SMIC and Hua Hong Beat Expectations, Turning Foundry Earnings Into Technology News

SMIC and Hua Hong reported second-quarter results above their own guidance on August 13, giving technology news readers a measurable shift in China’s chip market. Revenue, pricing, utilization, and margins all moved in the same direction. That combination matters more than the viral headline alone.

SMIC generated about $3.01 billion in quarterly revenue, roughly 20% more than during the first quarter. Its gross margin reached 25.3%, well above its earlier guidance of 20% to 22%. Hua Hong posted record revenue of $717.5 million and a 16.5% gross margin, also exceeding its forecast.

The results challenge the view that China’s foundry expansion is producing capacity without adequate pricing power. However, they do not show that SMIC or Hua Hong has closed the technology gap with TSMC. The stronger conclusion is narrower: demand for locally manufactured mature and specialty chips has become profitable enough to support higher prices.

That distinction establishes the central contest. SMIC and Hua Hong are strengthening a domestic foundry system built around broad local demand. TSMC remains overwhelmingly dominant in advanced manufacturing, particularly for high-performance computing and artificial intelligence processors.

The next test is whether China’s foundries can preserve their new margins as more factories enter production. Higher depreciation, export restrictions, and uneven end-market demand could still weaken the earnings story.

The Technology News Is in the Margins

Both companies delivered more than a routine revenue beat because higher sales arrived alongside better pricing and stronger profitability.

SMIC had previously forecast second-quarter revenue growth of 14% to 16% from the first quarter. Actual growth reached about 20%, taking revenue from $2.51 billion to approximately $3.01 billion. The company’s gross margin rose from 20.1% to 25.3%.

The margin result was the clearest surprise. A foundry’s gross margin measures the share of revenue remaining after direct production costs. It reflects pricing, factory utilization, product mix, manufacturing yields, and depreciation expenses.

A revenue increase driven only by additional factory output can produce disappointing profits. Every new fabrication plant brings substantial equipment and depreciation costs. SMIC’s results instead indicate that greater shipment volume arrived with higher average selling prices and lower unit costs.

The company’s official quarterly results provide the primary record for its operating performance. Its reported figures also show why the market focused on margins rather than headline revenue alone.

SMIC’s quarterly wafer capacity reached approximately 3.06 million eight-inch-equivalent wafers. An eight-inch-equivalent wafer is a standardized measure used to compare production across different wafer sizes. Capacity increased as new manufacturing lines continued entering service.

Shipments rose by about 14% from the previous quarter, while revenue grew faster. That difference points to a higher average selling price, a better product mix, or both. Reported pricing increases among selected mature-node customers began contributing to results after negotiations during 2025.

Mature nodes are established manufacturing processes used for products such as microcontrollers, power-management chips, display drivers, connectivity components, and automotive electronics. They are less advanced than processes used for leading AI accelerators. Yet they remain essential across consumer, industrial, and transportation systems.

SMIC guided third-quarter revenue to increase another 2% to 4%. It also forecast a gross margin between 26% and 28%, above the second-quarter result. That guidance suggests management expects favorable pricing and utilization to continue through September.

Hua Hong delivered the same directional signal from a smaller base. Its second-quarter filing reported record revenue of $717.5 million. Revenue increased 26.8% from one year earlier and 8.6% sequentially.

Hua Hong had forecast revenue between $690 million and $700 million. Its actual figure exceeded the top of that range by $17.5 million. Gross margin reached 16.5%, above its 14% to 16% forecast.

Profit attributable to the parent company’s shareholders reached $38.6 million. That represented an 84.6% sequential increase and a 385.9% rise from the prior-year period. Total company profit was lower because losses attributable to non-controlling interests affected the consolidated result.

Hua Hong’s overall utilization reached 102.8%, up from 99.7% in the first quarter. Utilization can exceed 100% when actual production surpasses a factory’s estimated standard capacity. That often reflects overtime, optimized production flows, or temporarily higher output.

The company shipped 1.54 million eight-inch-equivalent wafers during the quarter. Shipments increased 17.9% from a year earlier and 5.8% sequentially. Once again, revenue grew faster than shipment volume.

These figures verify the event behind the social-media claim. Both companies released their results on August 13, 2026. Both exceeded their own revenue and margin guidance. The underlying event was not an undated rumor or an isolated market estimate.

Mature-Node Pricing Changed the Earnings Equation

The central reversal is that mature-node capacity, often described as oversupplied, is now supporting better prices at China’s largest foundries.

China has invested heavily in semiconductor manufacturing capacity. Much of that expansion targets mature nodes because advanced production equipment remains harder to obtain. Critics have warned that simultaneous expansion by several foundries could create excess supply and destructive price competition.

