Quick Summary
Covering the latest research from top Wall Street investment banks

Semiconductor Momentum Is Broadening and Accelerating; 2027 Sales May Reach $2.25 Trillion

Institution
JPMorgan
Date
2026-08-10
Authors
Harlan Sur, Apoorva Kumar, Mayur Ramdhani
Company
-
Ticker
-
Industry
Semiconductors and Semiconductor Equipment
Rating
-
BullishLow confidenceJune WSTS data show simultaneous acceleration in industry sales, shipments, and non-memory growth, continued strengthening in memory pricing, and corporate feedback also indicates improvements in demand, orders, backlogs, lead times, and pricing power.
AuthorsHarlan Sur, Apoorva Kumar, Mayur Ramdhani
CoverageUnited States、Other
Business segmentsComputing chips、Memory chips、Networking chips、Micro components、Analog chips、Logic chips、Optoelectronic devices、Sensors and actuators、Discrete devices
Research firm divisions/subsidiariesJ.P. Morgan Securities LLC(Other)

AI summary card

Semiconductor Momentum Is Broadening and Accelerating; 2027 Sales May Reach $2.25 Trillion

Global semiconductor sales rose to $152.0 billion in June, with memory remaining strong and non-memory growth accelerating for the third consecutive month; JPMorgan expects industry sales to reach $1.68 trillion in 2026 and $2.25 trillion in 2027.

Industry view is positive; the report lists stock ratings as Overweight for AVGO, MRVL, MU, and NVDA, Neutral for AMD, and Underweight for INTC.
SemiconductorsArtificial intelligenceWSTSMemory chipsDRAMNANDComputing chipsNetworking chips
  • Industry sales reached $152.0 billion in June, up 134% year over year and 12% month over month, outperforming the historical average sequential increase of about 8% in June.
  • Non-memory sales reached $63.5 billion, up 38% year over year, with year-over-year growth accelerating for the third consecutive month.
  • Memory revenue hit a record $88.4 billion, including $56.8 billion for DRAM and $30.7 billion for NAND.
  • DRAM and NAND average selling prices rose 275% and 368% year over year, respectively, with year-over-year growth accelerating for the eighth consecutive month for both.
  • Artificial intelligence spending continues to support computing, memory, and networking semiconductors, with related beneficiaries including NVDA, AVGO, AMD, INTC, MU, and MRVL.

Report interpretation

Overview

Based on WSTS data for June 2026, JPMorgan judges that global semiconductor fundamentals remain strong and are beginning to spread from memory chips to more categories such as micro components, analog chips, and sensors and actuators. June growth was mainly driven by shipments, while the 1% month-over-month decline in industry blended average selling prices is viewed as normalization after a sharp 22% increase in May, rather than a weakening of fundamentals. Combined with corporate feedback on orders, backlogs, lead times, and pricing power, the report maintains a positive view on the AI-driven semiconductor upcycle.

Core views

First, June sales growth was significantly above seasonal levels, showing stronger industry demand momentum. Second, memory still accounts for about 77% of the year-to-date increase in industry sales, but non-memory sales growth has accelerated year over year for three consecutive months, indicating a broadening growth base. Third, DRAM and NAND both benefited from bit shipment growth and price increases, driving memory revenue to an all-time high. Fourth, the model expects industry sales to grow to $1.68 trillion in 2026 and further to $2.25 trillion in 2027, and because non-memory forecasts use relatively conservative seasonal assumptions, actual results may exceed forecasts. Fifth, computing, memory, and networking chips benefit the most from incremental AI spending.

Analysis framework

The report is based on WSTS monthly industry sales, shipment, and average selling price data, splitting year-over-year and month-over-month trends by memory versus non-memory and by major product categories, and comparing them with historical seasonal ranges. The forecasting model applies seasonal growth assumptions to most non-memory categories, while using the global research team's more detailed bottom-up forecasts for memory categories.

Methodology notes

  • Industry cycle analysisYear-over-year and month-over-month trend analysis

    Identify the direction of the cycle through year-over-year and month-over-month changes in sales, shipments, and average selling prices.

    June sales growth was mainly driven by shipments, while average selling price changes were also used to determine whether growth came from volume, price, or both.

  • Seasonality analysisHistorical seasonality comparison

    Compare current-month growth with the normal historical level for the same month.

    Historically, semiconductor sales have increased by about 8% month over month on average in June, while growth in June 2026 was 12%, indicating that categories such as micro components, analog chips, and sensors and actuators performed above seasonal levels.

  • Industry forecastingCategorized industry sales forecasting model

    Forecast and aggregate industry sales by major semiconductor category.

    Non-memory categories mainly use conservative seasonal growth assumptions, while memory categories use a more detailed bottom-up approach; therefore, the report believes the overall forecast has upside potential.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

