 |
| By Karen Riccio |
PCs don’t get much love anymore. And major developments for desktops come far and few between.
However, the biggest advancement for this taken-for-granted industry in more than 20 years is being made right now.
And this may be just what AI needs to reach that final, elusive stage in its development cycle: widespread adoption.
That’s good news because, if you look at Gartner’s Hype Cycle chart (below), AI is stuck between a rock and a hard place.
Or, as Gartner calls it, between the “Peak of Inflated Expectations” and “Trough of Disillusionment.”
Find out more about the Hype Cycle here.
Although AI has been around for decades, its meteoric rise has happened on a compressed timeline.
AI skated through the “Innovation Trigger” stage after OpenAI released ChatGPT to the masses nearly two years ago. Within just one month, the generational AI tool attracted 100 million users.
But growth stalled. Monthly users peaked in April and dropped ever since.
How AI Gets Unstuck
Clearly, potential no longer moves the needle. So, users are asking, “What has AI done for me lately?”
It’s a legitimate question that deserves legitimate answers.
We’ve seen multiple iterations of ChatGPT. But it’s clear that innovation has taken a backseat to infrastructure.
That’s a good thing, and here’s why.
AI models require the fastest silicon chips. Chips run on super-powerful processors (GPUs) that also consume extraordinary amounts of electricity — something already in short supply in the U.S.
Keeping the GPUs and data centers cool and running 24/7 every day of the year requires even more power and water, another threatened resource.
Sure, Nvidia (NVDA) has come through with superchips and super GPUs … and even its own AI factories.
And mega-rich tech companies like Microsoft (MSFT), Amazon (AMZN), Alphabet/Google (GOOGL) and others own dozens of data centers and can afford whatever it takes to implement AI on a very big scale.
However, for most companies, AI is cost-prohibitive. And all the above challenges have become barriers to wider-spread adoption.
Meanwhile, in the land of PCs … the industry is also facing a growth problem. During Covid-19 sales skyrocketed with remote work coming into vogue. Plus, Microsoft stopped support for Windows 7, so the refresh cycle provided a much-needed boost.
Research firm Canalys reported that PC shipments reached 297 million in 2020. That’s up an impressive 11% from 2019, the biggest growth since 2010.
Lenovo (LNVGY), Hewlett Packard (HPE) and Dell Technologies (DELL) represented the top three PC makers during this period.
Since then, though, users haven’t had much to get excited about. That’s one reason worldwide shipments of traditional PCs dipped 2.4% year over year to 68.8 million units during Q3 2024.
However, now that support for Windows 11 ended on Oct. 4, many will want or need to upgrade. Another refresh cycle is just what the doctor ordered for the PC industry … and AI.
Enter the First of Its Kind — the AI PC
AI PCs aren’t on some faraway horizon. Some are already on store shelves.
And they allow people like you and me to tap into AI without needing a data center and an industrial cooling system for it.
Chipmakers Qualcomm (QCOM), Intel (INTC) and Advanced Micro Devices (AMD) are working to bring AI-capable PCs to the mainstream.
AI PCs are not only a consumer market. Enterprises and developers seeking to upgrade their employee PCs should also drive the market forward.
A Powerful Upgrade Cycle Is Underway
With the holidays just weeks away, we’re entering a big upgrade cycle. And this one will have AI PCs giving the rest a run for their money.
Like any new tech, the best is yet to come. Still, even early on, expectations for their popularity are sky-high.
Canalys’ latest forecast predicts that an estimated 48 million AI-capable PCs will ship worldwide this year. That’s about 18% of total PC shipments.
These shipments are set to exceed 150 million in 2025 or 40% of all PC shipments.
And in 2028, Canalys expects PC makers to ship 205 million AI-capable PCs.
The total market will be buoyed by the emphasis on low power consumption, efficiency and zero latency of AI PCs.
Because AI PCs reside on the edge (i.e., not in a data center or connected to a network), it reduces reliance on cloud servers. This, therefore, minimizes the risk of data breaches.
On top of that, AI PCs offer real-time data processing and autonomy, both of which enhance overall performance.
Higher Upfront Cost, but More Savings Overall
Cost savings over time is another significant factor.
Minimizing dependence on costly cloud AI services can result in significant drops in total cost of ownership.
With an expected spike in adoption, over half of PCs priced at $800 and above will be AI-capable by the end of 2025, with this share increasing to over 80% by 2028.
As a result, PC shipments in this price range will grow to account for more than half the market in just four years.
This will help boost the overall value of PC shipments from $225 billion in 2024 to over $270 billion in 2028.
One company I like in the AI PC space as a longtime staple in the PC industry is Dell.
Overall, $95 billion Dell is a solid business that is performing well. Demand for Dell's AI servers is off the charts right now.
Revenue from infrastructure products jumped 38% year over year in the most recent quarter and helped fuel an impressive 86% increase in the company's earnings.
Investors clearly like what they see, boosting shares 69% year to date.
A recovery in PCs could boost Dell’s business and share price big time by this time next year, especially with the upcoming refresh cycle for Windows PCs.
The marriage of AI and PCs will provide opportunities for companies across the technology spectrum.
AI PCs will also create a wealth of investment options in these earliest stages of development.
The timing couldn’t be better.
Best,
Karen Riccio
P.S. While AI PCs will present one way to get around the out-of-control demand for power and cooling at new AI data centers, those problems remain. I urge you to check out this brand-new presentation that has the solutions to those problems.
