Back to top

AI Dreams and Reality: Investing in Advanced Technology

Read MoreHide Full Article

McKinsey Global Institute has estimated that AI technologies across 19 industries could create between $3.5 trillion and $5.8 trillion in value annually.

From their 2018 report, Notes from the AI frontier: Applications and value of deep learning, the authors wrote "This constitutes about 40 percent of the overall $9.5 trillion to $15.4 trillion annual impact that could potentially be enabled by all analytical techniques."

In the video that accompanies this article, I show a graphic that highlights the 19 industries impacted by true AI, not just data analysis and statistical inference. You won't be surprised to see that Retail tops the charts in terms of exponential value creation from fully engaging customer tastes and buying habits with big data.

I used this 2018 study on purpose to highlight what the expert view was before the coronavirus pandemic, which ignited a decade's worth of digital transformation in just 10 months.

Ever Seen Deep Learning in Action?

Today I presented a 45-minute deep dive into deep learning for fellow Zacks strategists and I tried to condense some of the best stuff down into under 15-minutes for you here.

One of the highlights is a 3-minute video clip from Simplilearn, the world’s #1 Online Bootcamp focused on helping people acquire the skills they need to thrive in the digital economy.

It is the best visual example of machine learning and how a neural network is trained that I have found on the web, explaining nodes, weighting, bias, activation function, forward propagation, and back-propagation.

While the dreams of futurist Ray Kurzweil about the technological Singularity may seem very remote when you watch how machine learning actually works, I think its important for everyone to understand the mechanics so that the veil of mystery is removed when someone uses the AI acronym.

There are many levels of sophistication and it helps to know what someone might be trying to sell you as a solution to your data challenges. For me, one of the best ways to keep up with how different industries are applying machine learning tools is to be a frequent visitor to the NVIDIA (NVDA - Free Report) Newsroom, where every day brings new stories of specific business applications.

Tesla Builds Giant Supercomputer Using NVIDIA GPUs

On Tuesday, we learned of Tesla (TSLA - Free Report) unveiling the in-house supercomputer the EV maker is using to train deep neural networks for Autopilot and self-driving capabilities. The cluster uses 720 nodes of 8x NVIDIA A100 Tensor Core GPUs (5,760 GPUs total) to achieve an industry-leading 1.8 exaflops of performance.

“This is a really incredible supercomputer,” senior director of AI at Tesla Andrej Karpathy said. “I actually believe that in terms of flops, this is roughly the No. 5 supercomputer in the world.”

In the video, I explain the evolution of NVIDIA super-computer architecture from a handful of "petaflops" (one quadrillion floating-point operations per second) to exaflops, or one quintillion FLOPS, equal to 1,000 petaflops.

The key to NVIDIA's dominance as the premier provider of the biggest engines for exascale computing is their ability to cram 21 billion transistors onto a GPU card that is smaller than a shoebox and then stack and link thousands of these GPUs in high-grade servers.

Don't Tell Jensen That AI Doesn't Scale

In the video, I also share more McKinsey intel about surveys they've done finding that AI is difficult to scale for some industries.

But also yesterday, the NVIDIA Newsroom had two posts about helping enterprises deploy "instant AI infrastructure" and "edge AI services." That's the power of Jensen & Co. when they combine top hardware with software built for custom machine learning applications.

And last week, I found this behind-the-scenes industrial story especially intriguing...

NVIDIA Unveils Jetson AGX Xavier Industrial Module

New ruggedized module is engineered to bring AI to harsh, safety-critical environments (from a June 15 NVIDIA blog release):

From factories and farms to refineries and construction sites, the world is full of places that are hot, dirty, noisy, potentially dangerous — and critical to keeping industry humming.

These places all need inspection and maintenance alongside their everyday operations, but, given safety concerns and working conditions, it’s not always best to send in humans.

Robotics and automation are increasingly used in manufacturing, agriculture, construction, energy, government and other industries, but many companies have struggled to incorporate the benefits of AI and deep learning in the most demanding applications.

With the new NVIDIA Jetson AGX Xavier Industrial module, NVIDIA is making it possible to deploy AI at the edge in harsh environments where safety and reliability are critical priorities.

(end of NVIDIA blog release)

AI Will Be the Engineer's Workbench

In today's video, I also talk about the new paper published by Alphabet (GOOGL - Free Report) scientists: A graph placement methodology for fast chip design.

Essentially, AI can now design semiconductors in hours that used to take engineering teams months of testing. Here's the opening of the abstract...

"Chip floorplanning is the engineering task of designing the physical layout of a computer chip. Despite five decades of research, chip floorplanning has defied automation, requiring months of intense effort by physical design engineers to produce manufacturable layouts. Here we present a deep reinforcement learning approach to chip floorplanning."

When I first saw this story, I was worried it might have a negative impact for a company I own in the EDA/SDE chip design space, Cadence Design Systems (CDNS - Free Report) . Those acronyms stand for electronic design automation and system design enablement, where software is already integral to semiconductor fabrication.

And speaking of Google, who allows so many tools and so much automation to be accessed through the cloud, Advanced Micro Devices (AMD - Free Report) was just chosen as a new heavyweight provider of datacenter chips.

Meanwhile, all eyes are on regulatory approval of the proposed $35 billion AMD acquisition of FPGA maker Xilinx. On April 7, both boards approved the deal and they described the marriage thus in the press release...

The acquisition will bring together two industry leaders with complementary product portfolios and customers, combining CPUs, GPUs, FPGAs, Adaptive SoCs and deep software expertise to enable leadership in computing platforms for cloud, edge and end devices. Together, the combined company will have the ability to capitalize on opportunities spanning some of the industry’s most important growth segments, including data centers, gaming, PCs, communications, automotive, industrial, aerospace and defense.

There are many ways to play the AI revolution, but direct software and hardware innovators are my favorites.

Disclosure: I own shares of NVDA, AMD, and CDNS for the Zacks TAZR Trader portfolio.

Breakout Biotech Stocks with Triple-Digit Profit Potential

The biotech sector is projected to surge beyond $775 billion by 2024 as scientists develop treatments for thousands of diseases. They’re also finding ways to edit the human genome to literally erase our vulnerability to these diseases.

Zacks has just released Century of Biology: 7 Biotech Stocks to Buy Right Now to help investors profit from 7 stocks poised for outperformance. Our recent biotech recommendations have produced gains of +50%, +83% and +164% in as little as 2 months. The stocks in this report could perform even better.

See these 7 breakthrough stocks now>>