Is the semiconductor industry growing now?

Why is the semiconductor growth answer so hard right now?

The semiconductor industry is growing right now, but the growth is uneven enough that a simple “yes” would be misleading.

The positive case is very strong. Global chip sales hit $110.5 billion in April 2026, up 93.9% year over year and 11% month over month, according to SIA. Q1 2026 sales were also up 25% from Q4 2025.

Obviously, there are a lot of good things happening in the semiconductor industry these days.

Nvidia reported $81.6 billion in Q1 fiscal 2027 revenue, up 85% year over year, with data center revenue up 92%.

Broadcom said Q2 AI semiconductor revenue grew 143% year over year and guided Q3 AI semiconductor revenue to $16 billion, up more than 200% year over year.

ASML said chip demand is outpacing supply and that customers had raised short- and medium-term demand expectations.

Applied Materials delivered record quarterly performance and lifted its 2026 semiconductor equipment growth expectation to more than 30%.

Memory is also flashing boom signals.

SK hynix reported its first KRW 50 trillion-plus quarter in Q1 2026, with a 72% operating margin.

TrendForce expects conventional DRAM contract prices to rise 58–63% quarter over quarter in Q2 2026, and NAND Flash contract prices to rise 70–75%. That pricing power shows that AI server demand is absorbing enough capacity to reshape the rest of the chip stack.

But we can also find a lot of negative signals.

Consumer and legacy end-markets are not behaving like a broad boom.

IDC said global smartphone shipments fell 2.9% year over year in Q1 2026 because memory shortages and record prices are forcing shipment cuts and price increases.

Omdia expects global smartphone shipments to fall about 7% in 2026, with a downside scenario worse than 15%.

IDC said Q1 2026 PC growth was partly pull-forward demand before price increases, not clean organic demand. STMicroelectronics improved, but still reported only $37 million of net income on $3.10 billion of Q1 revenue.

onsemi’s revenue was down 1% sequentially, even though management said demand had moved beyond the trough.

So, the answer is difficult.

If you want more recent data on this point, please see our latest semiconductor industry report.

What are analysts saying about semiconductor growth?

Semiconductor analysts are now broadly bullish, but their forecasts differ because they are measuring a market being distorted by AI and memory inflation.

Gartner forecasts worldwide semiconductor revenue above $1.3 trillion in 2026, up 64%. Its number is aggressive because it includes a huge memory-price reset: Gartner expects memory revenue to almost triple from $216.3 billion in 2025 to $633.3 billion in 2026, while non-memory grows from $589.0 billion to $686.9 billion. Gartner is basically saying the market is growing through two channels at once: real AI infrastructure demand and “memflation.”

IDC is directionally similar. It forecasts total semiconductor revenue of $1.29 trillion in 2026, up 52.8% from 2025. IDC also puts memory at the center of the move, with DRAM revenue projected to nearly triple to $418.6 billion. Its non-memory forecast is more measured at $693.5 billion. That matters because it shows the industry is not only selling more chips but also earning much more per unit in constrained categories.

WSTS/SIA data is even more forceful in the latest monthly read. SIA reported April 2026 global chip sales of $110.5 billion, up 93.9% year over year, and said annual semiconductor sales are projected to top $1.5 trillion globally in 2026. That estimate reflects the extraordinary late-2025 and early-2026 acceleration in memory and AI infrastructure demand.

Deloitte’s 2026 industry outlook is more cautious in tone. It says AI data centers could represent roughly half of industry revenue in 2026, but warns that the market is heavily exposed if non-data-center categories like PCs, smartphones, and automotive stay weak. That is the right framing: analysts agree the market is growing, but they disagree on how healthy that growth is.

Forecasts capture market value, but they do not fully show whether customers outside AI are buying more units, whether smaller OEMs are being priced out, or whether inventory is being pulled forward.

Analyst forecasts are one layer of evidence. The core of the answer has to come from real-world signals. And this is what we’re checking now!

Is AI chip demand still exploding?

