Semiconductor Momentum Is Spreading Faster; Global Sales Could Rise to $2.25 Trillion in 2027
AI summary card
Semiconductor Momentum Is Spreading Faster; Global Sales Could Rise to $2.25 Trillion in 2027
Global semiconductor sales reached $152.0 billion in June, up 134% year over year, memory prices continued to strengthen, and non-memory growth accelerated for consecutive months; J.P. Morgan expects industry sales to reach $1.68 trillion in 2026 and $2.25 trillion in 2027.
- June industry sales rose 12% month over month, above the historical June seasonal increase of about 8%, with growth mainly driven by shipments.
- Non-memory sales rose to $63.5 billion, up 38% year over year, with year-over-year growth accelerating for the third consecutive month.
- Memory revenue reached a record $88.4 billion, with DRAM and NAND sales at $56.8 billion and $30.7 billion, respectively.
- DRAM and NAND average selling prices rose 275% and 368% year over year, respectively, with both accelerating for the eighth consecutive month.
- The industry model forecasts sales of $1.68 trillion in 2026 and $2.25 trillion in 2027, and the report believes the forecasts may still have upside.
Report interpretation
Overview
Based on WSTS industry data for June 2026, J.P. Morgan concludes that global semiconductor fundamentals remain strong and are spreading from memory chips to more product categories. June sales growth was significantly above historical seasonality, with higher shipments offsetting monthly normalization in blended average selling prices. Memory remains the main source of year-to-date growth, but non-memory categories such as microcomponents, analog chips, and sensors and actuators are also showing improving demand, tightening supply, and stronger pricing power.
Core views
The industry upcycle is broadening rather than being driven solely by memory. Memory revenue and prices continue to deliver strong performance, while year-over-year growth in non-memory sales has improved for three consecutive months. Demand momentum, orders, backlogs, lead times, and pricing power in company feedback are strengthening in tandem, corroborating the WSTS data. AI-related spending remains the core driver, with computing, memory, and networking chips showing the most pronounced earnings leverage.
Analysis framework
The report breaks down WSTS monthly sales, shipments, and average selling prices into categories including memory, microcomponents, analog chips, logic chips, optoelectronics, and sensors, comparing them with historical monthly seasonality and 10-year ranges; it then incorporates the latest data into its industry forecast model. Memory forecasts use a more granular bottom-up approach from the global team, while other major categories are extrapolated using relatively conservative seasonal growth assumptions.
Methodology notes
Identify changes in industry momentum through global semiconductor sales, shipments, and average selling prices.
The report uses WSTS June 2026 data and separately disaggregates the overall industry, memory, and non-memory categories to assess the sources and breadth of growth.
Compare the current month-over-month change with the normal level for the same period historically.
Historically, semiconductor sales in June typically rise about 8% month over month, while June 2026 rose 12%, indicating performance above seasonality.
Decompose sales growth into changes in shipments and average selling prices.
Unit shipments rose 12% month over month in June, while blended average selling prices fell 1%; the report views the latter as normalization after a 22% jump in May rather than a weakening of fundamentals.
Forecast annual industry scale based on the latest monthly data and category-level growth assumptions.
Except for memory, which uses a more granular bottom-up forecast, major categories are broadly extrapolated according to seasonal growth, so the report believes 2026 and 2027 forecasts may still have upside.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- NVDAAI computing beneficiary
- Strengths
- Directly benefits from AI computing power spending and growth in computing chip demand.
- Weaknesses
- The report does not provide company-level earnings forecasts or valuation analysis.
- Comparison
- The report discloses an Overweight rating, with a closing price of $217.55 on August 10, 2026.
- Risks
- Slower AI spending, changes in supply and demand, and failure to meet high growth expectations.
- AVGODual beneficiary in computing and networking
- Strengths
- Serves both computing and networking markets, providing multiple exposures to incremental AI infrastructure spending.
- Weaknesses
- The report does not elaborate on company-specific valuation and operating assumptions.
- Comparison
- The report discloses an Overweight rating, with a closing price of $422.40 on August 10, 2026.
- Risks
- Slower cloud capital expenditure, declining order growth, and weaker industry pricing power.
- AMDAI and general-purpose computing beneficiary
- Strengths
- Can benefit from improving microprocessor demand and growth in AI spending in computing.
