The AI PC brings inference to the edge expanding benefits of AI beyond the browser
ABSTRACT: In this report we break down the AI PC – what it is, what it could mean for us, and why AI capabilities in your edge devices will be as ubiquitous as wifi. Our conversations with Dell executives spotlight how hardware innovations drive software breakthroughs, underscoring the critical synergy of the CPU, GPU, and NPU.. We look through the lens of Dell’s 2025 strategy unveiled at CES that includes a simplified branding approach (Dell, Dell Pro, and Dell Pro Max) and the Dell Pro AI Studio software platform, enabling users to easily find, train, and deploy AI models locally on AI PCs.
Market forecasts predict AI PCs will capture 64% of the PC market by 2028, reflecting an industry-wide shift toward faster, more efficient processing near the data source. 1.5 billion PCs are expected to be refreshed in the next five years. And with 50% of data already at the edge, there is considerable need for AI capabilities beyond the browser. Dell’s new generation of AI PCs is built with modular, sustainable materials, addressing both performance and environmental considerations. Real-world applications highlighted in interviews span healthcare, finance, education, and more, illustrating AI’s potential to transform workflows, enhance productivity, and unlock unprecedented creative possibilities.
By combining robust hardware, flexible design, and user-focused software tools, Dell aims to foster a comprehensive AI ecosystem that empowers individuals and enterprises alike. As businesses evaluate device refresh cycles, the question is no longer if AI PCs will revolutionize workflows, but how—and Dell is positioning itself as a leading partner to meet that demand. 2025 is the year our experiences with AI go beyond the browser.
Introduction:
There’s big nostalgia energy when people talk about AI, often linking our current tech evolution to the dawn of the internet. The early AI race has been about compute. AI Inference makes AI real, and now AI PCs bring that inference to the edge. However, the long-term dynasties of the AI game will be the ecosystems that empower the humans using them to do epic things. Dell aims to be that ecoysystem by helping customers meet their data with AI anywhere, whether with an infrastructure on-prem data center solution or at the edge with an AI PC. Hardware leads software, and while AI PCs aren’t new, they’re not everywhere yet, but their utility is poised to make them ubiquitous. As Michael Dell told us during their CES event, “nobody would think of buying a notebook without WiFi…that is what is going to happen here.”
When I think about AI PCs and the general AI ecosystem, I keep coming back to the human body. Imagining the CPU, GPU, and NPU communicating in sync like the human brain. The CPU acting as your Prefrontal Cortex, the GPU as your Visual Cortex, and the NPU as your Hippocampus. Extending this further, one can visualize the hands of the body gathering data and interacting with edge devices. The synergies and positive impacts of this seamless coordination solving problems and delighting the world around us. The human body is this incredible machine, each part in communication with the central nervous system. The AI PC is the closest consumer hardware manifestation we have of our finely tuned organism.
Market success hinges on the stories of the magic – a term Michael Dell used frequently in his keynote – the hardware makes real. The research conducted, the productivity gained, the fraud detected, the solutions deployed, and the lives saved. Buyers still aren’t sure what upgrading their laptop will help them achieve, but premium devices all have the ability to handle AI workloads inside. Translation: we’re about to understand the benefits of AI beyond the browser.
Highlights from our coverage include:
- 50% of data is at the edge
- AI PCs are expected to make up 64% of the PC market by 2028
- Dell announced new AI PCs that combine the power of AI with the convenience of a PC
- Dell has simplified its branding approach
- The Dell Pro AI Studio software platform helps customers find, train, and deploy AI models on their devices
- The AI PC’s potential to revolutionize various industries, including healthcare, finance, and education
- Q&A with Michael Dell, Jeff Clarke, and Sam Burd plus full interviews with Kevin Terwilliger, Jon Siegal, Sam Burd, and Marc Hammons
AI PC Market Overview:
Simply put, AI PCs are a traditional PC on steroids. They can compute complex AI workloads locally while optimizing the rest of your machine performance. On the ground in Las Vegas at CES, I spoke with attendees about how AI is factoring into their device refreshes and commercial lifecycles. Almost everyone said a version of the same thing “I’m waiting to see what the benefits are before I make the investment.” Speaking to a group of IAAPA leaders in the amusement industry here also afforded an interesting insight, they shared “it feels like the technology sector is only selling to technology companies, I’m not sure exactly how this tech will benefit our industry and achieve ROI.” Therein lies the key to gaining momentum in the market – connecting the capabilities to the solutions they provide.
