Back to top

NVIDIA GTC Lessons in 2 Minutes from Professor Huang

Read MoreHide Full Article

This week held the event that AI fans wait all year for: The NVIDIA (NVDA - Free Report) GPU Tech Conference.

And CEO Jensen Huang did not disappoint them. From gaming graphics to scientific research applications, the ability to process more data, at higher speeds, seems to know no bounds for his engineering teams.

Here was reaction from an excellent semiconductor analyst team led by William Stein at Truist...

NVIDIA price target raised to $360 from $257 at Truist: Analyst William Stein wrote that the company's software investments are "becoming more prominent" as management introduced new AI models and software tools at its GTC event.

Stein noted that while investors did not get the RTX40 series or other obvious model upside drivers, based on feedback from industry contacts, he is updating his model for higher growth in NVIDIA's Datacenter, Professional Visualization, and Gaming segments.

While there weren't specific gaming platform upgrades that some were looking for, it's worth remembering that innovations in gaming graphics are what drive new research and ideas across the NVIDIA "full stack." Here's how I explained that over three years ago...

NVIDIA Gaming Drives the Deep Learning-AI Revolution

Listen to 5 Big Ideas in 2 Minutes

In the video that accompanies this article, I show a 2-minute clip from Jensen Huang's 90-minute keynote speech on Tuesday morning.

And those 2 minutes are packed with big ideas that explain the dominance of the CUDA platform. I'll quickly list them here, but do yourself a favor and watch my video excerpt to hear the professor himself.

1) CUDA (Compute Unified Device Architecture) is an ever-expanding hardware + software stack that harbors growing libraries of algorithms to mine and model data and accelerate applications.

2) Software upgrades can "upgrade" older chip sets: end users experience "speed-ups" through the life of the product as application performance can increase by many x-factors on the same chip over several years.

3) GPU workbench stacks for all industries and scientific pursuits: wherever there's a data, research, or automation need and for countless scientific pursuits where large data sets must be modeled and simulated in real time.

The developer and engineering tools for the ISAAC Robotics platform are one of the most tangible mind-blowers you'll be experiencing in the next few years.

4) SDKs are the way: Over 150 software developer kits make using CUDA tools accessible, repeatable, and amazing for researchers across gaming and design, life and earth sciences, quantum computing, AI, cybersecurity, 5G, and robotics as discussed above.

It's also worth noting that the SDK approach is responsible for much of NVIDIA's success because Jensen & Co. essentially get to recruit an army of raving fans (the developers and engineers) who learn to build and research within the CUDA ecosystem and never see the need to leave.

5) Design Automation and Simulation are also the way: With partners like Cadence Design Systems and Synopsis, NVIDIA is constantly building the next iterations of its chips and also helping those companies create design automation and simulation for their industrial and scientific customers.

You'll have to watch the Jensen clip to visualize one application of these simulation efforts.

And if you decide to watch more of the GTC keynote, you can learn about how NVIDIA's Omniverse will be building worlds for creators, scientists, and engineers that will make Metabook look like a kindergarten playroom.

EVs and ADAS Are Taking Off

Last week, I profiled the latest and greatest innovation in the electric vehicle space from the oldest of old-school car makers...

Ford F-100 Eluminator Lights It Up: EV-Crate Motors Produce 480-HP

That video and article were fun to create because it's a completely new, back-to-the-future idea from Jim Farley & Co. where traditional gas gearheads and other custom propulsion enthusiasts are about to build some amazing EVs.

Further "innovation validation" of my thesis that the F-150 Lightning might just crush the Tesla (TSLA - Free Report) Cybertruck in a head-to-head sales battle.

While we enjoy open profits of nearly 50% in Ford shares near $20, another EV rocket blasted off today. Rivian (RIVN - Free Report) debuted on the Nasdaq at a valuation nearly twice what was expected only a week ago when we thought they'd be selling 135 million shares for around $60.

This morning's official IPO pricing was for 153 million class A common shares at $78 per share, to raise $11.9 billion in gross proceeds. Well, the EV-hungry mob sent shares as high as $119 at the 1pm ET launch.

And that's a valuation probably north of $100 billion. Good to know that Ford owns a chunk of Rivian too!

Meanwhile in ADAS (advanced driver assistance systems) news this week, my group of TAZR Traders was thrilled again to see our Luminar Technologies (LAZR - Free Report) get a big nod from... NVIDIA!

Luminar's LIDAR (light-detection and ranging) system was selected for the NVIDIA DRIVE Hyperion Autonomous Vehicle Reference Platform.

You can learn more about this in Thursday's Bull of the Day article on LAZR, where I explain why NVIDIA decided to make the shift to partnering with other sensor and software makers, whereas previously they left car makers "to their own devices."

In the video I also share the story of how Qualcomm (QCOM - Free Report) is slowly entering the ADAS game too with their recent $4.8 billion purchase of the Swedish company Veoneer.

Win a Call with Cooker Drawing

If you came here from the YouTube link, I have your code/instructions to enter the "Call with Cooker" drawing: #TechSuperCycle

Just go to Twitter and do the following...

1. Follow me @KevinBCook

2. ReTweet my pinned Tweet with the NVIDIA graphic using the code #TechSuperCycle

See you on Twitter and let's see if you're the next guest on a Call with Cooker!

Disclosure: I own shares of NVDA, LAZR, QCOM and Ford for the Zacks TAZR Trader portfolio.

Published in