AI chip demand is still exploding, and this is the cleanest growth signal in semiconductors right now.

Nvidia’s latest quarter is the strongest proof. Q1 fiscal 2027 revenue reached $81.6 billion, up 85% year over year and 20% sequentially. Data center revenue reached $75.2 billion, up 92% year over year. That means Nvidia’s growth is not just a year-over-year comparison against an old base. It is still expanding quarter to quarter at enormous scale.

Broadcom gives us a second signal because it is not selling the same product mix. In Q2 fiscal 2026, Broadcom’s AI semiconductor revenue reached $10.8 billion, up 143% year over year, driven by custom AI accelerators and AI networking. It guided Q3 AI semiconductor revenue to $16.0 billion, up more than 200% year over year. That tells us the AI chip boom is spreading beyond GPUs into custom ASICs, networking, and hyperscaler-specific silicon.

So, AI demand is becoming less about “one chip winner” and more about an infrastructure bill of materials. GPUs, custom accelerators, HBM, high-capacity RDIMMs, enterprise SSDs, optical links, power chips, networking silicon, advanced packaging, and lithography tools are all being pulled into the same cycle.

That makes the current expansion broader than a Nvidia-only story, even though Nvidia remains the visible center.

Is memory growth real, or just price inflation?

Memory is growing, but a large part of the current boom is price-led rather than clean unit-led expansion.

SK hynix’s Q1 2026 results show real demand. The company reported KRW 52.5763 trillion in revenue, KRW 37.6103 trillion in operating profit, and a 72% operating margin. Management specifically pointed to AI infrastructure investment, HBM, high-capacity server DRAM modules, and enterprise SSDs. That is not just accounting inflation. AI customers are reserving and consuming high-value memory products.

But the pricing signal is extreme. TrendForce expects conventional DRAM contract prices to rise 58–63% quarter over quarter in Q2 2026, while NAND Flash contract prices are expected to rise 70–75%. Gartner expects DRAM annual prices to increase 125% in 2026 and NAND prices 234%. That means revenue growth is being amplified by scarcity.

A semiconductor market can grow because more chips are being shipped, because better chips are being sold, or because buyers are paying much more for constrained supply.

Right now, memory growth is a mix of all three, but pricing is doing unusually heavy lifting. That still counts as current market expansion, but it is not the same as broad end-demand strength.

If you want more recent data on this point, please see our latest semiconductor industry report.

Are chipmakers expanding capacity right now?

Chipmakers and equipment suppliers are expanding capacity right now, and this is one of the best forward-looking proof points.

ASML said in April 2026 that demand for chips is outpacing supply and that customers are accelerating capacity expansion plans for 2026 and beyond. It guided full-year 2026 sales to €36 billion to €40 billion, after Q1 net sales of €8.8 billion. The key signal is not only ASML’s revenue but that customers are raising expected short- and medium-term demand for lithography systems and upgrades.

Applied Materials is seeing the same thing from a different part of the fab stack. It delivered record quarterly performance and now expects its semiconductor equipment business to grow more than 30% in calendar 2026. Management said AI infrastructure, leading-edge logic, DRAM, and advanced packaging are driving demand.

Texas Instruments adds a more mature-node angle. TI reported Q1 2026 revenue of $4.83 billion, with analog revenue up 22% year over year and embedded processing revenue up 12%. The company is also ramping U.S. 300mm capacity. That matters because the expansion is not limited to 3nm logic and HBM. Foundational analog and embedded chips are also moving into a recovery phase.

The caveat is timing. Capacity expansion is a confidence signal, but fabs take years to build and ramp.

Are consumer chips growing too?

Consumer chip demand is not broadly growing right now. Actually, it is being squeezed by the AI-driven memory shock.

Smartphones are the clearest weak signal. IDC said global smartphone shipments fell 2.9% year over year to 293.8 million units in Q1 2026, breaking a 10-quarter growth streak. IDC directly tied the slowdown to limited memory supply and record memory prices, which are forcing OEMs to reduce shipments and raise prices.