Morgan Stanley estimates AI PC sales made up 2% of the PC market in 2024. That’s expected to grow to 16% this year, 28% next year, and 64% by 2028. Their data also notes 75% of CIOs are already or are planning to explore AI PCs. Safe to say that future commercial PC purchases will be 100% in this category very soon, but when? It’s about more than just tracking device sales however, it’s about understanding the data landscape. 50% of data is already at the edge, and everyone is trying to meet the data where it is with AI.
The following visualization from ETR shows the Net Score on the vertical axis, which measures spending momentum. The horizontal axis shows Pervasion, which is a measure of penetration into the survey of more than 1,700 IT decision-makers. The lines show the change in position over time, in this case the last year. It looks like Apple has made dramatic improvements in market penetration. This confirms that Apple, which develops its own custom silicon and has been shipping Arm-based processors with NPUs for years, is Dell’s biggest competitor in this space.
The red dotted line is the Laptop industry Net Score average. This shows the battle that is being waged by Dell and Apple, where they are performing significantly above average. We also see that although Apple’s Pervasion in the survey has accelerated, Net Score has come down.
ETR.ai Technology Spending Intentions Survey (TSIS) Vendor Position – January 2025
Dell CES 2025 Announcements:
We’ve covered Dell’s AI Factory extensively, and now we’re getting a first look at their latest consumer-focused announcements to build notebook momentum in 2025. The AI PC is a new category of devices that combine the power of AI with the convenience of a PC. Dell has been working on this technology for several years and has made significant progress in recent months. We discuss AI at the edge and its potential applications in various industries, including healthcare, finance, and education. We will also dive into the Dell Pro AI Studio, a software platform that helps customers find, train, and deploy AI models on their devices.
Simplified Dell Branding:
Dell announced three tiers of offerings across their product lines:
- Dell: everyday Dell user at work, school, play
- Dell Pro: power users that need professional tools
- Dell Pro Max: maximum performance
Pretty straightforward. The tech community grumbled a bit as we often do when brands evolve, but that’s not the biggest story out of Dell this season. Plus, the renaming is data driven. Discussing the approach, Kevin Terwilliger shared an interesting data point, “Deloitte actually has a study that shows that 74% of consumers actually stopped in the purchase path of technology because they got confused.” Kevin adds further insight in our interview on the motivations behind their new three tier offering:
“It’s critical that we help those customers find the right device because if there’s a lot of confusion in trying to understand the brands, understand the tiers, when we get down to the device, there’s a lot of choices they need to make about the silicon options. They make decisions about the NPU and whether or not they want Copilot+. We want them focused on more of those use case decisions and not distracted with the sub brand structure or the tiering.”
Watch the full interview with Kevin Terwilliger
Dell AI PCs:
The hardware is indisputably evolved but what will we use AI PCs for? When I probe about use cases, productivity is often first off the shelf. It’s understandable, as productivity and efficiency are core pillars of the AI promise, but it’s about what the outcome is of us being more productive, efficient, and creative in our critical thinking about data intelligence. It’s about the cool thing, the epic experience, the cure, the climate, the <insert thing here> humans do with the hardware that matters. It soon will be about more than compute, it will be about what’s cool. But don’t worry there’s thermodynamics innovation happening across Dell’s AI Factory as well. The devices are also built to last with a modular design composed largely of sustainable and recycled materials. Dell isn’t looking to get you on a product replacement treadmill.
As Jon Siegal states in our interview:
“Why not buy a PC now that you know is going to serve what you need for the next couple years. No one’s buying a PC just for the next six months. They’re buying it for at least three, four years.”
Watch our full interview with Jon Siegal
AI PCs aren’t just the latest gadget to harness tech on the hype curve, they’re better PCs overall. These laptops will have to pique the interest of buyers on the commercial and the consumer side alike. And the key will be more than maximized compute, it’s demonstrating the advantage of the NPU, a modular design to give them runway, and the user experience that makes AI feel intuitive. Table stakes here is a tool that lessens toil. The exciting part is the z-axis of building we’re only just beginning to imagine let alone explore.
