Leikeji Digital Group | Editor: Chongjia | Supervisor: Luo Chao
Recently, market research firm IDC released Q3 2026 global Personal Computing Devices (PCD) market data. According to IDC's definition, PCDs include desktop computers, laptops, workstations, and tablets, essentially covering the PC + tablet segment.
Surprisingly, Lenovo ranked first in both shipment volume and revenue, leading brands such as Apple. Additionally, contrary to Leikeji's (ID: leitech) prior impression that overall shipments in the PC and tablet markets have been declining, Lenovo achieved counter-trend growth in revenue. This shift is closely related to the evident trend of AI integration in PCs and tablets.
Shipments Declining, Revenue Growing
First, let us examine the PCD shipment data released by IDC. The top five brands in Q3 were Lenovo, Apple, HP, Dell, and Samsung, with corresponding shipment volumes of 19.1 million units, 17.2 million units, 10.5 million units, 7.6 million units, and 6.1 million units, respectively.
Total market shipments reached 94.1 million units, representing a year-on-year decline of 19.3%. All top five brands experienced shipment drops; however, Lenovo and Apple maintained market shares of approximately 20%, retaining an advantage over other brands. It is evident that the current PC and tablet industry faces overall weak market performance, with even leading brands unable to escape the downward trend.
Looking at revenue data, the top five brands in Q3 were Lenovo, Apple, HP, Dell, and ASUS, with corresponding revenues of USD 19.5 billion, USD 19.0 billion, USD 13.2 billion, USD 11.3 billion, and USD 5.9 billion, respectively.
Notably, total global PCD market revenue in Q3 reached USD 91.0 billion, a slight year-on-year increase of 0.8%. Among the top five brands, Lenovo, Apple, and ASUS all achieved revenue growth. In terms of market share, both Lenovo and Apple exceeded 20%, capturing the largest portion of total market revenue.
The most direct factor behind declining shipments but increasing revenue is the rise in unit prices for PCs and tablets. The underlying reasons are not difficult to understand. On one hand, rising storage prices have significantly impacted the pricing of PCs and tablets, driving up their costs. On the other hand, storage pressures have led manufacturers to focus on high-end products while reducing entry-level offerings, thereby increasing the sales proportion of higher-priced devices.
Given the current market conditions, the volume-driven strategy is no longer viable for the PC and tablet markets. Convincing consumers to pay more for premium devices while maintaining revenue and profit margins has become a key priority for manufacturers. To sustain high product premiums, merely citing rising storage costs as a justification is clearly insufficient. Amid the sweeping AI wave, enhancing the AI capabilities of products has become critical to building competitiveness for PCs and tablets.
Betting on AI: Lenovo and Apple Take Different Paths
AI integration is the prevailing trend for both PCs and tablets, but specific implementation strategies vary across brands.
For Lenovo, its hardware AI implementation follows a "One Agent, Multi-Terminal" approach. This involves a single AI agent accessible across multiple terminals. The user's agent is linked to their personal account, but to ensure data security, the user's personal knowledge base is stored locally. If users switch devices, Lenovo offers an "AI Device Migration" feature to facilitate local AI data transfer.
Currently, on mainland China devices, Lenovo primarily promotes the Tianxi Personal Super Intelligent Agent, which employs a hybrid edge-cloud architecture and a cross-device personal knowledge base capable of accessing data accumulated by users across PCs, tablets, and smartphones. In overseas markets, Lenovo has also launched Lenovo Qira, positioned similarly as a cross-device personal super intelligent agent.
Regarding the hardware foundation, Lenovo's AI devices utilize heterogeneous computing solutions, fully leveraging the performance and computing power of CPUs, GPUs, and NPUs. The CPU handles global scheduling, particularly for Agent tasks involving cross-software calls. The NPU performs low-power AI inference using its dedicated computing resources, while the GPU manages high-load content generation. On the software side, Lenovo's hardware devices come pre-installed with various AI applications, including the system-level Agent "Lenovo Xiaotian," which provides AI office assistance features similar to WorkBuddy.
