DIGITIMES has reported, citing supply chain sources, that Intel plans to raise PC CPU prices by about 10% in October 2026. According to the report, which TechPowerUp covered on September 8, it remains unclear whether the price increase applies only to desktop chips or also includes notebook chips. Sources say the goal of the price hike is not to expand market share but to improve gross margins.
The report projects that global PC shipments will decline from 260 million units to about 250 million units by 2027. With little prospect of market expansion, Intel is reportedly seeking to boost profitability through higher prices. The report also included an estimate suggesting that if Intel's CPU shipments returned to 200 million units, its market share would recover to about 78%. However, this figure comes from the report's own calculation, not an official Intel projection. A simple division using recent PC shipment figures does not yield 78%, so the market size and scope used in this calculation would need verification.
Intel had already acknowledged price increases on certain products back in July. Beyond these price revisions, the latest report also describes production allocation that prioritizes server chips, as well as a review of lower-margin products. When considering the impact on the PC market, it's important to look not just at the size of the price increase but also at which products will continue to be supplied.
October Price Targets Still Unconfirmed; July Hikes Applied to Select Products Only
Details of the October price revision—including which products and regions it will apply to—remain unknown. By contrast, the July price revision clearly distinguished between products that saw price increases and those that didn't.
In a July 3 article, Tom's Hardware reported that an Intel spokesperson confirmed price revisions for some consumer and server CPUs. The company cited rising costs across the supply chain and strong demand as the reasons.
The top-tier server chip, Xeon 6980P, saw its recommended customer price rise from $12,460 to $13,955—an increase of $1,495, or about 12%. On the desktop side, the Core Ultra 7 270K Plus and Core Ultra 5 250K Plus saw price increases of $30 to $50 depending on the model.
Meanwhile, the previous-generation Core Ultra 200S series remained unchanged. The flagship Core Ultra 9 285K kept its recommended customer price at $599, while the entry-level Core Ultra 5 225 stayed in the $183–$236 range. As of at least July, Intel had not applied a blanket price increase across all products.
This difference alone doesn't prove that Intel deliberately targeted best-selling or high-margin products for price hikes. What is clear, though, is that the company's response to rising costs varied by product. It's also worth noting that recommended customer prices differ from actual transaction prices—how much of the increase gets passed on to PC makers' procurement costs or retail prices depends on purchase volumes and contract terms.
Repeated Price Revisions Add Upward Pressure on PC Prices
This isn't Intel's first price increase. The DIGITIMES report frames the October 5 price revision as the third such increase since late 2025.
However, earlier reports differ somewhat in their timing and scope. Some reports mention a 10–15% increase across product lines in February 2026, followed by roughly 15% in March, while other reports—citing South Korea's ET News—describe an approximately 10% price hike specifically for the consumer Core Ultra series in March. There were also reports anticipating additional increases in May.
Some views suggest Intel is aiming for price levels roughly 30% higher than 2025, but this figure comes from research cited by Minutes Logic Society rather than any cumulative price increase figure published by Intel. Since different reports cover different products and comparison points, these percentage increases cannot simply be added together to calculate the total cost burden on buyers.
Upward pressure is also being felt in PC prices themselves. Reports from Chinese retail markets describe price increases for Lenovo, HP, Asus, and Acer, with some models reportedly rising by more than 1,000 yuan. This comes on top of rising prices for memory and storage, not just CPUs.
In a February forecast, Gartner projected that combined DRAM and SSD prices would rise 130% by the end of 2026 compared to 2025, pushing PC prices up by 17%. The firm also estimated that memory's share of total PC component costs would rise from 16% in 2025 to 23%. Gartner further suggested that as profitability becomes harder to maintain, the market for entry-level PCs priced under $500 could disappear entirely by 2028.
IDC's forecast, published in June, projected that global PC shipments would fall 11.3% in 2026, while average selling prices would rise 18.3%. The outlook suggests that meaningful improvement in the memory shortage won't come until at least late 2027.
A roughly 10% increase in CPU prices doesn't automatically translate into a 10% increase in overall PC prices. But with other major components also becoming more expensive, PC makers have less room to absorb rising costs.
Server-First Production Allocation Is Squeezing PC Supply
One factor behind the price increases is CPU supply constraints. As demand for AI data center chips grows, Intel has reportedly been prioritizing its own fab capacity for server CPUs, which in turn has constrained the supply of PC CPUs.
The DIGITIMES report also points to this production allocation issue. If Intel increases server chip production further, it may need to shift more PC chip manufacturing to TSMC. How the company balances in-house production increases with external foundry outsourcing will affect both supply volume and cost.
Market share data reflects this shift. According to Mercury Research's Q4 2025 figures, AMD reached a record 29.2% share of overall x86 CPU shipment volume, while Intel held 70.8%. AMD posted 36.4% share in desktop, 26.0% in notebook, and 28.8% in server.