Omdia is more negative for the full year. It forecasts global smartphone shipments will decline around 7% in 2026, with downside risk above 15% if memory prices keep rising and geopolitical volatility worsens. Omdia also expects sub-$100 smartphones to decline nearly 31% year over year. That is a brutal signal for volume chips, low-end components, and price-sensitive emerging markets.

PCs look healthier on the surface, but the signal is not clean. IDC said Q1 2026 personal computing devices grew 2.8% year over year, with PC shipments up 3%. But IDC also said this was largely pull-forward demand as buyers accelerated purchases before memory price increases and product shortages. That means Q1 growth may have borrowed from later quarters.

The important conclusion is that AI is not lifting every semiconductor end-market. In consumer electronics, AI demand is actually crowding out supply and raising costs. That creates a strange market: semiconductor revenue rises, while some device shipments fall.

If you want more recent data on this point, please see our latest semiconductor industry report.

Are automotive and industrial semiconductors recovering?

Automotive and industrial semiconductors are recovering, but the recovery is still uneven and less explosive than AI.

Texas Instruments gives us a positive analog signal. Q1 2026 revenue was $4.83 billion, up materially from the prior year, with analog revenue up 22% and embedded processing up 12%. Management commentary also pointed to improving industrial demand, while data center demand remained strong.

STMicroelectronics shows recovery with margin pressure. Q1 2026 revenue was $3.10 billion, up 23% year over year, and management said bookings were strong and distribution inventory had normalized. But net income was only $37 million, operating margin was 2.3%, and gross margin was 33.8%. That is growth, but not a boom.

onsemi is somewhere in the middle. Q1 2026 revenue was $1.513 billion, up 5% year over year but down 1% sequentially. Management said demand strengthened through the quarter and that the company had moved beyond the cyclical trough. AI data center revenue grew more than 30% sequentially and more than doubled year over year, but some core segments remained soft.

Are semiconductor equipment orders confirming the boom?

Equipment demand confirms that the boom is real, because customers are spending ahead of future capacity.

ASML’s Q1 2026 statement is unusually direct: demand for chips is outpacing supply, customers are accelerating capacity plans, and order intake remains very strong. That is a hard signal because lithography tools are expensive, capacity-constrained, and ordered only when customers have confidence in future wafer demand.

Applied Materials is also raising its build plan, inventory positions, and logistics capacity. That is not the behavior of a supplier waiting for a soft market. It means customers are asking for more tools and longer visibility.

The equipment signal also tells us where the growth is concentrated. Applied called out leading-edge logic, DRAM, and advanced packaging. Those are exactly the areas needed for AI accelerators, HBM, high-capacity servers, and custom silicon. Equipment demand is therefore confirming the same story as chip revenue: AI infrastructure is pulling the whole front end and back end of the supply chain upward.

Are advanced packaging and interconnect becoming a growth engine?

Advanced packaging and interconnect are becoming one of the most important growth engines in semiconductors.

We see this indirectly across the strongest signals. Nvidia’s AI systems require advanced packaging and high-bandwidth memory integration. Broadcom’s custom accelerator and AI networking growth depends on high-speed interconnect. Applied Materials specifically names advanced packaging as part of the growth base for its equipment business. Startup funding is also flowing into chip-to-chip and data-center-level interconnect.

The startup evidence is especially telling. Semiconductor Engineering reported that Q1 2026 private semiconductor funding included 80 startups raising $8.4 billion, with many working on AI inference, bandwidth limits, and interconnect from chip level to data center level. Crunchbase said semiconductor startups had raised around $10.7 billion in 2026 through June 10, with Ayar Labs raising $500 million for optical interconnect technology.

This is one of the more under-discussed growth signals. The industry is not only scaling more wafers; it is rebuilding the system architecture around bandwidth. That favors HBM, silicon photonics, chiplets, advanced substrates, and packaging tools. In practical terms, the bottleneck is shifting from “can we make chips?” to “can we connect enough compute and memory at low enough power?”

Are startups still getting funded?