That risk has not disappeared. However, the latest results show a different near-term mechanism. Inventory rebuilding, domestic sourcing, and demand for memory-adjacent components increased factory loads. Higher utilization then spread fixed production costs across more wafers.

SMIC also benefited from price increases negotiated with selected customers. In June, foundry market data showed that some price increases agreed with eight-inch customers during 2025 were already taking effect.

That evidence predates the second-quarter release. It makes the latest margin improvement less likely to be an unexplained accounting fluctuation. The earnings numbers show that the pricing trend carried into a much stronger quarter.

The composition of demand matters. SMIC cited strength across industrial, automotive, personal-computer, and tablet applications. These categories require many supporting chips, even when the finished product’s central processor uses a more advanced manufacturing process.

A laptop needs power-management integrated circuits, display components, connectivity chips, controllers, and embedded memory. An electric vehicle requires microcontrollers, power devices, sensing components, and analog chips. Most do not require the smallest available transistor geometry.

Hua Hong’s results provide a detailed view of that market. Embedded non-volatile memory generated $200.1 million, an increase of 41.8% from the previous year. This category includes memory placed inside microcontrollers and smart-card chips, allowing stored information to remain when power is removed.

Standalone non-volatile memory revenue reached $68.8 million, rising 149.3%. The company attributed that growth mainly to flash demand. Analog and power-management revenue increased 13% to $183.1 million, while power-discrete revenue climbed 9.4% to $182.3 million.

Revenue from processes at 65 nanometers and below reached $205.1 million. It rose 63.4% from one year earlier, supported by flash and power-management demand. Hua Hong’s 90-nanometer and 95-nanometer revenue increased 19.9% to $174.4 million.

These figures show a broad improvement rather than one unusually large product order. Every technology platform recorded year-over-year growth. However, the strongest gains occurred in memory-related categories, where supply conditions and AI infrastructure investment have affected the wider component chain.

Hua Hong Chairman and President Peng Bai said rising volumes and prices drove the company’s improvement. He also said AI demand moved from memory products into related logic and analog products. That is a company assessment, not independent proof that every category is experiencing an AI boom.

The end-market data support a more cautious interpretation. Consumer electronics generated $481.3 million, or 67.1% of Hua Hong’s quarterly revenue. Industrial and automotive customers contributed $162.1 million, while communications revenue declined 12.8%.

Computing revenue grew 48.1%, but it remained only $11.7 million. Hua Hong is therefore benefiting mostly from the surrounding component economy, not from manufacturing leading AI accelerators.

That difference does not weaken the earnings result. It explains the mechanism. AI spending can tighten memory supply, stimulate server-related components, and encourage inventory building without converting every foundry into an advanced AI-chip producer.

The durable question is whether this secondary demand produces stable orders. Inventory restocking can raise shipments for several quarters before customers reduce purchasing. Genuine end-market consumption lasts longer and gives foundries greater pricing leverage.

SMIC and Hua Hong Still Face TSMC’s Scale

Stronger Chinese foundry earnings narrow a profitability problem, but they do not erase the gap in advanced manufacturing.

SMIC held about 5.1% of the global foundry market during the first quarter of 2026, according to TrendForce. It ranked third behind TSMC and Samsung. Hua Hong Group held approximately 2.5%, placing it sixth.

TSMC controlled about 72% of the market during the same period. That share reflects its commanding position in advanced processes used for AI accelerators, smartphone processors, server chips, and other high-performance products.

The revenue contrast is equally large. TSMC generated $40.2 billion during the second quarter, according to its quarterly disclosure. SMIC generated approximately $3.01 billion, while Hua Hong produced $717.5 million.

TSMC’s gross margin reached 67.7%, compared with 25.3% for SMIC and 16.5% for Hua Hong. Its two-nanometer process already represented 3% of quarterly wafer revenue. Advanced-node demand also helped high-performance computing reach 66% of its revenue.

Those comparisons define the primary competitive reality. SMIC and Hua Hong can improve quickly without approaching TSMC’s revenue, margins, or leading-node position. Their current advantage is access to a large domestic market seeking reliable supplies of established chips.

Domestic sourcing has commercial and strategic value. Chinese device manufacturers face continuing uncertainty around trade rules and access to foreign technology. Qualifying local foundries can reduce supply-chain exposure, even when their processes are not the industry’s most advanced.

SMIC can serve some relatively advanced domestic designs alongside a much larger mature-node business. Hua Hong focuses more heavily on specialty processes, including embedded memory, power devices, analog products, and microcontrollers.

That creates complementary roles. SMIC offers greater scale and a broader process range. Hua Hong concentrates on markets where specialized manufacturing characteristics matter as much as transistor density.

