DD
MM
YYYY

PAGES

DD
MM
YYYY

spot_img

PAGES

Home Blog Page 2

Samsung, SK Hynix Evaluate Chinese Chip Equipment As Hedge Against Tougher U.S. Export Controls

0

Samsung Electronics and SK Hynix have been evaluating semiconductor manufacturing equipment from China’s Advanced Micro-Fabrication Equipment (AMEC) for potential use at their Chinese production facilities, Reuters reports, citing people familiar with the matter.

The evaluations, which began roughly two years ago, are part of contingency planning by the world’s two largest memory chipmakers as they seek to reduce operational risks should Washington impose further restrictions on Western semiconductor equipment entering China.

While neither company has approved broader deployment of AMEC’s tools, the trials represent an important milestone for China’s semiconductor equipment industry, offering one of its leading manufacturers a rare opportunity to gain validation from global memory chip leaders.

The development also marks an unintended consequence of U.S. technology restrictions: measures designed to curb China’s semiconductor ambitions are simultaneously creating openings for Chinese equipment makers to displace Western suppliers inside foreign-owned fabrication plants operating in China.

However, Samsung told Reuters it has not tested AMEC equipment for use at its factory in China and has not considered doing so.

SK Hynix declined to comment.

Export Control Uncertainty Drives Contingency Planning

According to the sources, the chipmakers began assessing Chinese equipment as uncertainty grew over whether the United States would continue allowing them to import American chipmaking tools into China.

Washington designated Samsung’s and SK Hynix’s Chinese operations as Validated End Users (VEU) in 2023, allowing them to receive certain controlled U.S. semiconductor equipment without applying for individual export licenses. That arrangement changed in 2025 when the United States revoked the VEU designation before later granting both companies annual licenses covering equipment shipments into their Chinese facilities for 2026.

Although the licenses currently permit continued operations, executives remain concerned that future restrictions could become significantly broader. Rather than targeting only new equipment sales, future U.S. measures could also limit servicing, repairs, software upgrades or replacement parts for Western tools already installed inside Chinese fabs, the sources said.

As a result, Samsung and SK Hynix are evaluating Chinese suppliers primarily as a safeguard to maintain existing production lines if access to Western equipment becomes more restricted, rather than as a means of expanding manufacturing capacity in China.

China’s Growing Equipment Capabilities

The evaluations underscore the rapid progress made by Chinese semiconductor equipment manufacturers in narrowing the technology gap with established global competitors.

While Chinese companies remain behind international leaders in advanced lithography and certain inspection technologies, they have become competitive in other areas including:

  • Etching
  • Deposition
  • Cleaning
  • Chemical mechanical planarization (CMP)

According to Dan Hutcheson, Vice Chairman of research firm TechInsights, Chinese equipment is often priced 20% to 30% below comparable products offered by Western manufacturers, making it increasingly attractive for customers seeking lower costs alongside supply-chain diversification.

AMEC’s etching systems are already deployed by leading Chinese memory producer Yangtze Memory Technologies (YMTC), providing additional confidence that some of its technology has reached commercial maturity.

Western Suppliers Face Emerging Competition

Samsung’s NAND flash memory plant in Xi’an and SK Hynix’s NAND and DRAM facilities in Dalian and Wuxi currently rely heavily on equipment supplied by U.S. companies including Applied Materials and Lam Research.

A successful qualification by either Korean manufacturer would represent a significant commercial endorsement for AMEC and could eventually increase competitive pressure on established global equipment makers, including Applied Materials, Lam Research and KLA, as well as Japanese and European rivals.

China remains one of the industry’s largest markets. Applied Materials generated $8.53 billion in revenue from China during fiscal 2025, representing approximately 30% of its global sales.

Even if Chinese equipment receives technical approval, however, replacing entrenched Western suppliers would likely take years because semiconductor manufacturing tools undergo extensive qualification procedures before entering production.

Other obstacles include smaller service networks, intellectual property concerns and potential geopolitical pressure from Washington discouraging adoption of Chinese technology.

The sources also noted that it remains unlikely Samsung or SK Hynix would install Chinese equipment at fabrication facilities in South Korea because of security and intellectual property considerations.

The broader trend nevertheless reflects China’s accelerating progress toward semiconductor self-sufficiency.

According to Deutsche Bank estimates, Chinese equipment manufacturers including Naura Technology, AMEC, Piotech and ACM Research are each expected to generate more than $1 billion in revenue during 2026.