During this period, Intel's PC shipments—particularly notebook shipments—were weak, while AMD gained ground. Intel's supply constraints are seen as one contributing factor, but these figures alone cannot isolate the specific impact of production allocation decisions. Product competitiveness and each company's product lineup also play a role in share fluctuations.
On the financial side, Intel has been improving both revenue and gross margins. Second-quarter 2026 revenue rose 25% year-over-year to $16.1 billion. Gross margin improved by 12.9 percentage points to 40.4% (41.8% on a non-GAAP basis). Data Center and AI segment revenue grew 59% to $6.3 billion, while the Client segment grew 13% to $8.9 billion.
Rising average selling prices contributed to these results, along with efforts to secure long-term supply contracts. For the third quarter, Intel is guiding for revenue of $15.8–$16.8 billion, with gross margin of 41.0% (42.0% non-GAAP).
However, Intel Foundry, which handles chip manufacturing, remains unprofitable. Second-quarter foundry revenue rose 31% to $5.765 billion, but the segment posted an operating loss of $2.089 billion, translating to an operating margin of roughly negative 36%. Revenue from external customers was just $293 million, representing about 5.1% of segment revenue. Since the vast majority of foundry revenue comes from manufacturing Intel's own products, decisions about which products get priority for fab capacity affect both the product divisions and the manufacturing division.
Expanding supply will also depend on advanced manufacturing processes 18A and 14A. According to a report citing research from BlueFin Research Partners, wafer-to-wafer yield variation on 18A has improved, with production ramping up to 12,000–15,000 wafers per month at Fab 52 in Arizona and at the Oregon site. However, reducing wafer-to-wafer variation is not the same as achieving target yields across the entire product.
Panther Lake is the first PC product built on 18A. Reports indicate that the improved 18A-P variant has entered pre-production trial manufacturing at D1X in Oregon, with plans to eventually shift full-scale production to Fab 62. For external customers, Intel is expected to focus mainly on 18A-P and 18A-PT.
The next-generation 14A node is planned to begin trial production for Intel's own products in the second half of 2027, with full-scale production starting in 2028. Version 0.9 of the process design kit (PDK) needed for chip design is scheduled for release in October 2026. An Intel finance executive has said that defect reduction is progressing at its fastest pace since the 22nm generation.
That said, these are all efforts aimed at expanding capacity over the medium to long term. How much 18A production increases will translate into greater PC chip supply will still depend on allocation decisions relative to server chips. And 14A, with mass production not planned until 2028, won't directly resolve this October's supply constraints.
Review of Low-Margin Products Could Ripple Into Industrial PCs and IoT
Alongside the price revisions, another focal point is the review of lower-margin CPU product lines. DIGITIMES reported that Intel is considering discontinuing production and sales of what it calls "Small Core" products for profitability reasons, with potential impacts on industrial PCs, IoT devices, and embedded systems.
For these applications, low power consumption matters, but so does the ability to source the same product over an extended period. According to the report, CEO Lip-Bu Tan is selecting which products to continue supplying based on profitability criteria. Chipsets, LAN, and Wi-Fi-related products are reportedly not currently part of this review, but if the scope of the reassessment expands, related peripheral products could also be affected.
Some observers suggest that if Intel withdraws from certain product lines, it could create opportunities for companies like Qualcomm and MediaTek, potentially accelerating adoption of Arm-based SoCs in industrial PCs, IoT, and edge devices.
However, no formal announcement of discontinuation has been made. The specific products affected, the timeline for discontinuation, and the supply period for existing customers all remain unclear, so it would be premature to conclude that Intel's entire industrial and embedded CPU business is shrinking.
Workforce restructuring has also been reported. Industry sources cited by DIGITIMES suggested the possibility of an additional 5–10% workforce reduction, while also noting that with roughly 75,000 employees and ongoing hiring, the current headcount level is seen as appropriate. Neither point represents a confirmed announcement from Intel regarding further cuts.
Headcount figures need to be read carefully, since different reports use different scopes and reference dates. Core headcount at the end of Q2 was reported at 77,600, while total headcount at the end of 2025 stood at 85,100, and the company's previously stated core headcount target was 75,000. "Total headcount," "core headcount," and "target figures" are not directly comparable numbers.
Over the longer term, Intel has noted that headcount has fallen roughly 40%, from about 132,000 in 2022 to approximately 81,000 most recently. In July 2026, workforce reductions in the Data Center and AI division were confirmed, though the exact number of affected employees was not disclosed; the company stated this would not affect its product roadmap. Management layers have reportedly also been reduced from 12 to 6. Beyond product lineup adjustments, Intel appears to be restructuring its organization and operating costs to improve profitability.
For now, the key question is which products the October price increase will apply to, and how much of that increase will be passed on to PC makers' procurement costs and retail prices. Future pricing for Nova Lake and actual retail prices during the year-end shopping season will offer further clues about the impact.
For businesses relying on industrial PCs and IoT devices, continuity of supply matters just as much as pricing. How the review of low-margin products and capacity expansion efforts unfold will shape not only prices but also the range of CPUs available to choose from.