TSMC

AI chipmaker Cerebras Systems saw its stock drop 16-20% after reporting Q2 2026 revenue of $180.11 million, missing analyst estimates of $194.23 million. Despite a 74% year-over-year revenue increase and raised full-year guidance to $880-890 million, investors punished the miss amid high AI stock valuations. The company swung to a net loss of $450.53 million due to IPO-related stock compensation, while transitioning from hardware sales to cloud services, with cloud revenue surging 281%. Cerebras reiterated its $20 billion deal with OpenAI and plans to unveil its next-generation chip on August 18.

7 reports · 3 sources · Updated Aug 17, 07:27 PM UTC

Intel announced a $15 billion share sale on August 10, 2026, later upsized to $20 billion, to fund expansion of its chip contract manufacturing business and capitalize on AI-driven demand. The offering, priced at $95 per share, was managed by major banks. Despite a stock surge this year, shares fell over 3% on dilution concerns. The move underscores Intel’s strategy to challenge TSMC and boost capital expenditure to $20 billion.

17 reports · 5 sources · Updated Aug 13, 04:33 PM UTC

Sony and TSMC are planning a ¥1 trillion ($6.4 billion) joint venture to build an image sensor factory in Kumamoto, Japan, with production targeted for 2029. Sony will hold a 60% controlling stake, while TSMC owns 40%. The facility aims to meet growing demand from AI-powered robots and self-driving cars. Japan’s government is considering financial support. The partnership strengthens semiconductor supply chain diversification and reduces reliance on single manufacturing hubs.

8 reports · 3 sources · Updated Aug 11, 10:41 AM UTC

Intel announced a $15 billion stock offering on August 10, 2026, to fund AI chip growth, capital expenditures, and working capital. The offering was later upsized to $20 billion after institutional orders exceeded $100 billion, with major backers including the U.S. government, Nvidia, and SoftBank. Intel's stock initially fell 5% on dilution concerns but remains up 175% year-to-date. The company secured Tesla as a foundry customer and raised its 2026 capital spending target to $20 billion.

17 reports · 2 sources · Updated Aug 18, 07:33 PM UTC

On August 6, 2026, AMD announced the acquisition of Toronto-based startup Taalas, which specializes in AI inference chips hardwired for single models, offering faster speeds and lower power consumption than general-purpose GPUs. The deal, terms undisclosed, aims to strengthen AMD’s AI portfolio against Nvidia, which recently acquired Groq for $20 billion. Taalas’ technology will be integrated into AMD’s Instinct GPUs and Helios rack-scale systems, targeting low-latency applications.

8 reports · 2 sources · Updated Aug 9, 10:26 AM UTC

China has revised regulations for integrated-circuit layout designs, tightening originality requirements and allowing punitive damages for serious infringement. Signed by Premier Li Qiang on July 23 and effective October 15, the rules aim to safeguard domestic semiconductor know-how amid escalating US restrictions on advanced chip design software and manufacturing equipment. The revisions clarify licensing, transfer, and collateral use of layout-design rights, and mandate rewards for design personnel, reflecting Beijing’s strategic push for technological self-sufficiency.

7 reports · 3 sources · Updated Aug 4, 10:00 AM UTC

Samsung Electronics secured a $200 billion partnership with Broadcom to produce custom AI accelerators and advanced chips using 2-nanometer and below process technologies. The five-year agreement, announced at the AI Summit in San Francisco on July 26, 2026, covers memory (HBM), foundry, and advanced packaging. This deal boosts Samsung’s foundry business, helping it compete with TSMC, and underscores the growing trend of custom AI chip development by major tech companies.

6 reports · 3 sources · Updated Jul 30, 10:42 AM UTC

On July 28, 2026, a magnitude 7.1 earthquake struck off the coast of Kyushu, Japan, near Kumamoto Prefecture, triggering a brief tsunami warning. The quake caused significant damage, including a shopping mall collapse in Kashima that killed at least 18 people and injured 62. Power outages affected nearly 35,000 homes, and rail services were suspended. A government task force was established, and rescue operations involving Self-Defence Forces and international teams continued amid heat. No nuclear irregularities were reported.

