Morgan Stanley analyst Charlie Chan and others pointed out in the latest report that they are optimistic about the prospects of Windows on Arm (WoA) AI PC. Analysts believe that AI personal computers based on Arm architecture are expected to gain a larger market share in the coming years, with related semiconductor stocks being potential beneficiaries.
Currently, processors based on Arm (from Apple and Qualcomm) have begun to catch up with x86 in energy efficiency (measured in performance per watt), which is crucial for the future competition of AI PCs.
For the shipment volume of WoA AI PC chips, analysts estimate around 2 million units in 2024, increasing to 15 million units in 2025, and further doubling to 30 million units in 2026. Assuming that NVIDIA and MediaTek will occupy 50% of the WoA PC market share in 2028, the shipment volume will reach 33 million units, with a compound annual growth rate of up to 93% from 2025 to 2028.
In the WoA market, Qualcomm currently plays a leading role, but analysts expect that next year, manufacturers such as MediaTek and NVIDIA will join the competition.
<h2>Explosive Growth Expected in the WoA Market in the Coming Years</h2>
The report believes that the factors supporting the development of WoA AI PC are gradually maturing, and this market is expected to experience explosive growth in the coming years. Analysts predict that by 2025, AI PCs will account for 30% of total notebook shipments, and by 2026, this proportion will rise to 50%.
According to IDC data, in 2023, Microsoft's Windows operating system had a market share of 79% in the PC market, followed by Apple's macOS (11%) and Google's Chrome OS (9%).
Analysts believe that with the AI boom, in the upcoming AI PC era, &#34;energy efficiency&#34; will be the most critical selling point for electronic devices. Analysts believe that companies that can make breakthroughs in both aspects and have integrated advantages will become the winners in the future AI PC chip race:
<blockquote>
Companies that can launch the most &#34;energy-efficient&#34; chips may be the winners of AI PCs. Running AI applications (such as Copilot) on edge devices in the future (whether it's smartphones, laptops, etc.) will consume more power, and battery life has always been a key limitation of mobile devices. Therefore, we expect that the lower power consumption advantage of WoA will be one of the factors helping to increase its penetration rate in 2024.
The Arm architecture is a Reduced Instruction Set Computing (RISC) architecture, while Intel's x86 ISA is a Complex Instruction Set Computer (CISC) architecture. RISC instructions are fewer, simpler, and therefore have simpler decoding logic and fewer transistors
Although the performance of standard Arm cores may be lower than x86 products, Arm-based CPUs have shown significant advantages in power consumption.
</blockquote>
Intel still holds a 72% market share in the laptop CPU market today and produces most of its CPUs using its own architecture. Arm CPUs have a small share in the Microsoft Windows ecosystem.
Since 2016, MediaTek has been the supplier of Arm CPUs for Chromebooks. Arm-based CPUs are known for their compatibility with Android applications and longer battery life, helping Chromebooks gain a higher market share in the education sector.
For the shipment volume of WoA AI PC chips, analysts estimate around 2 million units in 2024, increasing to 15 million units in 2025, and further doubling to 30 million units in 2026. Assuming NVIDIA and MediaTek together capture 50% of the WoA PC market in 2028, the shipment volume is expected to reach 33 million units, with a compound annual growth rate of up to 93% from 2025 to 2028.
From an industry chain perspective, analysts point out that TSMC, MediaTek, NVIDIA, and a few other companies play important roles in the WoA AI PC race:
<ul>
<li>TSMC's 3-nanometer advanced process technology and CoWoS packaging technology will help Arm-based CPU compete with Intel chips.</li>
<li>As the most experienced Arm SoC design company, MediaTek is expected to collaborate with NVIDIA to launch a gaming AI PC chip integrating CPU, GPU, and AI accelerator in 2025.</li>
</ul>
<h2>Arm has a Technological Edge over Intel</h2>
Analysts note that in recent years, Arm has introduced more advanced CPU cores for smartphones. To enhance the performance of Arm-based CPUs, downstream companies have started developing custom Arm core CPUs, such as Apple's M-series CPUs for MacBook. The A-series chips on iPhones also benefit from Arm CPU cores, leading in performance among various manufacturers.
More importantly, Arm's leading edge in smartphones is expanding to the PC domain. As more manufacturers aim to integrate specialized computing units for AI tasks like NPU or APU into their designs, Arm's excellence in power consumption and edge AI inference tasks is becoming increasingly prominent.
Morgan Stanley's analysis shows that currently, Arm-based processors (from Apple and Qualcomm) are beginning to catch up with x86 in energy efficiency (measured in performance per watt), which is crucial for the future competition of AI PCs
In addition, Microsoft has also provided greater support for Arm in software, which is beneficial for manufacturers to transition to Arm architecture chips:
<blockquote>
Arm64EC may be a key point. Microsoft's Arm64EC announced in 2021 can convert code of x86-based applications to Arm-based code. It is similar to Apple's Rosetta 2, which was used for Apple Macbooks to transition from Intel chips to Arm chips. We believe that Windows 12 in 2025 may better support Arm architecture than the current Windows 11, with better virtualization technology (emulation) or software adaptability.
</blockquote>
<h2>AI PC is the key to Microsoft's competition with Apple. The outbreak of AI PC in 2027 will be reflected in Arm's financial reports.</h2>
Analysts point out that for Microsoft, maintaining the dominant position of the Windows operating system in the PC market is crucial for future deployment of AI services such as Copilot. Apple has already demonstrated the advantage of using Arm-based chips in MacOS. Analysts believe that AI PC may create a new entry point for Windows on Arm.
Although Arm has tried to enter the PC CPU market based on x86 architecture several times in the past decade, the results have been limited. Although Microsoft proposed the idea of WoA as early as 2011, it was not successful due to difficulties in integrating software with multiple chip manufacturers. It is expected that in this round of AI boom, Microsoft will increase its support for Arm design in software.
Analysts also stated, as Arm continues to strengthen its heterogeneous CPU design capabilities, especially with the introduction of the new generation &#34;Blackhawk&#34; high-performance core, Arm can provide top-notch hardware solutions for emerging players in AI and other fields, especially in the large-scale edge artificial intelligence inference field:
<blockquote>
Licensees now have the ability to design CPUs and NPUs that are consistent with GPUs and optimized for the PC environment, meeting Microsoft's 40x performance improvement requirements to support running AI applications like Copilot on edge devices. From the beginning, Arm is in a unique position to supply chips for Windows-based AI PCs.
</blockquote>
In the WoA market, Qualcomm currently plays a leading role, but analysts expect manufacturers such as MediaTek and NVIDIA to join the competition next year. At the same time, traditional PC chip giants like Intel and AMD are also accelerating the development of their own NPUs (neural network processors).
Analysts also predict that given that AI PC is just getting started, the short-term financial impact on Arm may be small, but in the long run, Morgan Stanley believes that the increasing application of Arm architecture in AI PC and other fields will support a stronger profit momentum, which will be reflected in the financial reports starting from the 27 fiscal year.

Analysts at Morgan Stanley believe that the company capable of launching the most "energy-efficient" chips may be the winner in the AI PC era

AI PC changes the chip industry landscape: From Wintel to Windows on Arm?