  • NVIDIA Corporation (NVDA)
    A core beneficiary of AI computing chips; the report lists the rating as Overweight.
    Strengths
    Directly benefits from growth in AI computing spending and stronger industry demand momentum.
    Weaknesses
    The report does not provide company-level earnings forecasts or valuation analysis.
    Comparison
    Along with AVGO and AMD, it is one of the main beneficiaries in computing.
    Risks
    Slower AI spending, industry growth below expectations, and valuation volatility after high growth.
  • Broadcom Inc (AVGO)
    Benefits from both AI computing and networking chip spending; the report lists the rating as Overweight.
    Strengths
    Spans two high-momentum areas, computing and networking, with more diversified benefit channels.
    Weaknesses
    The report does not provide a company-level target price or detailed financial forecasts.
    Comparison
    Compared with stocks mapped to only a single theme, AVGO has exposure to both computing and networking growth.
    Risks
    Cloud capital spending, order conversion, and changes in the semiconductor cycle may affect growth.
  • ADVANCED MICRO DEVICES INC (AMD)
    A beneficiary of AI computing growth; the report lists the rating as Neutral.
    Strengths
    Expansion in computing demand and stronger industry order momentum provide fundamental support.
    Weaknesses
    The Neutral rating indicates its risk-reward appeal is weaker than the Overweight-rated names in the report.
    Comparison
    Also benefits from the computing theme, but its rating is lower than NVDA and AVGO.
    Risks
    Competition, product execution, realization of AI demand, and valuation volatility.
  • Intel (INTC)
    Listed as a potential beneficiary of computing demand growth, but the report rating is Underweight.
    Strengths
    Industry computing demand and improving non-memory growth may provide cyclical support.
    Weaknesses
    The Underweight rating suggests company-level risk-reward remains relatively unfavorable.
    Comparison
    Among computing-related names, its rating is lower than NVDA, AVGO, and AMD.
    Risks
    Competitive pressure, execution risk, and failure of industry growth to fully translate into company performance.
  • Micron Technology (MU)
    A direct beneficiary of DRAM and NAND price and revenue growth; the report lists the rating as Overweight.
    Strengths
    Memory revenue reached a record high, and DRAM and NAND average selling price growth continued to accelerate year over year.
    Weaknesses
    Highly sensitive to memory pricing and supply-demand cycles.
    Comparison
    Compared with computing and networking names, MU has greater direct leverage to this round of memory price increases.
    Risks
    Increased memory supply, slower price gains, or weaker end demand.
  • Marvell Technology Inc (MRVL)
    A beneficiary of growth in AI networking spending; the report lists the rating as Overweight.
    Strengths
    Networking infrastructure demand is expected to grow as AI computing clusters expand.
    Weaknesses
    The report does not provide a company-level target price or earnings forecast.
    Comparison
    Together with AVGO, it forms the network chip beneficiary group favored by the report.
    Risks
    Delays in networking capital spending, customer concentration, and order conversion below expectations.

Key data

  • June 2026 industry sales$152.0 billionUp 134% year over year and 12% month over month.
  • Historical average month-over-month growth in JuneAbout 8%The 12% month-over-month increase in June 2026 was significantly above the historical seasonal level.
  • June 2026 non-memory sales$63.5 billionUp 38% year over year, with year-over-year growth accelerating for the third consecutive month.
  • Memory contribution to year-to-date sales growthAbout 77%Memory remains the main source of growth, but non-memory growth is broadening.
  • June 2026 memory revenue$88.4 billionA record high.
  • June 2026 DRAM sales$56.8 billionUp 8% month over month and 373% year over year.
  • June 2026 NAND sales$30.7 billionUp 15% month over month and 377% year over year.
  • DRAM average selling price growthUp 275% year over yearYear-over-year growth accelerated for the eighth consecutive month.
  • NAND average selling price growthUp 368% year over yearYear-over-year growth accelerated for the eighth consecutive month.
  • 2026 semiconductor sales forecast$1.68 trillionExpected to grow 111% year over year, or 30% year over year excluding memory.
  • 2027 semiconductor sales forecast$2.25 trillionExpected to grow 34% year over year, or 16% year over year excluding memory.

Impact & implications

The data indicate that the semiconductor upcycle is no longer relying solely on memory price recovery, but is gradually expanding to categories such as micro components, analog chips, and sensors, helping improve the sustainability of industry growth. AI infrastructure investment remains the core driver, and companies in computing, memory, and networking chips have the greatest revenue and pricing power elasticity. If orders, backlogs, and lead times continue to improve, industry sales could exceed current forecasts; however, categories that are highly dependent on price increases may also experience greater volatility if supply-demand conditions loosen.

Risks

  • The 2026–2027 forecasts rely on the assumption that major non-memory categories broadly follow seasonal growth; actual demand may deviate from the model.
  • Memory still contributes about 77% of the year-to-date increase in industry sales, meaning industry growth remains highly dependent on continued strength in DRAM and NAND prices.
  • If new supply is released, bit shipments decline, or end demand weakens, memory average selling prices may normalize or fall.
  • If AI capital spending slows, orders and backlogs for computing, memory, and networking chips may be below expectations.
  • If corporate lead times and pricing power stop improving, the current industry acceleration trend may be difficult to sustain.
  • The forecasts, views, and prices in the report reflect only information as of the report date, and actual results may differ significantly due to changes in markets, exchange rates, and regulation.

What to watch

  • Whether subsequent WSTS monthly sales can continue to maintain growth above historical seasonality.
  • Whether non-memory sales growth continues to accelerate year over year, and the breadth of improvement in micro components, analog chips, and sensors.
  • Whether DRAM and NAND bit shipments, average selling prices, and revenue growth continue.
  • Whether the month-over-month decline in the industry blended average selling price is indeed normalization after May's sharp increase.
  • Changes in orders, backlogs, lead times, and pricing power at major semiconductor companies.
  • The sustainability of capital spending on AI computing, memory, and networking infrastructure.
  • Upward or downward revisions to the industry sales forecasts of $1.68 trillion in 2026 and $2.25 trillion in 2027.
Zhejiang ICP No. 2022035445-5
Disclaimer: Market data, charts, indicators, research views, and other information provided on this website are intended solely for information display, research communication, and educational reference. They should not be regarded as personalized investment advice, securities recommendations, trading instructions, solicitations, or guarantees of return. While we strive to improve the reliability of our data and content, such information may still be subject to delays, errors, incompleteness, or untimely updates due to source differences, methodological limitations, system processing, or market volatility. Users should exercise independent judgment based on their own circumstances and bear all risks and responsibilities arising from the use of this website.

Settings

Sign in to view recent logins