Tag: TSMC

  • TSMC Faces $1 Billion Fine Over Huawei Ascend 910B Issue

    TSMC Faces $1 Billion Fine Over Huawei Ascend 910B Issue

    Key Takeaways

    1. Huawei has faced challenges with TSMC since being cut off, specifically regarding the Ascend 910B chip.
    2. TSMC may face fines up to $1 billion for allegedly violating export control regulations.
    3. TSMC previously informed the US Commerce Department about Huawei’s failed purchase attempts.
    4. The US Commerce Department has not publicly commented on the situation, and TSMC claims compliance.
    5. The outcome for TSMC remains uncertain, especially considering the US’s reliance on its chip manufacturing.

    It has been some time since Huawei was cut off from TSMC, with the Kirin 9000 being the last chip produced on TSMC’s 5 nm technology. Recently, Huawei faced difficulties when attempting to utilize TSMC nodes for its Ascend 910B chip. TSMC and Technsights discovered a chip resembling Huawei’s too closely and swiftly put a stop to it. Nonetheless, this might not be the end of the issue, as a new report suggests that the chip manufacturer could be facing significant fines.

    Potential Fines for TSMC

    Sources who wish to remain anonymous have informed Reuters that TSMC could be penalized up to $1 billion for reportedly breaching export control regulations. This situation seems questionable since TSMC had previously notified the US Commerce Department about Huawei’s unsuccessful attempt to make a purchase. However, there might be more details in this case that haven’t yet come to light. The Commerce Department has not issued any public statements regarding this matter, and TSMC claims it has maintained compliance throughout.

    TSMC’s Future in Question

    Considering the confusing timeline surrounding these events, it is quite possible that TSMC will manage to avoid penalties, particularly during these challenging times when the US relies on TSMC’s chip manufacturing capabilities to maintain an edge over China. Still, there might be some truth to the claims, especially since Chinese companies have increasingly turned to offshore or proxy firms to obtain materials that are otherwise restricted, like how DeepSeek recently acquired advanced Nvidia chips for AI development.

    Source:
    Link

    Tags: , ,
  • Trump Urges TSMC to Build in U.S. or Risk Up to 100% Tariff

    Trump Urges TSMC to Build in U.S. or Risk Up to 100% Tariff

    Key Takeaways

    1. Trump criticized financial support for foreign chip manufacturers and emphasized a tough trade stance during his presidency.
    2. He warned TSMC of a potential 100% tax if they did not build factories in the U.S., contrasting with the current administration’s $6.6 billion subsidy for TSMC in Arizona.
    3. TSMC plans a $100 billion investment in the U.S., including five new factories.
    4. Taiwan is engaging in diplomatic talks with the U.S. to address tariff threats and has proposed zero tariffs and increased investment.
    5. Taiwan’s stock market is volatile, prompting the government to activate a $15 billion stabilization fund to support investor confidence.

    U.S. President Donald Trump has ignited renewed discussion regarding trade ties between the U.S. and Taiwan, particularly focused on semiconductor manufacturing strategies. As reported by Reuters, Trump expressed strong opinions during an event for the Republican National Congressional Committee, where he critiqued the government’s financial support for foreign chip manufacturers and shared insights on his tough stance while in office.

    Trump’s Warning to TSMC

    At this gathering, Trump disclosed that he had cautioned Taiwan Semiconductor Manufacturing Company (TSMC) about a possible tax reaching as high as 100 percent if the company failed to build its factories in the U.S. In contrast to the present administration’s decision to grant TSMC’s American division a $6.6 billion subsidy for expanding its operations in Arizona, Trump noted that he provided no monetary aid, instead relying on taxation to drive his agenda. TSMC had earlier confirmed a substantial investment plan of $100 billion in the U.S., which involves constructing five new factories.

    Diplomatic Efforts by Taiwan

    In light of Trump’s tariff threats, officials from Taiwan indicated that they have begun talks with Washington to address the issue through diplomatic means. The Director-General of Taiwan’s National Security Bureau confirmed that there is “strategic communication” happening with U.S. representatives, with the goal of establishing a negotiation pathway that avoids any retaliatory actions. Taiwan has proposed zero tariffs, an increase in investment, and procurement in the U.S. as a way to ease tensions.