Collectively, those companies could capture 25% to 30% of China’s projected $28 billion wafer fabrication equipment market this year. Excluding lithography and metrology equipment, Chinese manufacturers’ market share could approach 40%, highlighting the rapid expansion of domestic suppliers in segments where technological barriers are lower.

However, the evaluations demonstrate how geopolitical tensions are fundamentally reshaping procurement strategies across the semiconductor industry. Rather than relying exclusively on established Western suppliers, multinational chipmakers operating in China are now exploring domestic alternatives as insurance against future export restrictions.

U.S. Refunds $100bn in Struck-Down Trump Tariffs as Legal Battle Over Trade Powers Continues

0

Customs filing shows more than half of invalidated tariff collections have been returned to importers, although some critics believe consumers have yet to benefit

The Trump administration has refunded approximately $100 billion in tariffs that were collected before the U.S. Supreme Court invalidated a large portion of President Donald Trump’s trade duties earlier this year, according to a court filing, marking one of the largest repayments of tariff revenue in U.S. history.

The disclosure, contained in a filing submitted Tuesday to the U.S. Court of International Trade by U.S. Customs and Border Protection officials, provides the clearest picture yet of the government’s efforts to unwind tariffs that the nation’s highest court ruled had been imposed without proper legal authority.

According to the filing, “refunds (duties plus interest) of approximately $100 billion have been completed using the Consolidated Administration and Processing of Entries Refund component, certified by the agency, and sent to the U.S. Department of Treasury for disbursement.”

The figure came from refunds processed through the end of July and represents more than half of the approximately $166 billion in tariff revenue invalidated by the Supreme Court’s February ruling. The repayments include both the original duties collected and accrued interest, with the money being returned primarily to businesses that imported goods subject to the tariffs.

The refunds stem from the Supreme Court’s landmark February 20 decision striking down most of Trump’s broad tariffs imposed under the International Emergency Economic Powers Act (IEEPA).

The court concluded that the decades-old emergency powers law does not authorize a president to unilaterally impose sweeping tariffs on imports from U.S. trading partners, curbing one of the administration’s principal trade policy tools.

The ruling marked a significant constitutional and legal setback for the White House, limiting the scope of executive authority in trade policy and reinforcing Congress’ central role in setting tariffs.

The decision affected roughly $166 billion in tariffs collected under the IEEPA framework, triggering an extensive refund process administered by Customs and the Treasury Department.

Tariffs have remained a defining feature of Trump’s economic agenda throughout his presidency. The administration has argued that higher import duties protect domestic manufacturers, encourage companies to relocate production to the United States, and provide leverage in trade negotiations with foreign governments.

However, the administration’s approach has faced sustained legal challenges from businesses, trade groups and state governments, many of whom argued that the president exceeded statutory authority by invoking emergency powers to impose broad-based tariffs.

The latest court filing illustrates the substantial financial consequences of the Supreme Court’s ruling, with federal agencies now tasked with returning tens of billions of dollars already collected. While businesses that paid the tariffs are receiving the refunds, critics argue that the repayments do not compensate American households that ultimately absorbed much of the higher cost through increased prices on imported goods.

“Trump is sending the ‘refunds’ to the companies, not working people. Every single cent of these refunds should go back to American consumers,” Democratic Representative Greg Casar said this week.

Economists have long debated who ultimately bears the cost of tariffs. Although importers pay the duties at the border, many companies pass some or all of those costs through supply chains to wholesalers, retailers and ultimately consumers in the form of higher prices.

As a result, consumer advocates argue that businesses receiving refunds may already have recovered much of the tariff expense by raising prices during the period the duties were in effect.

The Supreme Court’s decision has not ended Trump’s use of tariffs.

Following the ruling, the president sharply criticized the justices, describing them as “disloyal,” and quickly introduced a temporary 10% tariff on imports under a different statutory authority that, like the IEEPA, had not previously been used by any president to impose broad tariffs.

The administration subsequently expanded its trade measures by invoking Section 301 of the Trade Act of 1974, a long-established legal mechanism that authorizes the United States to respond to unfair or discriminatory trade practices by foreign countries.

Unlike the emergency powers statute rejected by the Supreme Court, Section 301 has been used by successive administrations to impose tariffs following investigations into foreign trade practices, making it a more established legal foundation for trade enforcement.