13 reports · 5 sources · Updated Aug 7, 06:06 PM UTC

Zeiss Semiconductor Manufacturing Technology is adding approximately 25,000 square meters of production space at its Oberkochen site in southern Germany, with the first building now operational four years after the 2022 groundbreaking. The site, along with Wetzlar, is the only place globally where optical columns for ASML's EUV lithography scanners can be built. ASML's annual report states that its system output is limited by Zeiss SMT's production capacity. ASML plans to increase Low-NA EUV output by roughly 30% in 2027 and is compressing its scanner build cycle. Zeiss SMT's purchases from ASML reached €4.41 billion in 2025, up from €3.95 billion in 2024. The article also details the complexity of High-NA optics, which contain over 40,000 parts and weigh 12 tons, and notes ASML's significant financial support to Zeiss through loans and advance payments.

1 reports · 1 sources · Updated Jul 26, 11:46 AM UTC

Zeiss Semiconductor Manufacturing Technology is expanding its Oberkochen, Germany facility by adding approximately 25,000 square meters of production space, with the first building now operational four years after the 2022 groundbreaking. This expansion is critical because Oberkochen and Wetzlar are the only locations worldwide capable of producing the optical columns for ASML's EUV lithography scanners, and ASML's annual report states that its system output is limited by Zeiss SMT's production capacity. ASML plans to increase Low-NA EUV output by roughly 30% in 2027, and its CFO noted the company is compressing build cycles from 22 to 15-16 weeks. The article also details the complexity of High-NA optics (over 40,000 parts, 12 tons), significant financial ties between ASML and Zeiss (€4.41 billion in purchases in 2025, €1.91 billion in loans receivable), and Zeiss SMT's broader expansion including a multifunction factory in Wetzlar and an innovation center in Korea.

1 reports · 1 sources · Updated Jul 26, 11:46 AM UTC

TSMC announced an additional $100 billion investment to build at least four advanced 2nm and below chip fabs plus packaging facilities in Arizona, bringing total US commitment to $265 billion. The announcement coincided with record Q2 profit of $22.35 billion (up 77% YoY) and raised 2026 capex forecast to $60-64 billion, driven by surging AI demand. The expansion aims to create a complete domestic supply chain for US customers, though execution faces labor, water, and visa constraints.

16 reports · 4 sources · Updated Jul 24, 04:08 PM UTC

On July 23, 2026, Amkor Technology announced a multi-year agreement with Nvidia valued at $1.5 billion to expand advanced semiconductor packaging and testing capacity in the United States, particularly in Arizona. Under the deal, Nvidia will make a prepayment to support Amkor's U.S. operations. The companies will jointly develop packaging and testing technologies for Nvidia's AI and accelerated-computing platforms, focusing on combining different chip types in a single package. Amkor already supplies advanced packaging for Nvidia's data center processors. The expanded partnership aims to meet growing AI infrastructure demand. Additionally, Amkor recently entered a 10-year partnership with TSMC and is also working with Advanced Micro Devices on chip packaging. Following the announcement, Amkor's shares rose 17% in extended trading.

1 reports · 1 sources · Updated Jul 23, 09:44 PM UTC

At its Advancing AI 2026 event, AMD revealed the Instinct MI455X GPU, a 320-billion-transistor chip built on TSMC's 2N GAA process with CDNA 5 architecture. The accelerator features four Accelerator Complex Dies (XCDs) stacked on Fabric and Cache Dies using hybrid bonding, connected to 12 HBM4 memory stacks via CoWoS-L packaging. AMD also introduced the Helios rack-scale architecture, linking 72 MI455X GPUs into a coherent system to rival Nvidia's NVL72 design. The MI455X offers theoretical peak performance of 40.26 PF in OCP MXFP4 format, surpassing Nvidia's Rubin in some metrics. Key architectural changes include a shift to 32-wide wavefronts and enhanced transcendental math units. AMD announced customer wins with Microsoft and Anthropic, though cautioned that real-world performance may be lower than theoretical peaks.

1 reports · 1 sources · Updated Jul 23, 06:05 PM UTC

At its Advancing AI 2026 event, AMD revealed the Instinct MI455X GPU, its most advanced AI accelerator built with 320 billion transistors using TSMC's 2N GAA process and CoWoS-L packaging. The chip features CDNA 5 architecture with 256 Work Group Processors, a wavefront size of 32, and up to 4x performance improvement in lower-precision formats like MXFP4 and MXFP8 compared to the previous MI355X. AMD also introduced the Helios rack-scale architecture, linking 72 MI455X GPUs into a coherent system to rival Nvidia's NVL72 design. On paper, the MI455X outperforms Nvidia's upcoming Rubin GPU in several peak FLOPS metrics. AMD announced key customer wins with Microsoft and Anthropic. However, the company acknowledged that real-world performance may be lower than theoretical peaks, emphasizing the need for software optimization.