    Market Reactions and Stability Measures

    Meanwhile, a report from Investing highlighted that Taiwan’s stock market has been experiencing fluctuations due to these developments, leading the government to activate a $15 billion stabilization fund to bolster investor confidence. While TSMC shares remained steady, significant suppliers like Foxconn faced noticeable drops in their stock prices.

    Tags: , ,
  • Samsung Foundry May Scrap 1.4 nm Node Plans

    Samsung Foundry May Scrap 1.4 nm Node Plans

    Key Takeaways

    1. Samsung Foundry is facing issues with low yields from its SF3 node, delaying the launch of the Exynos 2500.
    2. The upcoming 1.4 nm node (SF1.4) may be abandoned, with potential delays instead of a complete scrapping.
    3. Samsung is currently focused on the Exynos 2600 and AI chips while receiving new orders for its 4 nm nodes.
    4. Samsung Foundry’s market share is only 8.2%, significantly lower than TSMC’s 67.1%, raising concerns about its future.
    5. Possible restructuring within Samsung Foundry may include integrating the Exynos division into Samsung MX for better oversight.

    Samsung Foundry’s troubles could be getting even worse. After disappointing yields from its SF3 node, which caused the Exynos 2500 launch to be delayed, the company had to scale back some of its older 5 nm and 7 nm processes due to low interest. Now, reliable insider @Jukanlosreve has suggested that Samsung’s upcoming 1.4 nm node (SF1.4) might be completely abandoned.

    Potential Changes to SF1.4

    A prior roadmap from Samsung Foundry indicated that SF1.4 was set to begin high volume manufacturing (HVM) by 2027. This node was supposed to run alongside SF2A, which is geared toward automotive applications, and SF2Z, marking the first node featuring a Backside Power Delivery Network (BSPDN). However, there is a chance that SF1.4 could be delayed further instead of being entirely scrapped.

    Current Developments

    For the moment, Samsung Foundry is still working on the Exynos 2600 utilizing SF2, along with some artificial intelligence chips for PFN. Additionally, some of its 4 nm nodes have received new orders from a few Chinese fabless companies affected by trade restrictions. However, key players continue to turn away from Samsung, opting for TSMC or even Intel instead.

    Market Share Concerns

    According to the Korea Economic Daily, Samsung Foundry might undergo significant changes due to its low market share of just 8.2%, which pales in comparison to TSMC’s dominating 67.1%. This could lead to substantial restructuring within the organization. Business Post reports that the Exynos division may be integrated into Samsung MX, which would provide greater oversight over upcoming smartphone SoC designs.

    Source:
    Link

    Tags: , ,
  • Intel Partners with TSMC for Long-Term Strategy Amid 18A Yield Issues

    Intel Partners with TSMC for Long-Term Strategy Amid 18A Yield Issues

    Key Takeaways

    1. Intel is delaying the production of its Panther Lake mobile processors due to low yields from its 18A semiconductor manufacturing process.
    2. The company has formed a long-term alliance with TSMC to outsource wafer production, moving away from its goal to eliminate outsourcing.
    3. TSMC’s manufacturing process for 3nm and 2nm technologies is more efficient than Intel’s current in-house capabilities.
    4. The delay in Panther Lake production may push shipments to 2026, missing potential sales opportunities in late 2025.
    5. Intel’s new strategy suggests a recognition of the advantages of partnering with TSMC for future production.

    It has been recently reported that Intel is possibly delaying the production of its next-generation Panther Lake mobile processors due to disappointing yields from its own 18A semiconductor manufacturing process. In the latest Intel earnings call, the company revealed a long-term alliance with TSMC to outsource the wafers needed for making its processors. This call featured a discussion between Intel’s John Pitzer and Morgan Stanley’s Joe Moore, as transcribed by Seeking Alpha.

    Challenges with In-House Production

    Intel has faced difficulties in successfully bringing semiconductor manufacturing in-house, as the 18A process developed by Intel Foundry has not achieved satisfactory yields. Conversely, TSMC boasts a more seasoned and efficient manufacturing process for 3nm and 2nm technologies. During the earnings call, Intel’s Corporate Vice President of Investor Relations mentioned that the company has been outsourcing around 30% of its wafers, largely to TSMC, for some time. However, not long ago, Intel aimed to reduce this number to zero.