The administration has noted that the revised tariff framework complies with existing law while preserving its broader strategy of using import duties to address trade imbalances, protect domestic industries and encourage manufacturing investment in the United States.

While the government has already returned about $100 billion, roughly $66 billion in invalidated tariff collections remains to be refunded, suggesting the unwinding of the Supreme Court’s decision will continue for months.

SpaceX Aims To Challenge AT&T, Verizon And T-Mobile With Nationwide Starlink Mobile Network

0

SpaceX plans to transform Starlink Mobile from a satellite backup service into a full-fledged mobile network capable of competing directly with the three largest U.S. wireless carriers, marking one of Elon Musk’s boldest attempts yet to disrupt another established industry.

Speaking during SpaceX’s first earnings call since its June initial public offering, President Gwynne Shotwell said the company intends to launch its next-generation Starlink Mobile satellites in 2027 and begin offering a significantly enhanced mobile service before the end of that year.

The upgraded network will be powered by an additional 65 megahertz of wireless spectrum that SpaceX agreed to acquire from EchoStar, substantially expanding the system’s capacity and performance.

Shotwell said the enhanced service would be “100 times better” than Starlink Mobile’s current offering and made clear that SpaceX intends to compete directly with the dominant U.S. telecom operators.

“I anticipate us to be able to acquire quite a few of their customers because I think our service will be better,” she said, referring to AT&T, Verizon and T-Mobile.

The statement marks SpaceX’s clearest signal yet that it is no longer positioning Starlink Mobile merely as a complementary emergency communications service but as a potential nationwide mobile operator capable of attracting mainstream wireless subscribers.

Currently, Starlink Mobile functions primarily as a satellite-to-cell safety network, allowing compatible smartphones to connect to satellites when conventional terrestrial coverage is unavailable, such as in remote regions or during natural disasters and network outages.

In the United States, the service operates through a partnership with T-Mobile under the T-Satellite brand. However, the offering is designed to complement traditional cellular coverage rather than replace a conventional 4G or 5G subscription.

A T-Mobile spokesperson pointed to comments made by Chief Executive Srini Gopalan during the company’s July earnings call, where he emphasized that satellite connectivity remains a niche service. According to Gopalan, satellite-to-cell traffic accounted for just 0.0003% of T-Mobile’s total network usage during the busiest summer travel period, underscoring how conventional ground-based infrastructure continues to handle the overwhelming majority of mobile communications.

SpaceX, however, is pursuing a much broader strategy.

Rather than relying solely on satellites, the company plans to build a hybrid communications network that combines its expanding satellite constellation with a terrestrial cellular infrastructure.

Shotwell said the company intends to deploy a network of smaller, lower-cost cellular base stations integrated with Starlink ground terminals, a model that could significantly reduce the cost and complexity of expanding wireless coverage compared with traditional telecom networks that depend on large, expensive cell towers.

Musk said the combination of satellite and ground infrastructure would ultimately provide higher bandwidth and more reliable connectivity than existing cellular providers.

“Instead of having to deploy these very expensive and difficult-to-locate, large cellular base stations, we feel reasonably confident that we can deploy a large number of small stations,” Musk said.

The strategy points to SpaceX’s broader ambition to vertically integrate communications infrastructure by combining satellite connectivity, terrestrial wireless networks and spectrum assets under a single platform.

A key milestone came in May when the U.S. Federal Communications Commission approved the transfer of EchoStar’s spectrum licenses to SpaceX, giving the company access to valuable frequencies that can support satellite, terrestrial or hybrid wireless services.

Control of additional spectrum is considered critical for expanding network capacity, improving speeds and enabling SpaceX to compete more directly with incumbent mobile operators.

Shotwell declined to disclose how much the company expects to invest in building the new mobile network, indicating that capital expenditure plans remain under wraps.

Starlink has evolved into SpaceX’s largest commercial business, serving millions of customers across consumer broadband, enterprise connectivity, aviation, maritime and government markets. The satellite internet business has become the company’s primary source of cash flow and is increasingly funding investments in artificial intelligence, data centers, next-generation Starship rockets and new communications services.

By expanding Starlink Mobile into a full-service wireless network, SpaceX is seeking to challenge the decades-long dominance of AT&T, Verizon and T-Mobile with a hybrid satellite-terrestrial model that could extend coverage into rural and underserved areas while reducing infrastructure costs. The initiative would mark one of the most significant competitive threats to the U.S. telecommunications industry in years and further position SpaceX as an integrated communications and technology company rather than solely a space launch provider.