1 reports · 1 sources · Updated Jul 23, 06:05 PM UTC

At its Advancing AI 2026 event, AMD revealed the Instinct MI455X GPU, a 320-billion-transistor chip built on TSMC 2N GAA and N3P processes using advanced packaging. The MI455X features CDNA 5 architecture with 256 Work Group Processors, a wavefront size of 32, and up to 4x performance gains in lower-precision formats like MXFP4 and MXFP8 compared to the previous MI355X. AMD also introduced the Helios rack-scale architecture, linking 72 GPUs into a coherent system to rival Nvidia's NVL72. Theoretical peak performance exceeds Nvidia's Rubin GPU in several metrics, though AMD acknowledged real-world performance may be lower. Key customer wins include deals with Microsoft and Anthropic.

1 reports · 1 sources · Updated Jul 23, 06:05 PM UTC

TYLsemi, a new semiconductor startup, has publicly launched with $43 million in early-stage funding. The company aims to simplify custom processor development for AI infrastructure by offering pre-validated chiplets and custom ASIC design services. Co-founded by industry veterans Mohit Gupta and Sunil Bhardwaj, TYLsemi leverages mature advanced packaging, standardized die-to-die interfaces like UCIe, and supply-chain resilience trends. It targets the growing AI accelerator market, which is projected to reach $604 billion by 2033. TYLsemi claims its chiplet approach can reduce total cost of ownership by 57% compared to monolithic designs, making custom multi-chiplet processors more accessible to a wider range of customers.

1 reports · 1 sources · Updated Jul 23, 05:34 PM UTC

TYLsemi, a new semiconductor startup co-founded by industry veterans Mohit Gupta and Sunil Bhardwaj, has publicly launched with $43 million in early-stage funding. The company aims to simplify and reduce the cost of developing custom processors for AI infrastructure by offering pre-validated, standards-based chiplets (using UCIe connectivity) and full custom ASIC design services. TYLsemi targets customers who want to build multi-chiplet processors without the high costs and yield challenges of monolithic reticle-sized dies on leading-edge nodes. The company estimates its chiplet approach can reduce total cost of ownership by 57% at scale, from $350 million for a monolithic 700 mm² 3nm chip to $150 million for a chiplet-based design. The founders cite maturing advanced packaging, die-to-die standards, and supply-chain resilience needs as key market drivers.

1 reports · 1 sources · Updated Jul 23, 05:34 PM UTC

TYLsemi, a new semiconductor startup co-founded by industry veterans Mohit Gupta and Sunil Bhardwaj, has publicly launched with $43 million in early-stage funding. The company aims to simplify and reduce the cost of developing custom multi-chiplet processors, particularly for AI infrastructure. TYLsemi will offer pre-validated, standards-based chiplets (connectivity, power, memory) and end-to-end custom ASIC design services, enabling companies without silicon expertise to build their own multi-chiplet processors. The founders cite maturing advanced packaging, the arrival of the UCIe die-to-die standard, and supply-chain resilience needs as key market drivers. TYLsemi estimates its chiplet approach can reduce total cost of ownership by 57% compared to monolithic designs, from $350 million to $150 million at 100,000-unit volumes.

1 reports · 1 sources · Updated Jul 23, 05:34 PM UTC

Taiwan Semiconductor Manufacturing Co (TSMC) posted a record second-quarter revenue of T$1.27 trillion ($39.62 billion), a 36% year-over-year increase, driven by robust demand for AI chips from clients like Nvidia and Apple. June revenue surged 67.9% to T$442.68 billion, marking the company’s best month ever. The results slightly exceeded analyst estimates and topped TSMC’s own guidance. The preliminary data release was delayed due to Typhoon Bavi. TSMC’s shares have risen 57% in 2026, reflecting strong AI infrastructure spending.

20 reports · 3 sources · Updated Jul 22, 04:49 PM UTC

Intel and Fortinet announced a strategic collaboration to co-develop and manufacture Fortinet’s next-generation Security Processor 6 (SP6) on Intel’s 4 EUV process node. This marks the first publicly named external foundry customer for Intel 4, a node originally intended for internal use. The chiplet-based SP6 will power FortiGate firewalls, moving Fortinet away from TSMC-linked supply chains. The deal, announced under CEO Lip-Bu Tan, provides significant volume potential for Intel’s foundry business, which reported $307 million in external revenue in 2025.