- Weaknesses
- Compared with other key beneficiaries, the disclosed rating in the report is only Neutral.
- Comparison
- The report discloses a Neutral rating, with a closing price of $469.56 on August 10, 2026.
- Risks
- Intensifying competition, demand growth falling short of expectations, and supply-chain changes.
- INTCExposure to the computing chip industry
- Strengths
- Improving microprocessor demand and industry shipment growth can provide cyclical support.
- Weaknesses
- The report discloses an Underweight rating, indicating that improved industry momentum does not necessarily translate into a positive stock view.
- Comparison
- The report discloses an Underweight rating, with a closing price of $97.52 on August 10, 2026.
- Risks
- Competitive position, product execution, and insufficient transmission of industry growth into company performance.
- MUCore beneficiary of rising memory prices
- Strengths
- Significant growth in DRAM and NAND revenue and average selling prices provides strong earnings leverage for memory manufacturers.
- Weaknesses
- Highly sensitive to the memory pricing cycle and supply discipline.
- Comparison
- The report discloses an Overweight rating, with a closing price of $861.00 on August 10, 2026.
- Risks
- Memory prices peaking, overly rapid supply expansion, or weakening end demand.
- MRVLAI networking beneficiary
- Strengths
- Improvement in networking infrastructure demand, orders, and lead times is expected to support growth.
- Weaknesses
- The report does not provide a company-level target price or detailed earnings forecasts.
- Comparison
- The report discloses an Overweight rating, with a closing price of $208.56 on August 10, 2026.
- Risks
- Slower networking capital expenditure, customer concentration, and orders converting below expectations.
Key data
- June 2026 semiconductor sales$152.0 billionUp 134% year over year, above 128% in May.
- June 2026 month-over-month sales growth12%Above the historical June seasonal increase of about 8%.
- Month-over-month change in unit shipmentsUp 12%June growth was mainly driven by shipments.
- Month-over-month change in blended average selling priceDown 1%The report views this as normalization after a 22% increase in May.
- Non-memory sales$63.5 billionUp 38% year over year, with year-over-year growth accelerating for the third consecutive month.
- Memory revenue$88.4 billionA record high; memory accounted for about 77% of year-to-date industry sales growth.
- DRAM sales$56.8 billionUp 8% month over month and 373% year over year, with average selling prices up 275% year over year.
- NAND sales$30.7 billionUp 15% month over month and 377% year over year, with average selling prices up 368% year over year.
- 2026 industry sales forecast$1.68 trillionUp 111% year over year; up 30% year over year excluding memory.
- 2027 industry sales forecast$2.25 trillionUp 34% year over year; up 16% year over year excluding memory.
Impact & implications
The latest data reinforce the view that the semiconductor industry will continue to benefit from AI investment in 2026. Rising memory prices directly enhance the revenue leverage of DRAM and NAND manufacturers, while computing and networking demand benefits accelerator, processor, custom chip, and high-speed interconnect suppliers. Faster non-memory growth means the improvement in momentum is broadening, helping reduce concerns that this cycle is entirely dependent on memory prices, although product mix and ratings still differ significantly across companies.
Risks
- Memory accounted for about 77% of year-to-date industry sales growth, so industry growth still has a high concentration in the memory cycle.
- DRAM and NAND average selling prices have risen extremely sharply year over year; if supply increases or demand cools, price growth could normalize quickly.
- The 2026 and 2027 sales forecasts depend on seasonality and category-level model assumptions, and actual results may differ significantly from forecasts.
- If corporate orders, backlogs, lead times, and pricing power weaken, it would undermine the view that growth is spreading to non-memory categories.
- If AI capital expenditure slows, major beneficiaries in computing, memory, and networking may face earnings and valuation pressure.
What to watch
- Whether year-over-year growth in non-memory sales can continue to accelerate consecutively.
- Subsequent changes in DRAM and NAND average selling prices and bit shipments.
- Whether microcomponents, analog chips, and sensors and actuators can continue to deliver above-seasonal growth.
- Semiconductor companies' orders, backlogs, lead times, and pricing power.
- The sustainability of AI-related computing, memory, and networking capital expenditure.
- Whether the industry sales forecasts of $1.68 trillion for 2026 and $2.25 trillion for 2027 are raised further.