Semiconductor startups are still getting funded, and the funding pattern is highly specific: AI inference, photonics, EDA automation, and bandwidth.

Semiconductor Engineering counted 80 private semiconductor startups raising $8.4 billion in Q1 2026, including 18 mega-rounds above $100 million. It also noted two billion-dollar-scale rounds, Rapidus and Cerebras. That is not a dead private market.

Crunchbase’s June 2026 snapshot says investors had put around $10.7 billion into seed through pre-IPO semiconductor rounds so far in 2026, on pace to exceed last year. Notable recent rounds include MatX raising a $500 million Series B for AI-lab-specific chips and Ayar Labs raising a $500 million Series E for optical AI infrastructure.

The nuance is that investors are not broadly funding “semiconductors” in a generic way but actual important bottlenecks: inference cost, memory bandwidth, optical I/O, EDA productivity, and supply-chain localization.

That is healthier than hype-only funding because it maps to visible pain points in the current cycle.

Are public semiconductor companies reporting broad strength?

Public semiconductor companies are reporting strength, but it is concentrated in AI-linked revenue pools.

Nvidia, Broadcom, SK hynix, ASML, and Applied Materials all reported recent numbers or commentary consistent with live expansion. Nvidia’s data center revenue is still growing at extreme scale. Broadcom’s AI semiconductor revenue is accelerating. SK hynix’s margin profile looks like a supercycle. ASML and Applied are seeing customers increase capacity plans.

The less AI-centric names are more mixed. TI’s analog and embedded businesses are improving. STMicroelectronics is recovering but still operating with thin net income. onsemi is past the trough but only modestly up year over year and down sequentially.

This split is the whole market in miniature. Companies exposed to AI compute, memory, networking, advanced packaging, and equipment are in growth mode. Companies tied more heavily to automotive, industrial, low-end consumer, and legacy device volumes are recovering slowly or facing cost pressure. The public market is strong, but not evenly strong.

Are customers increasing semiconductor spending?

Customers are increasing semiconductor spending, but in many cases they are doing it defensively.

Cloud and AI customers are clearly spending more. Nvidia’s and Broadcom’s results show hyperscalers and AI infrastructure builders are buying accelerators, custom silicon, networking, and systems at much higher levels. TrendForce also says North American cloud service providers are accelerating AI inference deployments and negotiating long-term agreements for memory supply.

Device OEMs are spending more on memory too, but that is not the same kind of growth. PC and smartphone makers are paying more because supply is tight. IDC said PC growth in Q1 2026 was partly driven by pull-forward buying before price increases. IDC and Omdia both show smartphone vendors facing shipment cuts and price hikes due to memory pressure.

So spending is up, but the interpretation depends on the buyer. Hyperscalers are spending to expand AI capacity. Consumer device makers are spending to avoid shortages. The first is demand expansion. The second is margin defense.

If you want more recent data on this point, please see our latest semiconductor industry report.

Are semiconductors seeing real unit growth?

Semiconductor unit growth is harder to prove than revenue growth right now.

SIA’s revenue data is extremely strong, but revenue includes price effects. Memory pricing is rising so sharply that some market growth is coming from higher average selling prices rather than more end devices. Gartner and IDC both show memory driving a disproportionate share of 2026 revenue growth.

End-device units tell a weaker story. Smartphones fell in Q1 2026, according to IDC. Omdia expects the full-year smartphone market to decline. PCs grew in Q1, but IDC said the result was inflated by pull-forward demand. That makes it risky to infer broad unit growth from semiconductor revenue alone.

The best unit-growth signal is probably in AI infrastructure rather than consumer devices. More AI servers, more accelerators, more HBM stacks, more enterprise SSDs, more networking ports, and more power components are being deployed. But those units are concentrated in high-value systems, not mass-market devices.

That is why the market can be growing very fast in dollars while remaining mixed in units. The current semiconductor expansion is value-heavy, not volume-heavy.

Are regulations helping or hurting semiconductor growth?

Regulation is both helping and hurting semiconductor growth, depending on geography and product category.