7 reports · 3 sources · Updated Jul 22, 10:12 PM UTC

Taiwan Semiconductor Manufacturing Company (TSMC) announced a $100 billion increase in its US investment, bringing total planned spending in Arizona to $265 billion. The expansion aims to diversify manufacturing away from Taiwan amid geopolitical risks and meet surging AI chip demand. TSMC's Arizona facilities are already producing 4-nanometer chips with yields comparable to Taiwan. The company reported 34% year-over-year revenue growth to $40 billion last quarter, with record free cash flow of $40 billion. Management plans to spend $60 billion or more on new facilities in 2026. Despite strong performance, TSMC stock has fallen 15% from recent highs and trades at a P/E of 29. The investment is expected to benefit customers like Nvidia and equipment suppliers like ASML, though oversupply risks remain a concern.

1 reports · 1 sources · Updated Jul 22, 04:05 PM UTC

This article from Yahoo Finance predicts that Taiwan Semiconductor Manufacturing (TSMC) stock will outperform the S&P 500 over the next five years. It highlights TSMC's strong Q2 2026 results, with revenue of $40.20 billion (up 36% YoY) and net income growth of 77.4%. Despite a recent 8% pullback, shares are up 40.43% year-to-date. The analysis notes TSMC trades at a forward P/E of 25 and a PEG near 1, considered historically cheap for a company targeting a 25% revenue CAGR through 2029. The article outlines a path to $1,000 per share by 2031, requiring 18.6% annualized gains, contingent on successful 2nm and A13 node ramps and sustained AI hyperscaler capex. Key risks include geopolitical tensions across the Taiwan Strait. The piece also promotes a separate list of top AI stocks from an analyst who called NVIDIA in 2010.

1 reports · 1 sources · Updated Jul 22, 03:30 PM UTC

This analysis argues that Nvidia's competitive advantage lies not in its hardware but in its free CUDA software ecosystem, which locks developers into using Nvidia chips at scale. The article highlights Nvidia's Q1 FY2027 revenue of $81.6 billion (up 85% YoY), 71% gross margins, and $48.6 billion in quarterly free cash flow. It notes that China contributes zero to Nvidia's $91 billion Q2 guidance, while Blackwell and Rubin chips provide $1 trillion in revenue visibility through 2027. The author contrasts Nvidia with AMD (which lacks CUDA) and Broadcom (whose ASICs lack developer lock-in), concluding that the software moat justifies continued investment at a forward P/E of 23.

1 reports · 1 sources · Updated Jul 22, 02:03 PM UTC

A Motley Fool analysis published on Yahoo Finance reports that a major buying signal has emerged for Intel (NASDAQ: INTC) stock ahead of its Q2 2026 earnings release on July 23. Intel shares have surged 357% over the past year but recently declined 25% from their June 30 high. The bullish catalyst stems from TSMC's Q2 earnings call, where management noted that 'agentic AI' is driving a resurgence in CPU demand for AI data centers. Market research firm TrendForce estimates a 50-50 CPU-to-GPU split in agentic AI data centers, compared to a 1:4 to 1:8 ratio in traditional LLM-centric centers. Intel, holding an estimated 54% share of the server CPU market, stands to benefit significantly. Analysts expect Intel to swing to earnings per share of $0.22 in Q2 from a loss of $0.10 a year ago, with potential upside from growing server CPU revenue, which AMD projects could reach $120 billion annually by 2030.

1 reports · 1 sources · Updated Jul 22, 01:36 PM UTC

Taiwan Semiconductor Manufacturing (TSMC) reported strong Q2 2026 earnings with 34% revenue growth and expanding margins, yet its stock fell 5% amid market concerns over excessive AI spending. CEO C.C. Wei expressed high conviction in the AI megatrend, and management raised capital expenditure guidance to $60-64 billion, with an additional $100 billion planned for Arizona facilities. The article warns Nvidia investors that despite likely strong upcoming earnings, the market may not reward the report due to growing skepticism about the sustainability of AI investment cycles. Drawing parallels to pandemic-era boom-bust patterns, the author suggests TSMC and Nvidia are more resilient like Amazon and Shopify rather than Zoom or Fiverr, but near-term volatility is expected.

