NVIDIA’s GTC: The near term Future of Advanced AI

Datamation

This post was originally published by Rob Enderle at Datamation

One of the major trends this decade is in AI development, and one of the areas that are moving the fastest is Automotive AI.  It has become such a significant trend that NVIDIA’s coming GTC conference has made a massive investment towards AI developers and innovators this year.  In general, AI has favored GPU technology for accelerators and has become a massive driver for its future potential revenue.

Automotive AI has also become one of the fastest areas of AI growth. Companies discovered that this technology, once mature, can be applied to other types of workloads because it is massively parallel and data-rich. The same base technology can be applied to ever broader transportation areas (including flying cars, planes, spacecraft,  and ships). Some of the benefiting areas include robotics, building management, security, and other uses where the AI must make rapid decisions based on complex data streams.

I’ve been looking through the agenda as I work on my attendance plans and figure I’d share the sessions I think will be the most interesting for those exploring the near-term future of advanced AI development.  Registration is free this year, so it’ll likely be worth the effort if you have the time and interest.

GTC Keynote

One thing you can never miss at an NVIDIA event is the keynote by their founder Jensen Huang.  He is better now that these events are virtual than in person because NVIDIA parses the keynote into segments. Jensen tends to do a complete overview of the event. Still, not everyone is interested in everything, so this allows you to focus on the things you are specifically interested in exploring.

If there is only one thing you attend at this event, this would be what I’d suggest as it typically provides a broad overview with decent depth in each of the areas where the event is focused. This session runs from 8:30 to 9:30 PDT on April 12.

Deep Learning Certificate

As I mentioned, the public event is free to attend, but they also have certification courses in Deep Learning for around $249, which could come in handy if you are a practitioner.

Deep Learning is where the AI trains itself, and it is replacing Machine Learning as the predominant way to both initially train an AI and update its training long term.  The training one of many training sessions runs from 6 AM to 2 PM PDT on April 13.

USD

One of the more exciting platforms NVIDIA has been developing is Omniverse, which is focused on developing game-level Artificial Reality environments.

USD or Universal Scene Description is a technology from Pixar Animation Studios used as a core element in NVIDIA’s Omniverse.  From 1-1:40 PDT on April 12, this USD session would be a good primer for those new to VR development or those who want to see how far VR environments and tools have advanced.

Tracks

One of the best ways to attend this show is to select a track and then decide which sessions are best in that Track.

For instance, they have an exciting track for Venture Capitalists that would likely be very useful for those who want to invest in technology and understand it better to increase their yield while reducing investment risks. Starting with the above keynote, this Track has things like AI Startups worldwide, Korean Startups behind the self-driving car revolution, AI-Powered Evolution of Agriculture, AI Startups  Disrupting Media, and Entertainment, and a variety of sessions focused on AI startups all over the world.

Another Track aggregates the AI startups and is more appropriate for those looking to use, partner, or work for emerging AI companies. This Track looks at global AI trends, AI Smart Factory acceleration, and many VCs’ sessions.

IT professionals have three distinct Tracks on Data Center/Cloud Infrastructure, one on Data Center Networking, and one focused on Virtualization. The Data Center/Cloud Infrastructure track looks the most interesting starting with a look at 10 AM on April 12 with a Vision for the Next Decade of Computing. Other sessions look at the AI-on-5G opportunity, AI Infrastructure, Accelerating Prototyping, and how best to deploy AI in the enterprise.  That last starts at 2 PM on the 12th

Automotive AI (or Where I Plan to Spend Most of My Time)

Now I mentioned that this year would have a substantial automotive component, and some of the sessions should be fascinating.

Audi will be talking about transforming the automotive future; Rimac will be talking about building the fastest electric car in the world (I’m hoping for a virtual ride); VW (which doesn’t stand for VotsWagon after all) will be talking about AI in commercial vehicles; Zoox will be presenting on Robotaxis; and Cruise will be talking about how to develop an advanced automotive AI from the ground up.

Two other exciting sessions are the one given by Aurora on the 12th at ten on AI drivers and at 11 the same day by Raquel Urtasun, a professor out of the University of Toronto on the future of Self-Driving vehicles. Both are on my personal shortlist.

Wrapping Up

I attend GTC every year and have even moderated or judged some sessions over the years. This year I’ll only be in the audience. I will likely prioritize the automotive sessions because much of the revolutionary advances in AI and AI development are likely to be. But I’ll likely work in as many sessions from the Autonomous Machines and Robotics Track as I can. I’m particularly interested in the Autonomous Race Drones session because, well, Autonomous Race Drones sound cool.

Oh, I should add, there will also be some exciting panels for those interested in the automobile’s future. GM and Ford, both looking to stand out in this increasingly electric autonomous driving world, will have dedicated panels, as will Lyft And Uber.  Fortunately, most of these are either on-demand or on the 13th.

Anyway, this event is a great place to explore our AI, Robotic, and Autonomous future if you have the time. Oh, and given many, if not most of these sessions can also be seen on-demand, you can make up your schedule if you want.

Spread the word

This post was originally published by Rob Enderle at Datamation

Related posts