, passionate speeches, jaw-dropping Exhibits and truly revolutionary technology, is unique in my experience. And NVIDIA has done an amazing job of making practically every event available online. See this link to screen them.
The 2015 GTC conference was a delight. I came away inspired and very interested in pursuing further research, particularly on Deep Learning. If you have any chance of attending next year’s conference in early April, 2016, I can’t recommend it more highly. You will love it no matter what level of knowledge you have of graphics technology.
My sincere thanks to Shannon Deoul at Raz Public Relationsfor her help in co-ordinating my trip and to NVIDIA for inviting me to this remarkable conference. Gail Laguna and Sean Kilbride were excellent presenters and overcame logistical problems to help me understand in more practical detail what NVIDIA is currently doing with computer graphics. I’d also like to thank the many people I spoke with in the halls and meeting rooms. These spontaneous conversation were so enjoyable. The generosity of the people I met at the GTC was appreciated.
Note: You can discover practically everything you need to know at NVIDIA’s GTC website. I’ve also set up a Flickr photo album of my GTC experiences (see above). Information on NVIDIA’s graphics cards and new graphics technology can be found here. A complete list of GTC sessions with detailed descriptions can be found here.