Here is what tech investors can learn from CES
This post is sponsored by Seeking Alpha's Alpha Picks. Starting today, new subscribers who join Seeking Alpha Premium will receive $50 off their annual membership. Join today to learn about the Top Stocks for 2025 at the upcoming Investing Forum on January 8th.
Here is what tech investors can learn from CES
I’ve been to CES several times. It’s a magical experience for any tech enthusiast. Before COVID, CES probably had 170,000 people attend Las Vegas every year. It has always been a major tech event. Since 1967, major products like the VCR and CD players have been released there.
This week, CES started on January 6th with a massive presentation. I haven't been this excited for CES since the launch of mobile phones. Software is great but some of our greatest innovations happen at the hardware level. This week it starts with Nvidia’s GPU announcements.
Jensen Huang, the CEO of NVIDIA, held his opening keynote at Mandalay Bay, focusing on AI and accelerated computing innovations. I linked to the full keynote at the bottom of this post. To help you get started, I've included some of the highlights from his keynote.
Find out what Nvidia released last night at CES
When I first started playing video games, I remember how expensive it was to buy GPUs for PCs. Since PCs had multiple components, most kids defaulted to video game consoles for simplicity and more power. Microsoft’s Xbox, Sony’s PlayStation and even the Nintendo 64, required custom GPUs to power state-of-the-art games.
At the time, I think most people overlooked Nvidia as a company for two reasons: 1) it is an enterprise technology that sits behind a server or in a console; 2) video games don’t appear to have much real-life applications. Nvidia also didn’t do the same branding exercise as Intel. However, none of this ever stopped Nvidia from investing billions in chip technology.
In many ways, video games have to augment real-life. Which means Nvidia’s GPUs had to learn how to compute power for multiple variables in real-time. For example, Intel’s CPUs were designed to execute tasks sequentially or with limited parallelism. Meaning one application at a time, like a web browser or word processor. Nvidia’s GPUs were designed to execute many tasks simultaneously in parallel. If you have ever played video games or edited videos, you know how essential parallel processing is.
Over time, Nvidia’s GPUs began to power more than video games. In fact, parallel processing has been at the core of multi-modal A.I. applications like OpenAI’s ChatGPT and Google’s Gemini. It is required for fast applications.
These advancements in technology are why Nvidia’s stock has been on an absolute tear for the past five years. What seems like an overnight success, has been 32 years in the making.
Three Highlights from Jensen’s CES Keynote
Let’s get started with the announcements.
Project Digits
This is a personal A.I. supercomputer priced at $3,000 and launching in May 2025. While still early, I think the shift towards A.I. hardware devices are coming. Right now Digits is designed for developers. But this tends to be an early indicator that a consumer product is coming. The next evolution will be an A.I. laptop and then mobile phone.
Jensen mentioned that Nvidia’s entire A.I. stack runs on its latest A.I. supercomputer. It is becoming obvious that this will power more people’s devices in the future. The opportunity for investors is where Digits will be deployed in the future. Nvidia has already become the market leader in GPUs, but who will become the next winner when it comes applying A.I. tech?
Autonomous Mobility
Sure, ChatGPT has been great for most consumers, but autonomous vehicles are the real Nvidia winners for me. Only in the past two years have I noticed the advancements in hardware, software and engineering. These are three attributes that Tesla has been well known for. But with Nvidia launching so many new products, many startups are playing catch up.
Below are a few of the product announcements when it comes to autonomous mobility:
Each of these models will enhance safety and security of autonomous vehicles, capture valuable data and deploy large, robot fleets. Companies like Toyota, Aurora, and Continental are joining Nvidia to accelerate compute when it comes to commercial vehicles.
While everyone thinks in terms of commute costs, I think the manufacturing and logistics will benefit most from these developments. The humanoid robotic element is hardest to predict here but will have the greatest real-world application for humans. Outside of Tesla, I don't know when the robotics company will go public but it seems inevitable now.
Agentic AIs are coming
When I use A.I. tools, building agents probably excites me the most. I’m seeing more developers build tools that can communicate across channels to do micro tasks. Nvidia is leading this front with agentic A.I. frameworks.
In a recent podcast interview, Jensen mentioned that Nvidia will have 100,000 A.I. agents in the future. These agents will be working on internal and external issues around the clock. He will still have the need to hire the best engineers, but the demand for intelligence agents will outpace what humans can physically do in real-time.
Jensen said it best yesterday that,
“The ChatGPT moment for robotics is coming. Like large language models, world foundation models are fundamental to advancing robot and AV development, yet not all developers have the expertise and resources to train their own,”
NVIDIA introduced Blueprints which are pre-designed workflows to help developers build custom AI agents. Right now it is applicable for tasks like data analysis, video summarization, and PDF-to-podcast conversion. But the next step will be A.I. systems capable of reasoning, planning, and executing complex, multi-step tasks across industries.
At the Orchestration Layer, multiple agents will be able to work together across the enterprise. Jensen described 2025 as the ‘Year of A.I. Agents’ which I agree with. These agents will ‘digital employees’ and be able to perform 50% of human workloads across industries.
In many ways, everyone’s IT department will begin to look and work like HR to mange these digital agents in real-time. This sounds like an exciting future and I’m ready for it.
Jensen’s Full 2025 CES Keynote