India’s FX Reserves Near $693bn As RBI Inflows Bolster Buffer, But Economists See Limited Upside for Rupee

0

India’s foreign exchange reserves climbed to a near three-month high at the end of July, strengthening the Reserve Bank of India’s (RBI) ability to shield the rupee from external shocks, but economists expect the currency to remain largely range-bound as the central bank continues to manage sizable dollar obligations accumulated from previous market interventions.

Data released by the RBI on Wednesday showed foreign exchange reserves rose by nearly $10.5 billion to $692.9 billion in the week ended July 31, marking the largest weekly increase since late January.

The jump reflects the success of the central bank’s foreign-currency deposit initiative launched in June, which was designed to shore up India’s external buffers as oil price volatility linked to the U.S.-Israeli conflict with Iran heightened risks to the country’s balance of payments and currency stability.

The RBI has so far received $36.7 billion through Foreign Currency Non-Resident (FCNR) deposits mobilized by banks under the special scheme through July 31. The facility allows banks to swap those foreign-currency deposits with the RBI under a zero-cost hedging arrangement that remains open until the end of September.

The stronger reserve position provides the central bank with greater firepower to smooth excessive volatility in the rupee, which has faced pressure this year from elevated crude oil prices, capital flow swings and geopolitical uncertainty affecting global financial markets.

“India’s foreign exchange reserves continue to be adequate in terms of the standard metrics of reserve adequacy with import cover of over 10 months and external debt cover of 90.8%,” RBI Governor Sanjay Malhotra said while presenting the central bank’s monetary policy decision in Mumbai.

The RBI left its benchmark repo rate unchanged at 5.25% on Wednesday, a widely anticipated decision that underscored the central bank’s preference to preserve policy flexibility while monitoring inflation and global risks.

The reserve build-up coincided with a sharp appreciation in the rupee. During the week covered by the data, the currency gained 1.2% against the U.S. dollar to close at 95.38, its strongest weekly advance in four months.

However, analysts caution that stronger reserves do not necessarily translate into sustained currency appreciation.

A Reuters poll of 36 foreign exchange strategists found broad consensus that the rupee will remain largely stable over the coming months before weakening modestly over the next year.

The median forecast projects the currency at 95.25 per dollar in three months and at the end of January 2027, before easing to 95.95 in twelve months.

The relatively subdued outlook reflects expectations that much of the incoming foreign capital will be used to offset the RBI’s sizeable forward dollar commitments rather than support a stronger exchange rate.

“Inflows will be enough to fund the RBI’s requirements rather than be used for currency appreciation. I expect more sideways movement for the rupee rather than any upside,” said Anitha Rangan, chief economist at RBL Bank.

Dhaval Shah, founder and managing director of De-Risk Forex Consultancy, said the FCNR mobilization has materially strengthened India’s external position and could push reserves above another major milestone.

“As flows from the FCNR scheme have gained pace, we expect the headline FX reserve figure to cross $700 billion in coming weeks, and it will also help the RBI to reduce its short FX book,” Shah said.

“The bigger picture will continue to favor rupee appreciation.”

Economists estimate the RBI’s June measures could ultimately attract around $50 billion in foreign currency by year-end, further boosting India’s external buffers.

Nevertheless, analysts say the central bank’s sizeable forward dollar book, estimated at more than $100 billion as of June, will likely absorb much of those inflows, limiting their impact on the exchange rate.

The forward positions stem from previous interventions aimed at smoothing volatility in the foreign exchange market. As those contracts mature, the RBI will need dollar inflows to meet its obligations, reducing the scope for reserves to translate directly into a stronger rupee.

Unlike several emerging-market central banks that have raised interest rates to defend their currencies against imported inflation, the RBI has so far avoided using monetary policy as an exchange-rate tool, instead relying primarily on its substantial reserve stockpile and targeted market intervention to maintain orderly currency movements.

However, the latest reserve increase bolsters India’s position among countries with the world’s largest foreign exchange buffers, providing policymakers with greater flexibility to manage external shocks even as higher oil prices, geopolitical tensions and global monetary uncertainty continue to cloud the outlook for emerging-market currencies.

AMD Shares Slide After Earnings Beat as Lofty AI Expectations Overshadow Strong Data Center Growth

0

Chipmaker posts 50% revenue growth and doubles data center sales, but investors demand stronger upside after stock’s 132% rally this year.

