Memory Shortage Impacts Cars, Smartphones & Appliances

Market Pulse Metrics

At-a-glance benchmarks, each traceable to a government, association or company filing.

Read these as value measures. WSTS states its memory figure is revenue growth, and SIA reports dollar sales, so much of the 2026 rise reflects price rather than additional supply.

Executive Summary

The state of the shortage

Between October 2025 and July 2026, SIA's three-month average of global chip sales rose from $72.7B to $146.8B a month. Q2 2026 alone totalled $403.3B, 35.1% above Q1. WSTS's August update puts the full-year 2026 market at about $1.655 trillion and raises its memory-segment growth estimate to 302%.

The same upstream strength shows in Korea, the world's largest memory exporter. Chip exports reached $46.65B in August 2026 and made up 47.5% of the country's total exports. In the first half they totalled $192.4B, already 11% above 2025's full-year record of $173.4B.

Why this matters beyond servers

When memory is scarce, makers ration it. Entry-level phones and low-margin appliances carry the least pricing power, so they feel allocation first. The question for buyers is not only "what does memory cost" but "who gets supply". This report separates the segments where that question already has data behind it from those where it does not.

How large is the move?

Measured from the start of the SIA series used in this report, the three-month average of monthly global chip sales roughly doubled in nine months, from $72.7B in October 2025 to $146.8B in July 2026 (NextMSC calculation: 2.02 times). The climb was not smooth. Monthly gains stayed between $2.6B and $3.6B through January 2026, rose to $6.3B in February and $10–11B through March to May, and reached $17.3B in June, the largest step in the series, before easing to $8.9B in July.

Each SIA reading is a three-month average that is later revised, so month-to-month changes are smoothed and one month should not be read as a turning point. June itself was first reported at $134.5B and later revised to $137.9B. The July slowdown, from +14.3% on the month in June to +6.5% (NextMSC calculations), is worth watching, but it is one data point.

The table below restates the scale of the move using only figures reported elsewhere in this article. Percentage changes and ratios are NextMSC calculations, and they inherit the revision risk of the underlying series.

Source: SIA, WSTS, SEC Form 8-K filings, U.S. Bureau of Labor Statistics, MOTIR and NextMSC analysis.https://www.nextmsc.com/industry-deep-dive/memory-shortage-cars-smartphones-appliances

Key Drivers: why scarcity travels downstream

Allocation, not only price

Reuters reporting of 16 September 2026 quotes executives at three independent device makers saying access to memory is now the binding constraint. Fairphone's CEO said memory can be almost 60% of the bill of materials on a roughly $400 handset. TrendForce, cited in the same report, expects DRAM contract prices to rise 13–18% this quarter versus 93–98% in the first, which suggests price momentum is easing while scarcity is not.

Supply is booked ahead

Micron's Q3 FY2026 prepared remarks state it had signed 16 Strategic Customer Agreements by June 2026, covering about 20% of DRAM and about a third of NAND volume over their term. Micron also said server DRAM content growth is being trimmed as customers maximise unit shipments amid very tight allocation. These are company statements, not independent market data.

Capacity is not an immediate relief valve

Micron guided fiscal 2026 capital expenditure at about $27B. New capacity takes years to reach output, and SK Hynix's CEO said in July that 2027 would be the industry's worst supply year (as reported by Reuters). Executives at the smaller device makers quoted by Reuters expect scarcity to last through 2027.

Why sales, price and revenue measures differ

SIA reported global chip sales up 135.1% year on year in July 2026, while the BLS index covering semiconductors and electronic components rose 26.7% year on year in August. These are different measures. SIA counts industry revenue in dollars, the BLS index tracks producer prices for all semiconductors and components, and WSTS's +302% memory-segment estimate is likewise revenue growth, not price.

NextMSC analysis: the gap between the sales and index growth rates is consistent with memory revenue growing much faster than the broad index, but this report holds no free, memory-only price series that would confirm it. For that reason every price statement here is anchored to the BLS series and labelled with its scope.