    New Direction for Partnership

    Currently, it seems Intel has adopted a new approach that includes a prolonged partnership with TSMC. Pitzer reportedly commented, “We think it’s always good to have at least some of our wafers with TSMC. They’re a great supplier. It creates a good competition between them and Intel Foundry.”

    The announcement of this strategy comes soon after industry analyst Ming Chi Kuo indicated that production of Intel’s Panther Lake mobile processors has been postponed due to the underperformance of the 18A process. This delay means that the company will likely not be able to ship Panther Lake notebooks until 2026, causing them to miss potential sales in the fourth quarter of 2025. This will negatively impact the revenue, profits, and trust in the supply chain for the latter half of 2025.

    Implications for Future Production

    This strategic shift may indicate that Intel has acknowledged the benefits of a longer-term collaboration with TSMC. Still, it remains uncertain how this will influence the Panther Lake processors, which were expected to be the first mobile SoCs developed using the 18A process.

    Source:
    Link

    Tags: , ,
  • Apple announces $500 billion US investment plan amid tariff concerns

    Apple announces $500 billion US investment plan amid tariff concerns

    Key Takeaways

    1. Apple plans to invest over $500 billion in the US over the next four years, including a new server manufacturing facility in Houston.
    2. The company is doubling its US Advanced Manufacturing Fund to $10 billion and launching a skills academy in Michigan to boost manufacturing skills.
    3. Apple will collaborate with thousands of suppliers nationwide, creating direct jobs and infrastructure for its Apple Intelligence systems and data centers.
    4. The new Houston server plant will be crucial for Apple Intelligence and is expected to create thousands of local jobs by 2026.
    5. Apple aims to add around 20,000 new positions focused on research, development, and advanced technologies over the next four years.

    Apple has just revealed an unprecedented investment strategy, planning to pour over $500 billion into the US within the next four years. This initiative encompasses the establishment of a server manufacturing facility in Houston, aimed at powering its Apple Intelligence systems. Additionally, Apple is doubling its US Advanced Manufacturing Fund to $10 billion and will be initiating a skills academy in Michigan to enhance manufacturing skills.

    Collaborations and Job Creation

    In this ambitious plan, Apple intends to collaborate with thousands of suppliers across all 50 states. This effort will create direct employment opportunities, strengthen the infrastructure for Apple Intelligence, and develop data centers. Furthermore, it includes building corporate hubs and supporting Apple TV+ productions in 20 different states. Currently, Apple supports more than 2.9 million jobs in the US through direct employment, partnerships with American suppliers and manufacturers, and roles for iOS app developers.

    New Server Plant in Houston

    The new server facility in Houston, which will cover 250,000 square feet and is set to begin operations in 2026, is anticipated to generate thousands of jobs in the area. These servers, previously manufactured overseas, will now be essential for Apple Intelligence and will lay the groundwork for what’s being referred to as Private Cloud Compute, a system that integrates AI processing with high-level security for cloud services.

    Investment in Advanced Manufacturing

    As part of its enhancement to the US Advanced Manufacturing Fund, Apple is making a significant multibillion-dollar investment in advanced silicon production at TSMC’s Fab 21 in Arizona, where it stands as the largest client. Notably, mass production of Apple chips commenced just last month at this state-of-the-art facility, which already has over 2,000 workers on its payroll.

    Future Hiring Plans

    Looking forward, Apple aims to add approximately 20,000 new positions over the next four years, primarily in research and development, silicon engineering, software development, as well as AI and machine learning. The company has nearly doubled its investment in advanced R&D in the US over the past five years and is keen to maintain this growth.

    All of this is happening amid potential tariff challenges, as the upcoming Trump administration signals a crackdown on Chinese imports. Recently, Apple CEO Tim Cook had a meeting with President-elect Trump, who suggested that Apple might withdraw from Mexico and increase its investments in the US. Meanwhile, the administration is considering a potential 25 percent tariff on chips and has already introduced a 10 percent tariff on goods from China.

    Source:
    Link

    Apple announces $500 billion US investment plan amid tariff concerns

     

    Tags: , ,
  • Arm Creates In-House Chip, Meta May Be First Customer

    Arm Creates In-House Chip, Meta May Be First Customer

    Key Takeaways

    1. Arm is reportedly developing its own customizable CPU for data centers, potentially launching it this summer.
    2. The company will collaborate with TSMC for the production of these new chips as a fabless chipmaker.
    3. Softbank, which owns Arm, recently partnered with OpenAI to create extensive AI infrastructure, with Arm as a key player.
    4. Arm’s chip designs are widely used in smartphones, mobile devices, and the latest Apple Macs, known for their power efficiency.
    5. The new processors may support AI applications, increasing Arm’s relevance in the evolving tech landscape.