Advanced Micro Devices’ shares fell sharply in premarket trading on Wednesday, even after the chipmaker delivered better-than-expected second-quarter results, as investors looked beyond another quarter of robust artificial intelligence-driven growth and questioned whether the company’s performance justified its elevated valuation.

AMD shares were down about 8.7% before the opening bell, illustrating how rapidly rising expectations have become one of the biggest challenges for companies at the center of the AI investment boom.

The company reported second-quarter revenue of $11.54 billion, a 50% increase from a year earlier and above analysts’ average estimate of $11.28 billion, according to market data. The results reinforced AMD’s emergence as one of the primary beneficiaries of surging global demand for AI computing infrastructure.

Much of the growth came from its data center business, where revenue more than doubled to $6.7 billion, up 107% year-on-year. AMD attributed the performance to strong demand for both its central processing units (CPUs) and graphics processing units (GPUs), which are increasingly being deployed in AI training and inference workloads by cloud providers and enterprise customers.

The data center division has become AMD’s most important growth engine as hyperscale cloud companies continue expanding AI infrastructure and corporations accelerate investment in generative AI applications.

Chief Executive Lisa Su said AI demand continues to reshape the semiconductor industry and expand computing requirements across virtually every market the company serves.

“AI is driving a significant expansion in demand for compute across all of our markets,” Su said in a statement on Tuesday.

“Our leadership portfolio and growing customer visibility position us exceptionally well to capture this expanding opportunity and deliver substantial revenue and earnings growth in the years ahead.”

Even so, investors appeared unconvinced that the results were strong enough to justify the stock’s rapid ascent. AMD shares had surged about 132% this year before Wednesday’s decline, fueled by optimism that the company is narrowing the technology gap with Nvidia in AI accelerators while also strengthening its position in server processors.

That rally has significantly raised the performance threshold investors expect from each earnings report.

Analysts at Deutsche Bank said AMD’s quarterly results exceeded Wall Street’s consensus forecasts but failed to surpass the more optimistic expectations that had built up among investors ahead of the release.

“Second-quarter earnings came in slightly ahead of consensus,” the bank said in a research note, adding that they fell short of the market’s most bullish projections.

The reaction follows the increasing need for AI companies to deliver results that substantially exceed expectations rather than merely beat consensus estimates. As AI-related stocks have rallied to record valuations over the past year, investors have become less tolerant of earnings that fail to provide meaningful upside surprises or stronger-than-expected guidance.

AMD’s latest results come as competition in the AI semiconductor market intensifies.

Although Nvidia remains the dominant supplier of AI accelerators used to train and deploy large language models, AMD has steadily expanded its presence by introducing increasingly powerful GPU products and capitalizing on customers seeking greater supplier diversification. The company has also benefited from strong demand for its EPYC server processors, which continue to gain market share from Intel in enterprise data centers.

Last month, AMD significantly raised its long-term outlook for the semiconductor industry, reflecting management’s growing confidence that AI will transform global computing demand.

The company now expects the semiconductor market to reach approximately $2 trillion annually by 2028, up sharply from previous projections. Of that total, AMD estimates AI accelerators, primarily GPUs, will account for roughly $1.4 trillion, nearly tripling its earlier forecast of $500 billion for the same period.

AI investment is increasingly reshaping the semiconductor industry’s growth trajectory, as governments, cloud providers and technology companies commit hundreds of billions of dollars to AI infrastructure.

Still, investors remain focused on valuation.

Morningstar Chief Equity Strategist Michael Field said AMD’s decline was less about the company’s operational performance than the exceptionally high expectations embedded in its share price.

“It’s simply a case of market expectations being too high,” Field told CNBC.

“The stock has trebled in the last 12 months and now trades on a P/E multiple of 170, meaning expectations are commensurately high,” he said.

“We view the dip as a potential buying opportunity, and investors may take a similar view in the coming weeks.”

The market reaction reveals that AI leaders are facing intense pressure to deliver more. As enthusiasm for artificial intelligence continues to fuel record investment across the semiconductor industry, companies such as AMD are increasingly judged not only on their financial performance but also on their ability to consistently exceed the market’s already elevated expectations.

Even strong revenue growth, expanding margins and surging AI demand may no longer be sufficient if investors have priced in even more optimistic outcomes.