We discuss with “Mr. AI PC” himself, Sam Burd:
“We’ve gone all in on AI, we’re putting it into our devices. It’s basically accelerators that are sitting next to the CPU, so you can think about that as enabling the apps that we use every day to be more intelligent, so capability you’re going to get in the work that you do. And then for our commercial customers, the most exciting piece to me is they’re going to be able to take their data generated at the edge on their devices, their PCs, and run their AI models on those devices. So taking workflows in their business space, making them smarter and more intelligent. So everything from the home user of today, your apps, using the NPU, seamlessly bringing more intelligence and capability to our commercial customers, working in a much smarter, more enabled world.”
What does that mean for the greater good of the world? He adds:
“Now you think about AI coming on the device, it is that kind of personal companion with capability that just makes you better. It saves you time, it allows you to be more efficient. It allows you to get more out of life and that’s really exciting.”
Watch our full interview with Sam Burd
Dell Pro AI Studio:
The ecosystem layer, the Dell Pro AI Studio, is a software that simplifies the process of developing and deploying AI models on AI PCs. We spoke with Marc Hammons about the NPU, Dell-validated tools, and their commitment to the developer user experience.
“There’s a new component, a new player, the neural processing unit. The really interesting thing is there’s a lot of differentiation out there when it comes to the NPU, but that adds complexity. Complexity in tool chains, complexity in software frameworks, complexity in knowing, hey, is this thing going to perform? Not only are you dealing with the silicon complexity that’s out there, you’re dealing with model complexity. What models run best on these machines? Which ones are going to solve the problems that I want to solve?”
To address this complex challenge, Dell is taking a three-pronged approach. First, they are curating a set of AI models that have been tested and validated to work well on AI PCs. These models will be available through the Dell Enterprise Hub, along with reference code and documentation to help developers get started. Second, Dell is providing a set of Dell-validated designs for AI PCs, which will ensure that the hardware and software are optimized to work together. Third, the Dell Pro AI Studio will provide a simplified development environment that allows developers to quickly and easily deploy their AI models on AI PCs.
Most impressively, Marc notes that this will reduce the time it takes to develop and deploy AI models from six months to just six weeks. Learn how in our full interview.
Watch our full interview with Marc Hammons
The Savvy Angle:
I knew Dell was all in on AI PCs when Michael Dell was standing in front of a 1985 Turbo PC to open their CES 2025 preview. Turns out someone had stashed it in a Round Rock closet. The mood was exciting, yet familiar. Steve Jobs stood in front of a 1984 Macintosh when he announced the iPhone 18 years ago. You can only pull the ‘founder in front of vintage hardware saying something profoundly ambitious’ move once. Michale Dell kept referencing magic, a term he saves for special occasions. Occasions like today.
If the devices are the magic as Michael Dell predicts, the AI Studio is the fairy dust that’s going to be the glue that keeps the community loyal to Dell’s AI hardware. They’re not only thinking about the adaptability, the sustainability, the modular design of the physical device, but also making sure that the developers who are using them (or whoever it might be, it might not be a traditional dev at this stage using an AI PC) are within an ecosystem that’s also going to be able to grow and match that same simplified user experience all the way through the ecosystem.
The reality is, at an estimated only 2% market share we don’t know what people are going to do with AI PCs yet. We know what Dell has done with them internally. We know a few examples of teams increasing productivity. But we don’t yet know what industries and creatives at scale will do. I’m looking forward to the stories we’re not expecting. The problems solved. The potential unlocked. It’s okay that we don’t know exactly what the future holds. We didn’t realize our smartphones would do more than send emails until Apple showed us a different way. It’s on Dell to show us the way with AI PCs.
Conclusion
Silicon geopolitics are impossible to ignore at this stage in AI. Supply chain security is critical. So is user adoption. We see that one of the challenges Dell may have is the variety of chips that they’re engaging with, but this optionality could also be a great thing for both Dell and their users.
We believe AI PCs will hit the mainstream market in 2025. We’ll see core and unexpected use cases across industries, from scientific research to storytelling and from medicine to movies, that emerge and inspire other applications. It will be critical for there to be collaboration across the AI ecosystem aiming to provide a simple UX for devs and creators, bring AI to data, and the humans benefiting from it. AI PCs are more than their AI workload capabilities. They represent the evolution of PCs into better, more sustainable, modular hardware to achieve and solve problems and workloads that haven’t even been invented yet that will require the new capabilities afforded by AI PCs. What people will do when able to leverage the other half of our data at the edge remains to be seen. And many need to see to believe – and upgrade their device. The next 18 months will be telling.