Supplier revenue shows the same shape

Micron's fiscal Q4 2026 revenue of $54.23B was 379.1% above the $11.32B of a year earlier (NextMSC calculation). Sequential growth was 74.9% in Q2 FY2026, 73.8% in Q3 and 30.8% in Q4, so the ramp is steep but already slowing in percentage terms even as the dollar level keeps rising. The company's guidance midpoint of $61.5B implies a further 13.4% step, and guidance is not an actual result.

Micron's remarks on Strategic Customer Agreements and its capital expenditure guidance of about $27B are company statements. They indicate how the supplier is committing volume and capacity, but they are not independent confirmation of how much supply reaches the wider market.

Spillover Evidence Matrix

NextMSC compiled this matrix to separate what is documented from what is assumed. It rates the strength of evidence, not the importance of the segment.

Source: MOTIR, Reuters, Detroit News, S&P Global Mobility, Yicai, Micron and NextMSC analysis.https://www.nextmsc.com/industry-deep-dive/memory-shortage-cars-smartphones-appliances

Figures: sales, supplier revenue and prices

What the figures show

Figure 1 shows the SIA three-month average rising in every month from October 2025 to July 2026, with June's revised $137.9B the largest single step and July's $146.8B the latest reading. Because the series is revised, the most recent bars carry the most uncertainty.

Figure 2 shows Micron revenue between $8.05B and $13.64B for the first four quarters, then $23.86B in Q2 FY2026, $41.46B in Q3 (3.04 times Q1 FY2026, a NextMSC calculation) and $54.23B in Q4. The gold bar for Q1 FY2027 is guidance of $61.5B ±$1.5B, a range of $60.0B to $63.0B, and not an actual.

Figure 3 shows the BLS index at 58.3 in January 2025 and 61.6 in January 2026, then 67.0 in February 2026 (+8.8% on the month) and 75.5 in August. Nearly a third of the 17.2-point rise since January 2025 came in the single month of February (NextMSC calculation), which suggests a repricing step rather than a gradual drift. The index covers all semiconductors and components, not memory alone.

Segment Analysis: what the evidence supports

Smartphones and PCs: the entry tier absorbs the shock first

When memory is close to 60% of a roughly $400 handset's parts cost, allocation decides which models ship. Reuters reports smaller makers dropping low-end models, adding paid memory upgrades or designing alternate boards. Micron's own outlook is that PC and phone industry revenue grows even as units fall, which fits a market shifting toward higher-priced devices. Implication: buyers of budget-tier devices face more supply risk than buyers of premium tiers.

The Counterpoint figure (−11% in Q2, the lowest Q2 since 2013) and the June full-year forecast (−13.9% to 1.08B units) measure different periods and come from secondary reporting, so they are shown separately rather than combined. Neither isolates memory, since other factors also move shipments, which is why the evidence matrix lists memory-only attribution as not yet shown.

If memory is close to 60% of a roughly $400 handset's parts cost, that is about $240 per unit (NextMSC calculation), leaving little room to absorb a price increase through margin. This is consistent with makers redesigning boards, dropping low-end models or charging for memory upgrades.

The same allocation logic extends to the PC market, which we cover in more detail in AI memory shortage 2026 and PC prices.

Cars: cost today, output risk to monitor

Automakers' public statements through February 2026 describe DRAM as a cost headwind that could be managed, and S&P Global Mobility did not expect a light-vehicle output impact. The data behind this report documents no memory-driven production cut. Korea's August auto exports fell 29.8%, but MOTIR attributes that to holiday timing and partial strikes. NextMSC therefore treats cars as exposed but unproven. Software-heavy newer models carry more memory, so Q3 earnings commentary is the next test.

Ford said DRAM and inflation would add about $1B to its 2026 costs and that it expected an offset. GM said it saw no production impact, and S&P Global Mobility expected none for light-vehicle output. These are cost-management and outlook statements, not output warnings, and all three date from December 2025 to February 2026, before much of the rise in the BLS index, so they should be re-tested against Q3 commentary.