1 reports · 1 sources · Updated Jul 22, 12:50 PM UTC

Adata chairman Simon Chen stated that the global DRAM shortage will persist for another decade, dismissing talk of an AI bubble as premature. He argued that electricity (especially green power) and memory will be the world's two scarcest resources over the next decade. Chen rejected concerns that cloud providers renting spare compute capacity indicate overinvestment, instead predicting AI applications will spread across business, government, and consumer markets. He cited edge devices including robots, driverless vehicles, and satellites as future demand drivers. Adata forecasts DRAM contract prices will rise 20-30% in Q3, with NAND flash up 35-40%. Chen noted that major memory suppliers learned from past downturns and will expand cautiously, while Chinese makers face equipment restrictions. Adata has stockpiled over NT$30 billion in chip inventory, benefiting from rising prices.

1 reports · 1 sources · Updated Jul 22, 11:00 AM UTC

Singapore stocks ended higher on Wednesday, July 22, 2026, with the Straits Times Index (STI) gaining 1.2% or 68.70 points to close at 5,595.42, driven by a rebound in global chip stocks. Yangzijiang Shipbuilding was the top blue-chip gainer, rising 5.6%, while Singapore Airlines was the worst performer, declining 1.8%. The three local banks all ended higher. The Singapore Exchange saw the debut of its first three US Singapore Depository Receipts (SDRs) – Sea, Grab, and SpaceX – which turned in a muted performance, with all three closing slightly below their opening prices. Market strategist James Ooi noted the debut as a positive milestone that could broaden market participation and attract growth-oriented investors. Across the broader market, gainers outnumbered losers 290 to 279, with 1.2 billion securities worth S$2.4 billion changing hands. Regional indices were mixed, with Hong Kong's Hang Seng Index falling 1% and South Korea's Kospi rising 0.7%.

1 reports · 1 sources · Updated Jul 22, 10:17 AM UTC

Pressure from President Donald Trump to manufacture advanced semiconductors in the U.S. is increasing costs and squeezing margins at TSMC, the world's leading chipmaker. Since Trump returned to power in 2025, TSMC has announced $200 billion in U.S. manufacturing commitments, including a recent $100 billion investment in Arizona facilities. While TSMC reported a 77.4% jump in second-quarter profit driven by AI demand, CFO Wendell Huang said overseas expansion is diluting gross margins by 2-4% over the next several years. Analysts estimate TSMC's U.S. chips cost 20-50% more than those made in Taiwan. The company is reportedly raising prices by up to 10% in 2027, with customers expected to absorb most of the higher costs due to TSMC's market dominance. The White House credits Trump's trade and economic policies for driving the investments.

1 reports · 1 sources · Updated Jul 22, 05:00 AM UTC

Taiwan-based Wistron Corporation shares jumped 9.7% on Wednesday after the company announced the opening of its first U.S. facility in Fort Worth, Texas. The $700 million, 324,000-square-foot D1 AI smart facility currently produces Nvidia's GB300 Grace Blackwell Ultra Superchip and will later expand to produce Nvidia's Vera Rubin Superchip, critical for next-generation AI computing. Wistron Chairman Simon Lin stated the location will be one of the most important for building AI infrastructure in the United States. The move follows TSMC's recent announcement of expanded investment in Arizona, highlighting a broader trend of Taiwanese tech giants expanding U.S. operations. Nvidia CEO Jensen Huang emphasized that demand for AI factories is incredible and must be produced everywhere. Nvidia's partner manufacturing facilities span 43 U.S. states, contributing an estimated $485 billion to U.S. GDP in 2026.

1 reports · 1 sources · Updated Jul 22, 04:10 AM UTC

Micron Technology stock surged 13.4% on July 21, 2026, driven by two key factors. First, Taiwan Semiconductor Manufacturing Company (TSMC) announced price increases of up to 10% for contract chip manufacturing in 2027, citing rising costs and strong AI chip demand. This signals that memory chip maker Micron can also raise prices, as AI inference requires significant memory. Second, Bank of America analyst Vivek Arya dismissed concerns that cheap Chinese AI models threaten Micron's business, arguing that lower-quality GPUs may actually require more memory chips. The article presents a bullish thesis for Micron based on pricing power and AI-driven memory demand, though it notes that Motley Fool's Stock Advisor did not include Micron in its top 10 recommended stocks.

1 reports · 1 sources · Updated Jul 21, 05:40 PM UTC