Industrial policy is helping capacity buildout. The U.S. CHIPS Act, European chip strategies, Japanese support for Rapidus, and Indian design incentives all continue to push semiconductor localization. Those programs create demand for fabs, tools, materials, packaging, workforce development, and local supply chains.

But export controls and trade restrictions are also adding friction. NXP explicitly lists tariffs, export restrictions, supply-chain disruptions, Middle East conflict, and trade-policy changes as business risks. ASML remains exposed to China restrictions, even though AI demand elsewhere is strong. Advanced AI chips and semiconductor manufacturing tools remain politically sensitive.

The practical effect is that regulation increases spending while reducing efficiency. Governments are subsidizing local capacity because chips are strategic, but geopolitics also fragments markets, limits some sales, and forces duplication. For growth, that means more capex and more regional supply-chain buildout. For margins and planning, it means more uncertainty.

Are layoffs and restructuring contradicting the boom?

Layoffs and restructuring do not disprove growth, but they show the boom is not uniform.

The strongest AI-linked companies are investing aggressively, but weaker or transitioning segments are still cutting costs. STMicroelectronics has been restructuring after a difficult automotive and industrial cycle. Intel has been working through a multi-year turnaround in foundry and product competitiveness. Smaller device-linked suppliers face pressure as memory costs squeeze smartphones and PCs.

This is normal in a bifurcated market. Companies with the right exposure are expanding. Companies with the wrong mix are protecting margins, exiting lower-value revenue, or restructuring. onsemi, for example, says it is beyond the trough and seeing AI data center growth, but its Q1 revenue was still down sequentially.

The question here is whether layoffs are broad enough to offset growth signals. Right now, they are not. They show rotation inside the market, not industry-wide contraction.

Are M&A and partnerships showing confidence?

M&A and partnerships are showing confidence, especially around design complexity, AI systems, and supply assurance.

Synopsys completed its acquisition of Ansys in July 2025, creating a larger silicon-to-systems design and simulation platform with a stated $31 billion total addressable market. That matters right now because AI chips are increasingly systems problems, not just transistor problems. Thermal behavior, power delivery, packaging, photonics, and software all need to be optimized together.

Nvidia and SK hynix have also moved toward deeper memory roadmap alignment, reflecting the importance of long-term HBM and advanced memory planning. Broadcom’s AI growth is tied to long-cycle custom silicon relationships with hyperscalers. Memory suppliers are negotiating long-term agreements with key customers, according to TrendForce.

So, clearly, companies are locking supply, roadmaps, and design tools earlier. That usually happens when demand is strong enough that waiting becomes risky.

Is China demand still supporting the market?

China demand is still important, but, to be honest, it is more complicated than a simple growth driver.

China remains a major semiconductor buyer and a major source of mature-node capacity investment. Equipment suppliers have benefited from China’s past spending, while export controls have encouraged Chinese firms to build domestic alternatives. That supports some demand for tools, materials, and local design activity.

At the same time, restrictions on advanced chips and tools limit the upside for U.S. and European suppliers. ASML, Applied Materials, Lam Research, KLA, Nvidia, AMD, and other advanced-chip ecosystem players all face policy risk around China. NXP also flags tariffs and export restrictions as risks.

So China is not disappearing from semiconductor demand. Rather, it is fragmenting.

Is the market growing because supply is tight?

A meaningful part of the semiconductor market is indeed growing because supply is tight, and that is both bullish and risky.

Tight supply is clearly visible in memory. TrendForce says DRAM suppliers are reallocating capacity toward HBM and server applications, while NAND capacity is increasingly allocated to enterprise SSDs. That pushes prices higher across conventional DRAM and NAND, even where end-device demand is weaker.

ASML also says chip demand is outpacing supply and customers are accelerating capacity expansion plans. Applied Materials is increasing build plans and logistics capacity. These are classic tight-market behaviors.