Korea as a control case

Korea's August chip exports of $46.65B were about 12 times its auto exports of $3.85B (NextMSC calculation). Autos fell 29.8% in the same month, but MOTIR attributes the fall to shifted holidays and partial strikes. Treating that decline as evidence of a memory squeeze would mix two unrelated stories, so this report keeps it as a control and not as support.

Appliances: price pass-through first

Chinese makers reported wholesale price increases of 2–10% in April 2026, with Hisense explicitly citing memory costs; other inputs, including oil, were also cited. No volume data tie appliance output to memory supply. Implication: expect margin and pricing effects before supply effects.

The reported increases cover a range: TCL wholesale prices rose 2–10%, Skyworth about 10% and Hisense TVs 3–10%. They come from April 2026 press reporting, and oil is also cited as a cost, so the memory share of each increase cannot be isolated. The reading that fits the evidence is that appliance makers with thin margins pass costs on before they cut output.

Implications by Stakeholder

NextMSC analysis (not a forecast of record): the points below follow from the evidence above and are not additional data.

Device makers

Reuters reports that smaller makers now struggle with availability more than price. Makers outside large supply agreements, such as Micron's 16 Strategic Customer Agreements covering about 20% of its DRAM and about a third of its NAND volume, may have less visibility on supply. Planning around allocation, including memory-light redesigns and tiered configurations, is already visible in the reported behaviour of smaller makers.

Automakers and appliance makers

For both groups the evidence so far describes cost. The useful signal is a change in language on earnings calls from cost and pricing to production and availability. Until that appears, the prudent position is to treat the exposure as real but unproven and to track it quarterly.

Procurement teams and buyers

With TrendForce expecting DRAM contract-price rises of 13–18% this quarter against 93–98% in the first, the immediate pressure point is allocation and contract renewal terms, not headline prices. Buyers of budget-tier devices carry more supply risk than premium-tier buyers, so renewal timing and secondary-source options deserve attention.

Investors and analysts

The key discipline is separating value from volume and actuals from guidance. SIA and WSTS report dollars, Micron's $61.5B is guidance, and the BLS index is not memory-specific. Percentage growth is already slowing at Micron and in the SIA series, so the next prints test whether the level holds, not whether the rate of increase continues.

Outlook: what stakeholders should monitor

NextMSC analysis (not a forecast of record): price momentum appears to be easing while availability stays tight, so the next stress points are allocation and contract renewals rather than headline spot prices.

• Micron FQ1 2027 results against the $61.5B ±$1.5B guidance.
• BLS PPI release due 15 October 2026 for the August-to-September move.
• SIA August and September sales, and WSTS's next forecast update.
• Auto and appliance Q3 earnings calls for any shift from cost language to production language.
• DRAM contract-price commentary from TrendForce, which is published commercially.

Signals that pressure is easing

Three indicators in this report point to slower momentum: TrendForce's 13–18% expected DRAM contract rise this quarter against 93–98% in the first, Micron's guided 13.4% step against 30.8% in the quarter just reported, and the SIA step of +6.5% in July against +14.3% in June (NextMSC calculations). Easing price momentum is not the same as easing availability.

Signals that scarcity is deepening

SK Hynix's CEO said in July that 2027 would be the industry's worst supply year, as reported by Reuters, and executives at the smaller device makers expect scarcity to last through 2027. New capacity takes years to reach output, and signed multi-year agreements commit part of the volume that does exist. Any further revision of the WSTS memory estimate above 302% would also point to a tighter market.

What would change the view on cars and appliances

A move from cost language to production language in Q3 automaker or appliance-maker commentary, or any source documenting a memory-driven production cut, would change the current "exposed but unproven" reading. The BLS release due 15 October 2026 and the Q3 earnings calls are the next checkpoints.