Hua Hong is also pursuing the acquisition of Huali Microelectronics. The company says the transaction will expand its technology portfolio, operating scale, and profitability. Regulatory registration had been obtained by the second-quarter announcement, although integration remained an execution challenge.

The proposed combination could strengthen Hua Hong’s 12-inch production base and broaden its process offerings. It could also add depreciation, integration costs, and management complexity. Investors should not assume that greater capacity automatically creates equivalent demand.

SMIC faces a related tension. Its capacity expansion supported shipment growth and reduced unit costs during the second quarter. Future additions will need enough orders to maintain that benefit.

This contest is therefore not simply SMIC versus TSMC. It is domestic breadth versus global advanced-node leadership. China’s foundries are attempting to build a large, profitable manufacturing system under technology constraints.

TSMC can command higher prices because it offers manufacturing capabilities that few competitors can match. SMIC and Hua Hong rely more on availability, local relationships, specialty processes, and demand across many established product categories.

That model can support a substantial business. It also remains vulnerable to price competition when supply exceeds demand. The latest quarter demonstrates operating leverage, not permanent immunity from the foundry cycle.

Export Controls Make the Beat More Important

The earnings surprise carries strategic weight because it arrived while access to advanced manufacturing equipment remains restricted.

United States export controls target equipment, software, and components used to produce advanced semiconductors in China. The restrictions cover selected lithography, deposition, etching, inspection, and other manufacturing technologies.

A December 2024 control package added restrictions on 24 equipment categories and three software categories. It also expanded controls covering high-bandwidth memory and added numerous Chinese entities to the Entity List.

The stated American objective is to limit capabilities associated with advanced computing and military applications. Chinese officials and companies have criticized the restrictions as barriers to technological and economic development.

For SMIC and Hua Hong, the operational consequence is a harder path toward leading-edge production. They must secure permitted equipment, extend the usefulness of existing tools, qualify domestic alternatives, and manage licensing uncertainty.

Mature-node businesses offer a partial response. Production at established nodes generally presents fewer equipment challenges than manufacturing at the leading edge. Those processes also serve large markets that remain commercially important.

The second-quarter results suggest this response is producing financial returns. Strong utilization and improved pricing can fund research, factory construction, and process development. Cash generation also helps absorb the high depreciation associated with new fabrication plants.

However, mature-node profitability does not eliminate equipment dependence. Modern specialty processes still require reliable deposition, etching, metrology, inspection, and automation systems. Restrictions can affect maintenance, upgrades, spare parts, and future factory configurations.

The policy environment can also change. New licensing rules might affect equipment that companies previously expected to obtain. Conversely, approvals for selected tools could ease bottlenecks without changing the overall advanced-node restrictions.

SMIC’s stronger margin gives it more room to operate within that uncertainty. A company earning 25.3% gross margin can finance expansion more comfortably than one operating near break-even. Hua Hong’s move from a 10.9% margin one year earlier to 16.5% creates a similar improvement.

That progress should not be confused with technological independence. Neither company disclosed enough information to establish complete localization of manufacturing equipment or materials. Public results focus on commercial performance, not a full supplier map.

The AI narrative also needs restraint. Advanced AI processors depend on leading logic nodes, high-bandwidth memory, and advanced packaging. The results do not show that China has removed its constraints in any of those areas.

They show that AI-related spending can support surrounding markets. Flash memory, controllers, power-management chips, analog components, and industrial electronics can all benefit from data-center construction and broader digitization.

That creates an important strategic feedback loop. Domestic demand improves utilization. Better utilization raises margins. Higher margins support investment. Investment expands capacity and process capability.

The loop can strengthen China’s semiconductor base even if the leading-edge gap remains. It can also fail if demand weakens before new capacity becomes productive.

This is why the earnings beat matters beyond one quarter. It offers evidence that constrained access to advanced tools has not prevented China’s major foundries from improving the economics of their available manufacturing base.

What the Earnings Do Not Prove

One strong quarter cannot establish that higher margins, AI demand, or domestic pricing power will persist through the next capacity cycle.

The first uncertainty concerns demand quality. Customers sometimes place orders early when they expect price increases, supply shortages, or trade restrictions. That behavior can produce strong utilization before a later correction.

Consumer electronics accounted for more than two-thirds of Hua Hong’s revenue. The category grew 34.7% from one year earlier, making it the company’s largest growth contributor by value. Consumer demand can change quickly when inventory levels or replacement cycles shift.

Communications revenue at Hua Hong declined 12.8%. That weakness shows the improvement was not universal. A broader slowdown in smartphones or networking equipment could offset growth elsewhere.

SMIC’s demand also included personal computers and tablets. Inventory rebuilding helped these markets during early 2026, but channel restocking does not always indicate comparable growth in final purchases.

