In Q1 this year, computers and mobile phones, two consumer electronics markets that have entered the era of stock, have shown completely different trends. Data released by IDC on April 8 shows that in Q1 2026, global PC shipments will be approximately 65.6 million units, a year-on-year increase of 2.5%; during the same period, Omdia gave a growth rate of 3.2%, corresponding to approximately 64.8 million units shipped. The figures from the two organizations are slightly different, but the conclusion is the same: the PC market has grown again at the beginning of the year.



After experiencing a downward cycle from 2022 to 2024, the PC market has maintained growth for several consecutive quarters, with the growth rate rising from 6.5% in Q2 of 2025 to 9.4% in Q3, and close to 10% in Q4.

However, this "recovery" at the beginning of this year was not evenly distributed. According to statistics from IDC and Omdia, the growth rates of Lenovo and Dell are much higher than the market. Apple has maintained growth, and HP has become the only one among the top five to decline.

The PC market continues to grow, but life for mobile phones is getting increasingly difficult. Old issues such as channel inventory, price pressure, and replacement cycles are frequently raised. The IDC report shows that global smartphone shipments in Q1 2026 will be approximately 290 million units, a year-on-year decrease of 6.8%. What's even more serious is that IDC has significantly lowered its full-year shipment forecast for 2026 to about 1.1 billion units, a year-on-year decline of about 13%. This figure represents the largest decline in nearly a decade.

PCs and mobile phones are both core categories of consumer electronics, and both face the same upstream cost pressure. Why are the trends so opposite? Where did this slight growth in PC come from? What cards are the major PC manufacturers playing?

1. Stocking up, replacement, and new products support PC’s Q1

To answer this question, we must first understand the background of the continuous growth of PCs in the past few quarters.

In Q3 of 2025, the growth rate of the global PC market is already not low. "There are factors of 'low base' and 'centralized replacement'." According to the analysis of PC practitioner Li Ze, on the one hand, the overdraft replacement demand during the epidemic has led to the previous low base. On the other hand, the timetable for the end of Windows 10 support has forced many organizations to focus on updating equipment.

By Q4, manufacturers and channels placed orders in advance to lock in storage capacity and costs, pushing some of the demand that was supposed to be released in the first half of 2026 until the end of the year. This has pushed up shipments by the end of 2025, and naturally squeezed out space in Q1 this year.

Against this background, Q1 in 2026 can still achieve a growth rate of 2%-3%, which is worthy of analysis.

The primary factor is channel stocking.

The global AI data center construction wave since the second half of 2025 has allowed storage giants such as Samsung, SK Hynix, and Micron to smell the more profitable HBM (high-bandwidth memory) business opportunities and begin to shift advanced production capacity to AI-specific memory production on a large scale. The supply side is shrinking and the demand side is strong, causing the price of memory chips to soar.


Omdia data shows that the cost of DRAM (running memory chips) and memory materials for mainstream PCs has increased by a cumulative US$122-237 from 2025Q1 to 2026Q2. Ye Maosheng, chief analyst of Omdia, said when the Q1 data was released that the memory and storage costs in Q1 increased moderately and are expected to increase significantly starting from Q2.

Faced with this "one price a day" situation, manufacturers and channels want to stock up more stocks and speed up shipments before prices completely rise. By Q1 of 2026, this "defensive hoarding" is still continuing.

This means that some shipments are actually "pad inventory" for sales in subsequent quarters. The Q1 data contains elements of channel stocking, which cannot be fully explained as a substantial recovery in terminal demand.

The second factor is that we are at the end of the Windows 10 replacement wave.

Microsoft will officially stop supporting Windows 10 in October 2025, and devices running the old system will no longer receive security updates. For enterprise IT departments, this means compliance risks and security breaches. Therefore, starting from Q3 of 2025, global enterprises have been accelerating machine replacement. As of Q1 this year, more than a quarter of Windows 10 users have not completed the migration. This group of users, especially enterprise users, continued to replace their phones in Q1 this year, contributing a considerable portion of shipments.

This can be found in regional markets. Omdia's regional data shows that EMEA (Europe, the Middle East and Africa) Q1 shipment growth reached 7.4% and Asia-Pacific 4.3%. Both regions still have relatively strong replacement demand.

The third factor is the concentrated release of new products by manufacturers.

This spring, mainstream manufacturers Lenovo, HP, Dell, and Asus have intensively updated a round of new products. The prices of models with local AI acceleration units have continued to drop, objectively stimulating some consumer demand.

Taken together, the growth in Q1 is the result of channel dealers rushing to stock up before price increases, enterprises switching machines due to the suspension of Windows service, and consumers rushing to buy before prices get higher. It is precisely because this part of demand has been overdrawn in advance that when the agency released Q1 data, it lowered its full-year PC shipment forecast for 2026 to a year-on-year decline of 11.3%.

Looking at consumer electronics as a whole, the pace of PCs is different from that of smartphones. Mobile phones neither have restrictions like Windows replacement, nor do they have enough support for corporate and educational orders. Therefore, facing the same memory crisis, PCs can survive for one or two more seasons, while mobile phones will enter the decline cycle earlier.

