The 2017 NVIDIA GPU Technology Conference (GTC) is a yearly conference produced and sponsored by NVIDA. The conference brings together thousands of scientists, technology experts, artists, virtual reality experts and media producers from all over the world to share and discuss their discoveries in the field of graphics technology.
NVIDIA's 9th annual GPU Conference closed recently with attendance at record levels. Approximately 8,000 artists, engineers, scientists and researchers came to the San Jose Convention Center to share their knowledge and ideas related to computer graphics, artificial intelligence, deep learning and virtual reality.
The conference, which featured an staggering array of events, was a unique learning experience for me. In addition to reporting back to Renderosity Magazine readers, I was able to advance my knowledge and understanding regarding deep learning and virtual reality and other related technologies. Plus, you meet all kinds of interesting people at the GTC who want to exchange ideas with you.
The main event of the GTC was Jensen Huang's keynote address which took place on Wednesday. As founder and CEO of NVIDIA, Mr. Huang has been a driving force in graphics technology for over a decade. His presentation is always an event and this year he did not disappoint. And what I personally like about his speeches is that he always relies on science and data to make his points. Rather than a charismatic presentation, he wows you with science and ideas.
Highlights from his keynote of interest to Renderosity readers include:
Project Holodeck, a new virtual reality conference mechanism that includes interactive physics and photo-realistic models,
AI for ray-tracing - advances the iterative design process to accurately predict final renderings by applying artificial intelligence to ray tracing
- NVIDIA Volta technology - cutting edge graphics technology that will be "the essential tool for deep learning going forward".
The Graphics Technology Conference
is where the future is invented
- Project Isaac - a virtual environment robot simulator that allows researchers to train robots for real-world accomplishments.
Toyota announces that they have selected NVIDIA's Drive PX for their autonomous cars.
NVIDIA makes Xavier DLA (deep learning technology) is now Open Source with a general release in September.
A complete recording of Jensen Huang's GTC 2017 Keynote at the NVIDIA GTC is available to view at the conference website: http://www.ustream.tv/gpu-technology-conference
Another highlight of my GTC was learning about the Health Industry's use of deep learning and AI. I was surprised to hear that there is a very great need for compiling and analyzing data by doctors and technicians in order to combat corrupt patient information and prevent misdiagnosis.
VR at the GTC
Virtual Reality is advancing at light speed. Perhaps because the promise (and profit) of quality VR is so immense in every field, NVIDIA is putting everything they have into its development.
Two trends in NVIDIA's VR technology stood out for me at the 2017 GTC.
- Multi User VR Experience - where more than one person shares the virtual space. This is especially important for multiplayer video games and conferencing. See NVIDIA's Project Holodeck for how this problem will be addressed.
- Haptic VR Experience - where you can actually touch and manipulate the objects you see in a virtual space. NVIDIA is helping to address this aspect of VR through its VRWorks technology.
Although the entertainment aspects of VR/AR get the most press, I believe that the more significant impact will be in the areas of medicine, data analysis, engineering, architecture and training of all kinds. Just standing at a booth watching someone practice a virtual knee surgery is a jaw-dropping experience. I predict that next year's GTC will feature more breakthrough VR technology in these areas.
We are moving towards a
convergence of AI, Augmented
Reality and Virtual Reality
My own experience in NVIDIA's superb VR Village was a combination of excitement and frustration. The actual VR experience is still so great. The suspension of disbelief is almost instant for me, so throwing Batman's batarang in the VR Batman Arkham is a blast, but the hardware feels like you are in a deep sea divers outfit: it's cumbersome and really requires a helper to keep you from tripping over cable. But I think researchers and developers are acutely aware of this problem and are working like mad to make better (and more affordable) hardware.
The 2017 GPU Technology Conference was extremely well organized. The convention center and show floor had excellent traffic flow so you aren't bumping into people all the time. Guidance and help is ubiquitous and easy to find. NVIDIA treats all attendees very well. It's a reflection of the larger company identity, I believe - treat others decently, share with them and, above all, inspire them to develop their own creative ideas.
It's exciting to witness the actual technology that was only theory in last year conference. Deep Learning technology has spread so rapidly to fields like medicine (a big eye-opener for me), engineering and urban planning (another surprise to me). The NVIDIA booth had an enlightening presentation of their Metropolis project which aims to help cities become "smart" with AI/Deep Learning tech.
We are approaching significant changes in how our data-focused world society will function in the future. Literally ever part of society will be affected with Deep Learning, AI and Virtual technologies. That NVIDIA saw this coming and is in the fore-front of development and technology sharing with events like world-wide GPU technology conferences is a testament to NVIDIA: a smart, innovative and inspiring company.
My sincere thanks to NVIDIA and, in particular, Gail Laguna, for sponsoring my trip to San Jose and the conference. Every year just gets better and better. Thank you for the trust you've placed in me.
- Everything begins and ends at the NVIDIA GPU Technology Conference website for 2017: http://www.gputechconf.com/
- I'd also recommend their YouTube channel for video coverage: https://www.youtube.com/user/nvidia
- The official NVIDIA blog also has a lot of GTC coverage, plus coverage of deep learning, virtual reality and artificial intelligence: https://blogs.nvidia.com/