Computer manufacturers, at least in their declarations, have high hopes for the Al PC. The new generation of devices is expected to provide users with new, previously unknown experiences and improve work efficiency.
Analysts and computer manufacturers have been stirring up the topic of Al PC for several months. Canalys predicts that the global market will absorb approximately 48 million of these computers this year, representing 18% of the global market for new PCs. According to Gartner, this figure is expected to reach 22% (54.5 million devices), while IDC predicts that this year's sales will reach 50 million units.
The Polish market, at least for now, shows no signs of business activity. Aside from varying degrees of media hype, there are no significant promotional campaigns or seminars dedicated to the new product group.
"We don't have much to say on this topic. This is a new issue for us, and we won't pretend to be experts," says the head of a well-known Polish integrator.
And this is not an isolated case. Nobody in Poland is really excited about the arrival of the AI PC. Some argue that it's a clever marketing ploy aimed at accelerating the computer replacement cycle. Gartner estimates that the average lifespan of a business notebook is four years, and for consumers, it's even longer. Manufacturers looking to improve their financial results can only rely on the aging Windows 10 alongside artificial intelligence (Microsoft is providing free support for this operating system until October 14, 2025).
Many people, even those familiar with the IT industry, have some difficulty defining what an AI PC actually is. Canalys uses this term to describe desktop or notebook computers equipped with a dedicated chipset or processor to handle the device's workloads. Examples include AMD XDNA, Apple Neural Engine, Intel NPU, and Qualcomm Accelerator. According to IDC, these are personal computers with a system-on-a-chip (SoC) integrated circuit designed to perform tasks related to generative artificial intelligence. Gartner defines an AI PC as a PC "equipped with special AI accelerators or cores, neural processing units (NPUs), accelerated processing units (APUs), or tensor processing units (TPUs), designed to optimize and accelerate AI tasks on the device." Processor manufacturers define AI PCs slightly differently. Robert Hallock, Senior Director of Technical Marketing at Intel, defines them as devices containing a GPU with a neural processing unit and supporting two instructions. The first, VNNI (Vector Neural NetWork Instructions), significantly improves the performance of deep learning engines compared to previous generations. The second, DP4, supports GPUs in image processing. Importantly, this combination is already available in Intel Core Ultra processors, also known as "Meteor Lake." In short, for Intel, the AI PC "badge" signifies compliance with specifications or special features not found in other computers. In this context, it's worth recalling that the manufacturer previously employed a similar marketing maneuver, introducing the "Centrino" brand, distinguishing computers with Wi-Fi cards, as well as the "Viiv" brand, representing multimedia home entertainment centers.
Experts from the chip development startup Ampere see it completely differently. They believe that training AI models requires enormous computing resources, and the cloud is the best place for this. However, the inference process doesn't require extraordinary power, and traditional processors are sufficient for such tasks. Therefore, the entire hype surrounding AI PCs resembles a storm in a teacup.
Until recently, the common narrative was that a computer doesn't need to have massive computing power because the average user wouldn't be able to utilize it anyway. According to this approach, a large hard drive is also unnecessary, as data can be stored in the cloud. It's important for a computer to be lightweight, convenient, and energy-efficient. Perhaps the best example of this concept is the Chromebook, which relies primarily on web applications.
However, after two lean years, computer manufacturers have concluded that this is a road to nowhere. In any case, someone has flipped the switch. Therefore, voices are increasingly heard claiming that a computer with high processing power will provide better performance and efficiency when handling AI-related workloads than external servers or cloud services. The sustainability narrative has also been silenced, as an AI PC will consume significantly more energy than a regular PC.
Until now, the average person has only had the opportunity to interact with intelligent chatbots trained on vast public data sets. With the expansion of AI PCs, this will change, as users of computers with new processors will be able to perform tasks related to artificial intelligence without the need for an internet connection. They will have a private assistant performing tasks similar to ChatGPT. These include "intelligent" tools for editing videos, planning family trips, or taking conference minutes. Of course, as hardware and software develop, the list of possibilities will grow.
Analysts expect that in the initial phase, most AI PCs will go to commercial buyers. We are already seeing interest in these devices among manufacturing companies seeking to apply AI to design, construction, research, and development. It's possible that over time, AI PC users will also be joined by individual consumers, looking for solutions that facilitate the creation of digital photos or new experiences while interacting with multimedia entertainment, including computer games.
While the AI PC is ready in terms of hardware, the same cannot be said about the software. Manufacturers are looking towards Microsoft, after all, a great advocate of artificial intelligence. The Redmond-based company has invested $13 billion in OpenAI and is also financially supporting promising startups like Inflection AI, Adept AI, and Builder AI. Microsoft is building its AI strategy around the Azure cloud. The situation is worse for computers, as the Windows 11 operating system can only minimally utilize the NPU's capabilities, except for improving webcam performance and training its own artificial intelligence models, which will interest only a few users. A new version of Windows 11 adapted for Apple PCs is expected to be released at the end of this year.
However, users of newer computers with a bitten Apple don't have to wait. The NPU has been available in Apple computers for four years, offering intelligent features that increase user productivity and creativity, including translation, text prediction, and creating near-professional TikTok videos. Additionally, there are solutions that improve user comfort during communication, such as Portrait mode, studio lighting, and noise filtering.
Apple specialists have managed to integrate many features into macOS that rely on machine learning. It's also significant that the company's products often appeal to specific user groups – graphic designers, filmmakers, and programmers. Creativity is valued in these professions, and the use of artificial intelligence will significantly increase their work efficiency. For example, Adobe Photoshop uses NPU to speed up operations on filters, layers, and when adding special effects.
The actions of many other software vendors remain a separate issue. Companies like Adobe are trying to introduce increasingly advanced AI-based features, which increases the demand for more powerful computers. As hardware becomes more efficient, more opportunities arise for software innovation.
Source: https://crn.pl/
RSS Feed