2. Lenovo is winning, HP is under pressure, and Apple is conservative

The 2%-3% growth of the PC market in Q1 this quarter did not fall evenly on the leading manufacturers.

Lenovo and Dell rely on the supply chain, commercial and small and medium-sized enterprise customers to "eat meat"; HP has encountered pressure on the consumer side, indicating that this round of supply and cost shocks is even more unfriendly to brands facing the mass market; Apple has the strongest pricing power, but it cannot escape the influence of this cycle; second-tier manufacturers are trying to pull themselves out of the purely cost-effective competition track by segmenting groups and scenarios.

Lenovo is still the player that has benefited the most from this rebound, with shipments of about 16.5 million units in Q1 of 2026, a year-on-year increase of 9%, and a global market share of more than 25%.

Behind this, on the one hand, Lenovo has stronger supply chain capabilities. Chen Jing, a practitioner who has been tracking the hardware industry for a long time, said that Lenovo’s global supply chain control capabilities are ahead of most manufacturers, and it can lock in production capacity and prices earlier when memory and storage prices rise and local shortages occur.

On the other hand, Lenovo focused earlier on the enterprise and commercial market, using the "device + service + solution" model to meet the needs of Windows replacement and local computing power.

IDC's research on the Chinese market shows that since 2025, PC demand for small and medium-sized enterprises and professional workstations has increased by more than 30% year-on-year for several consecutive quarters. Scenarios such as engineering design, video editing, and local AI inference have driven the shipment of high-end models. This market is less sensitive to selling prices and has higher requirements for stability. Lenovo focuses its resources on these customers in advance.

In contrast, HP, the world's second-largest company, is clearly under pressure.

In 2025, HP is not doing too badly: it has reaped the dividends of Windows 10 replacement, and Q4 shipments in 2025 will still have double-digit growth. However, Q1 shipments in 2026 (12.1 million units) fell 4.9% year-on-year, and the market share fell back to about 19%.

In Li Ze's view, HP is highly dependent on the C-end consumer market in Europe and the United States, especially the American market. This type of user is "not in a hurry to change phones", and HP's own cost-passing ability is limited, so sales pressure will naturally be greater. IDC data shows that the Americas is the only market with negative overall shipment growth in Q1 of 2026.


Source/IDC

What's more critical is the rhythm of play. Li Ze added that in the past two years, Lenovo's AI PC and hybrid AI routes have been clear, and Dell's overall plan on the enterprise side is also very clear. HP is in the CEO transition period, relying on cost reductions in the short term, and the long-term strategy is relatively vague to the outside world. "When companies are settling accounts, this is easily amplified by investors and major customers."

Dell, ranked third, had a bit of a surprise performance this season.

Dell went from a 3% year-on-year decline in Q2 of 2025 to a year-on-year increase of 18% in Q4. In Q1 of 2026, it maintained a growth of 7.8% (shipping approximately 10.3 million units), and its share returned to close to 16%. This trend shows that Dell is not simply following the recovery of the market.

Li Ze analyzed that Dell's basic fundamentals as an "old-school PC manufacturer" are still at work. It is a large buyer in the supply chain and has the ability to pass costs down, so it can raise prices for some enterprise customers and high-end models after the end of 2025.

On this basis, Dell’s more obvious change in this round is its product structure. After 2025, it will no longer rely solely on traditional commercial customers, but will focus on promoting thin and light notebooks, creator notebooks, and mobile workstations. This type of model can not only accommodate the budgets of enterprises and professional users, but also cover the needs of "local computing power, AI applications, and content production." Essentially, growth is focused on high unit price and high-performance products.

Apple, ranked fourth, has been conservative in this round of "AI PC wave", focusing more on stuffing computing power and local capabilities into existing product lines. Its Q1 shipments (7.11 million units) in 2026 will grow at single digits, and its market share will rise to 11%, mainly driven by the M5 chip MacBook Pro released at the end of 2025. The new MacBook Air and MacBook Neo had limited contribution to Q1 due to their late launch.

For Apple, PCs are just one part of the entire device and service ecosystem. What they care more about is the combination of “high unit price + high stickiness”.

It should be noted that the Mac business’s shipments in the first quarter of fiscal year 2026 (corresponding to Q4 2025) increased, but revenue declined. This shows that in the current PC environment, Apple has chosen to sacrifice part of its profits in exchange for a larger plate of active devices, hoping that iCloud and subscriptions will slowly make money later.

Looking further down the road, second-tier manufacturers such as Asus are also taking advantage of this market situation to accelerate shipments.

In Q1 of 2026, Asus shipped approximately 4.6 million units, growing the fastest among the five major manufacturers, with its market share increasing to about 7%. Part of the reason behind this is the rebound in demand for game notebooks and creator notebooks, and part of the reason is that it has launched new products in the mid-to-high-end price range with NVIDIA and AMD, and used configuration upgrades to increase the unit price.

3. AI PC: Life-saving straw or bubble?

If you look at global shipments, the keywords for the PC industry will still decline in 2026. Under this round of impact, manufacturers have grabbed the same life-saving straw: AI PC, labeling the machine AI and selling it at a higher price than "ordinary laptops".