According to data disclosed by Lenovo, AI has positively boosted its PC shipments. In Q2 2025, AI PCs accounted for over 30% of Lenovo's total PC shipments. In Q3 of this year, Lenovo's global market share for AI PCs reached 25%, ranking first. In the domestic market, AI PCs comprised more than 50% of Lenovo's laptop sales. Furthermore, AI has had a positive impact on Lenovo's tablet sales. In the Q3 domestic tablet market, Lenovo ranked third in shipments and saw an increase in sales volume.
In comparison, Apple's approach remains relatively basic. First, Apple has not developed a targeted AI solution specifically for PCs and tablets, largely relying on the same Apple Intelligence framework and features used on iPhones. Second, the rollout of Apple Intelligence has been bumpy, with the mainland China version only recently entering the testing phase.
Apple's hardware AI solution generally combines edge and cloud processing. Edge-side models handle simple tasks without uploading data, while complex tasks rely on external support—using GPT cloud models in overseas markets and Alibaba's Qwen model in mainland China. For core productivity scenarios on Macs and iPads, Apple AI still has limited capabilities. Features such as photo editing, document polishing, and visual recognition are available, but actual performance lags behind other brands.
Nevertheless, regarding Mac products, Apple has benefited from the AI boom. Based on Q3 data this year, Macs generated over USD 10.0 billion in revenue for Apple, a year-on-year increase of 29%.
Mac's ability to capitalize on the AI trend stems primarily from its hardware architecture and operating system, which are naturally suited for running Agent tools and locally deployed models. Most AI users purchase Macs not because of Apple Intelligence, but due to its superior cost-performance ratio and stability in local computing. In contrast, iPad sales and revenue both declined in Q3, indicating that Apple's proprietary AI features lack significant appeal.
Why Are AI Roadmaps for PCs and Tablets Diverging?
As AI technology advances further, the AI roadmaps for PCs and tablets will exhibit more pronounced divergence, given their differing form factors, usage scenarios, and computing characteristics. To support more complex AI and Agent capabilities, the underlying hardware ecosystem—including CPUs, GPUs, and storage—is undergoing changes.
From a computing perspective, heterogeneous computing has largely become a consensus, with all hardware resources involved in AI operations being fully utilized. Future AI PCs will rely more heavily on fine-grained allocation of computing power: CPUs will handle global scheduling and cross-software collaboration for system-level Agents; GPUs will provide high-load computing power for local large models, complex image, and video generation; and NPUs will serve as low-power AI engines for real-time online device functions, handling background tasks such as knowledge base updates and screen visual perception.
However, the local computing power of most consumer-grade devices still cannot support models with massive parameter counts. Therefore, edge-cloud integration will remain mainstream. When ordinary users employ various Agent tools for office work and productivity tasks, they will still need to leverage cloud computing resources.
Storage is also a key determinant of AI experience. The sustained rise in storage prices this year is primarily driven by the massive data processing demands brought about by the AI explosion. Smoothly running large models locally and building cross-device personal knowledge bases require large-capacity memory. Simultaneously, frequent AI prompt caching and vector database retrieval impose higher requirements on the random access speed of solid-state drives.
Based on these hardware foundations, the AI evolution paths for PCs and tablets will show distinct differences:
For PCs, the future core direction lies in heavy-duty productivity and automation. Leveraging the stronger computing power and performance support of PC chips, PC-based Agents will gain higher system permissions, enabling them to operate across applications like humans—for example, automatically scraping web data, deeply analyzing long financial reports to generate charts, and taking over parts of workflow processing—to become true digital employees.
Tablets, on the other hand, will focus on light-weight AI scenarios. Limited by size and power consumption, tablets' local computing power ultimately cannot match that of PCs. They will prioritize lightweight, high-frequency AI functions such as real-time subtitles, meeting transcription, and document summarization.
Additionally, as part of each brand's hardware ecosystem, tablets can "borrow" computing power from other devices through multi-device collaboration. For instance, heavy AI generation tasks can be handed off to a PC under the same account. In this scenario, tablets can serve as an entry point within the AI interaction system.
In summary, Lenovo was able to achieve revenue growth in the sluggish Q3 PC and tablet markets and secure a high AI PC market share precisely because it laid out a comprehensive AI implementation plan early on and effectively managed AI connectivity across devices.
In the future hardware market, manufacturers will no longer just sell configurations and specifications; they must provide practical AI implementation capabilities. Whoever excels in this area will hold greater initiative in the next device replacement cycle.