The second uncertainty concerns pricing. Foundries gain leverage when utilization approaches full capacity. Customers gain leverage when available capacity grows faster than orders.

Numerous Chinese manufacturers are expanding mature-node production. If those factories target overlapping customers, competitive pricing could return. SMIC and Hua Hong would then need product differentiation and manufacturing efficiency to protect margins.

Hua Hong’s utilization of 102.8% leaves little idle capacity today. Its third-quarter revenue forecast of $770 million to $780 million points to continued growth. The company expects a gross margin of 16% to 18%, which would maintain or improve the second-quarter level.

Yet new capacity brings additional depreciation. Hua Hong acknowledged that higher depreciation partially offset the benefits of better prices and cost controls during the second quarter. That pressure will remain as new lines begin production.

SMIC guided to a 26% to 28% third-quarter gross margin, a notably favorable range. Meeting that target would strengthen the argument that price increases and cost absorption are durable.

Missing the target would not erase the second-quarter beat. It would suggest that the margin surge was more sensitive to product mix, subsidies, timing, or temporary factory economics than the headline implies.

The third uncertainty is the meaning of AI demand. Hua Hong said the AI upturn had spread from memory into associated logic and analog products. Its own product figures show strong growth in embedded and standalone memory.

However, computing contributed only 1.6% of its revenue. Consumer electronics remained the dominant market. Readers should therefore view AI as one demand catalyst among several, not as the sole explanation for Hua Hong’s performance.

SMIC’s advanced-node activity is less transparent than TSMC’s reporting. TSMC discloses revenue contributions for major process generations. SMIC does not provide an equivalent public breakdown for its most advanced production.

That limits independent analysis. Investors and customers cannot precisely separate mature-node price improvements from advanced-node growth using the headline figures alone.

Geopolitical uncertainty adds another layer. Export-control changes could raise equipment costs or slow process development. Trade tensions could also encourage more local orders, producing both constraint and demand at the same time.

The results therefore support a specific claim: both companies executed better than their own forecasts during the June quarter. They do not support a claim that China has achieved semiconductor self-sufficiency or overtaken global manufacturing leaders.

Maintaining that boundary is essential. Viral financial headlines often compress revenue, profit, technology, and geopolitics into one narrative. The operating data tell a more useful story when each element remains distinct.

Three Signals to Watch Next

Third-quarter margins, capacity discipline, and customer mix will determine whether this technology news marks a durable cycle or a temporary peak.

The first signal is SMIC’s third-quarter gross margin. Management expects 26% to 28%, following 25.3% in the second quarter and 20.1% in the first.

A result inside that range would confirm that higher prices and factory utilization continue absorbing new depreciation. A result above the range would indicate even stronger pricing or product mix.

A material miss would weaken the thesis. It could reveal faster cost growth, delayed customer orders, or weaker average selling prices. Revenue alone would not answer those questions.

The second signal is Hua Hong’s ability to deliver $770 million to $780 million of third-quarter revenue while preserving a 16% to 18% margin. The revenue target implies another sequential increase after a record quarter.

Both figures need to move together. Revenue growth with a falling margin could mean the company is buying volume through lower prices. Stable or rising margins would support management’s claim that volumes and prices are improving simultaneously.

Huali Microelectronics integration will also affect this test. Greater scale can improve customer coverage and manufacturing efficiency. Integration expenses and added depreciation can weaken near-term profitability.

The third signal is the balance between shipments and final demand. Industry reports should show whether orders remain broad across microcontrollers, power management, flash, automotive electronics, and consumer devices.

If foundry shipments keep growing while electronics inventories remain controlled, the upturn will look healthier. If inventories rise faster than device sales, a correction becomes more likely.

TSMC provides a useful reference point, even though it serves a different product mix. Its advanced-node growth shows how much global semiconductor value remains concentrated in high-performance computing. SMIC and Hua Hong need not match that model to succeed.

Their immediate test is whether domestic breadth can sustain returns. A profitable mature and specialty foundry system would support thousands of products and provide funding for technical development.

For developers and enterprise buyers, the near-term effect will appear in hardware availability rather than software features. More reliable supplies of controllers, power components, embedded memory, and connectivity chips can reduce production delays.

Product teams should watch supplier qualification, lead times, and component substitutions. Procurement decisions may increasingly include Chinese foundries indirectly through local chip designers and integrated-device manufacturers.

The August results make the direction clearer, but they do not settle the outcome. SMIC and Hua Hong have shown that China’s foundry expansion can generate stronger pricing and margins under the right demand conditions.

Now the companies must repeat the performance while new capacity arrives. Watch their third-quarter margins, customer inventories, and integration costs. Those signals will reveal whether the earnings beat represents a durable manufacturing shift or another turn in a notoriously cyclical market.