But tight supply can create deceptive growth. Revenue rises because customers pay higher prices, reserve supply earlier, and pull orders forward. If downstream demand weakens, some of that revenue can later unwind. The strongest evidence of sustainable growth is therefore not just pricing. It is the combination of tight supply, long-term agreements, hyperscaler capex, and capacity expansion.

Right now, all four are present in AI infrastructure and memory. They are not equally present in smartphones or low-end consumer devices.

If you want more recent data on this point, please see our latest semiconductor industry report.

Is the semiconductor industry growing right now?

The semiconductor industry is growing right now, but it is an AI-and-memory-led expansion rather than a broad-based device-cycle recovery.

The direct evidence is too strong to call the market flat or declining.

SIA’s April 2026 sales were up 93.9% year over year and 11% month over month. Nvidia, Broadcom, SK hynix, ASML, and Applied Materials all show live demand expansion. Analysts from Gartner and IDC now forecast 2026 semiconductor revenue well above $1 trillion, with growth rates above 50% in their latest views. Startup funding is also active, especially around AI inference, photonics, interconnect, and EDA.

But the market’s health is uneven.

Smartphones are shrinking. PC growth is partly pull-forward. Automotive and industrial semiconductors are recovering, not booming. Memory inflation is lifting revenue while pressuring consumer OEMs. Export controls and geopolitics are adding friction. Some companies are still restructuring.

So, yes, the semiconductor industry is growing right now.

But it is growing through a concentrated AI infrastructure supercycle, memory scarcity, and capacity expansion, while several legacy and consumer-linked segments remain under pressure. If we measure by revenue, the market is clearly expanding. If we measure by broad end-device unit demand, the answer becomes mixed.

Question	Verdict	Comment
What do analysts say?	Yes	Gartner, IDC, and SIA/WSTS all point to major 2026 revenue expansion.
Is AI chip demand exploding?	Yes	Nvidia and Broadcom show very strong AI data center and custom silicon growth.
Is memory growth real?	Mixed	HBM/server demand is real, but price inflation is doing heavy lifting.
Are chipmakers expanding capacity?	Yes	ASML and Applied Materials both report stronger capacity-expansion demand.
Are consumer chips growing?	No	Smartphones are declining, and PC growth is partly pull-forward demand.
Are automotive and industrial chips recovering?	Mixed	TI, ST, and onsemi show recovery, but not broad acceleration.
Are equipment orders confirming growth?	Yes	Lithography and process-equipment suppliers are raising expectations.
Is advanced packaging becoming a growth engine?	Yes	AI systems are increasing demand for packaging, HBM integration, and interconnect.
Are startups still getting funded?	Yes	Q1 and 2026-to-date funding data show active AI-chip and photonics investment.
Are public chip companies broadly strong?	Mixed	AI-exposed names are booming; legacy and device-linked names are uneven.
Are customers spending more?	Mixed	Hyperscalers are expanding capacity; device OEMs are spending defensively.
Are chip units really growing?	Mixed	Revenue is clearly up, but unit evidence is weaker outside AI infrastructure.
Are regulations helping growth?	Mixed	Subsidies support capacity, while export controls and tariffs add friction.
Do layoffs contradict the boom?	Mixed	Restructuring shows rotation, not an industry-wide decline.
Are M&A and partnerships bullish?	Yes	Supply agreements, EDA consolidation, and AI roadmap partnerships show confidence.
Is China still supporting demand?	Mixed	China supports mature-node/local demand, but restrictions cap advanced-chip upside.
Is tight supply driving growth?	Mixed	Scarcity supports revenue and capex, but can overstate organic end-demand strength.
Is the market growing now?	Yes	The industry is expanding now, led by AI infrastructure, memory, and equipment.

Who is the author of this content?

NEW MARKET PITCH TEAM

We track new markets so founders and investors can move faster

We build living "market pitch" documents for emerging markets: AI, synthetic biology, new proteins, and more. Instead of outdated PDFs or hallucinated LLM answers, our clients get a clean, visual, always-updated view of what's really happening: key players, deals, regulations, and signals that matter. Learn more about us.