Lenovo only offers the AI series in its product line, and AI models have accounted for more than 30% of the total shipments; HP has made Copilot+ adaptation (the official certification standard set by Microsoft for AI PC) on commercial notebooks, while consumers are more cautious; Dell mainly promotes "AI-Ready" workstations and commercial AI PCs, selling "a complete set of IT refreshing experience"; Asus promotes local AI on game notebooks and creator notebooks; Apple uses M series chips and Apple Intelligence does part of the "device-side AI" thing, but does not use the term AI PC.

But to make an AI PC that is adequate and easy to use, the threshold itself is not low.

On the one hand, AI PC has higher requirements for memory and storage. Taking the official caliber of Microsoft Copilot+ as an example, a "qualified" AI PC must have at least 16GB of memory, starting storage of 256GB, and an NPU of more than 40TOPS to run a complete set of system-level AI assistants, real-time translation, background cutout, multi-modal search and other functions.

In this price increase cycle of memory (DRAM) and flash memory (NAND), large-capacity memory and SSD (solid-state drive) are the parts with the fastest rising costs. Manufacturers can either maintain the configuration and pass the cost to users; or they can downgrade the configuration and use smaller-capacity hard drives to reduce the BOM (complete machine material cost). However, this operation will inevitably greatly reduce the AI ​​PC experience.

On the other hand, AI PC and the “disruptive changes” mentioned in the propaganda are still a long way away.

From a functional perspective, the differences between current AI PCs and ordinary PCs are mainly concentrated in three parts. First, there is an extra piece of computing power dedicated to running AI, so tasks such as speech transcription, image generation, and local small models run more smoothly; second, a full set of AI functions are built into the system, such as meeting minutes, document summaries, and smart screenshots, which can basically be called with one click; third, peripherals such as cameras and microphones are tied to AI to optimize experiences such as eye correction, intelligent noise reduction, and speaker recognition.

For these functions to be implemented, real money must be spent on hardware. As storage prices increase, terminal selling prices will naturally rise significantly. A channel source concluded that the starting price of AI PCs of mainstream brands is more than 5,000 yuan, and the main models are generally in the range of 8,000-15,000 yuan. If it is a high-end model, the price is 20,000 yuan or even higher.

In this price range, ordinary consumers are naturally cautious about whether to pay more for "AI". People in the above-mentioned channels told "Dingjiao One" that there are mainly two types of people who are willing to buy: one is heavy productivity and content creators who want more local computing power; the other is early adopters. The attitude of most users is that if they have enough budget and similar configurations, they will choose AI PCs; if the price gap is too large, they will give priority to traditional high-performance laptops.

The consumer side is waiting and watching, and the enterprise market's attitude towards AI PC is obviously divided.

Chen Jing mentioned that in the United States, large companies are now more inclined to invest money in the cloud and SaaS instead of replacing AI PCs. Currently, those who are willing to use AI PCs on a large scale are mostly in industries with higher requirements for data security and offline capabilities, such as medical, financial, and government contractors.

European companies move at a slower pace. Although local AI has advantages in terms of compliance, budget approval is strict and the hardware update cycle is long. Most companies will conduct small-scale pilots first, and then decide whether to roll out the technology after determining the actual benefits.

Looking back at the Chinese market, AI native companies, small and medium-sized software companies, and design and content teams are significantly more active in purchasing AI PCs and high-performance PCs. Such companies regard high-performance PCs with local AI capabilities as standard upgrades in the next two to three years.

Judging from the data, the penetration rate of AI PC is indeed rising. The proportion of the global market in 2024 is about 15–16%, rising to 31% in 2025. Gartner predicts that AI PC shipments will be approximately 143 million units globally in 2026, accounting for approximately 55%. So how to understand the value of these data?

Chen Jing told "Focus One" that the industry's current statistical caliber for AI PC is relatively broad. If it has an NPU or reaches a certain level of AI computing power, and claims to support local large models, it will be counted as an AI PC. It is unlikely to take a closer look at how much memory it actually has, or whether the pre-installed AI software is called.

This means that from high-end to entry-level, products in different price ranges will be counted together as AI PCs: among high-end models, some are AI PCs that truly have sufficient memory, storage, and NPU, and are equipped with software and hardware adaptation; mid-range or entry-level models, which meet the minimum configuration and have labels such as "AI PC" and "AI Ready", will also be included in AI PC shipments.

4. Conclusion

Going back to the original question, has the computer market really picked up?

In terms of shipment figures, Q1 of PC did show positive growth, but this wave of growth was largely supported by channel stockings and corporate replacements, and cannot fully represent the real demand of the consumer market.

For the world's top five manufacturers, they need to rely on the wave of AI PC hardware upgrades to retain more high-end, commercial and high-computing users. The trend of AI PC will indirectly squeeze out small and medium-sized manufacturers faster, further increasing industry concentration.

Judging from the penetration rate of AI PC, the number is rising, but the consumer side is more cautious, and the European, American and Chinese markets on the enterprise side have different levels of payment, which may affect the subsequent market structure.

Computers are just "not that miserable" for the time being, and there is no real recovery yet.