    Arm could be set to reveal its own processor this year, according to new rumors. The well-known chipmaker might be developing in-house chips, potentially making Meta their first client. Here’s what we’ve learned so far.

    Arm’s New Chip Venture

    A report from The Financial Times suggests that Arm is creating a CPU aimed at data centers, which will be customizable to meet the needs of various clients. As a fabless chipmaker, Arm plans to collaborate with TSMC (Taiwan Semiconductor Manufacturing Company), the largest contract chip manufacturer globally, to produce these new chips. Industry insiders believe these in-house processors might be launched as soon as this summer.

    Partnership with OpenAI

    This development comes just a month after Softbank, which owns Arm, partnered with OpenAI to create up to $500 billion in AI infrastructure. This large-scale initiative will include Arm along with Microsoft and Nvidia as key tech partners. Arm could play a significant role in this project, possibly connecting with AI-driven personal devices being developed by John Ive (a former Apple designer) and Sam Altman from OpenAI.

    Arm’s Ubiquity in Technology

    For those who might not know, Arm’s designs power almost every smartphone available today. They’re also found in most mobile devices and even run the latest Apple Macs and Qualcomm-powered Windows PCs. CPUs that utilize the ARM architecture offer impressive power efficiency without sacrificing performance, rivaling Intel and AMD chipsets. This is a key factor in their growing popularity in data centers that support AI applications.

    Tags: , ,
  • Arm Enters Chip Making with Meta as a Confirmed Customer

    Arm Enters Chip Making with Meta as a Confirmed Customer

    Key Takeaways

    1. MediaTek and Nvidia are entering the “AI PC” sector, increasing competition for Intel, AMD, and Apple.
    2. Arm is planning to launch a new chip aimed at data centers and servers, supported by the acquisition of Ampere.
    3. Arm has secured Meta as a significant client for its upcoming chip, indicating potential use in consumer devices.
    4. There could be conflicts of interest as many companies rely on Arm’s intellectual property while competing in the market.
    5. Production may involve TSMC and potentially Intel, with a likelihood of using older TSMC nodes for the initial chip version.

    With MediaTek and Nvidia set to enter the “AI PC” sector later this year, Intel, AMD, and Apple are about to face new competition. However, a fifth player is also looking to join the competition: Arm. CEO Rene Haas may reveal the new chip “as early as this summer.”

    Arm’s Ambitions

    This upcoming Arm chip is intended for use in data centers and server markets. The efforts of Arm will be supported by Ampere, a company that Softbank (Arm’s parent) is planning to acquire soon. Ampere has struggled to gain significant market share against Intel and AMD in the server sector, but things could shift once Arm steps in.

    A New Major Client

    Moreover, Arm has secured a significant client for the aforementioned chip: Meta. This chip could also be utilized in Sir Jony Ive’s secretive device. This indicates that Arm might be developing a low-power chip aimed at consumer gadgets as well. However, this scenario could lead to serious conflicts of interest, given that many companies, such as Qualcomm, Apple, and recently Intel, rely on Arm’s intellectual property.

    The report indicates that production might be outsourced to TSMC, although there’s a strong possibility that Intel will play a role as well since Arm mentioned it would utilize Intel’s 18A node for a mobile chip. If this chip actually debuts this year, it’s likely that the design has already been finalized. It’s probable that the first version will be produced on an older TSMC node like N3 or N4.

    Source:
    Link

    Tags: , ,
  • Samsung Foundry Resumes 4nm Production Lines Amid New Demand

    Samsung Foundry Resumes 4nm Production Lines Amid New Demand

    Key Takeaways

    1. Samsung Foundry halted operations at its 4 nm, 5 nm, and 7 nm fabs due to low demand, focusing on 3 nm products instead.
    2. Full production at the 4 nm line is expected to resume in June, primarily serving Samsung LSI and cryptocurrency mining ASIC manufacturers.
    3. U.S. sanctions against China may force Samsung to lose Chinese clients, impacting its business.
    4. Some Chinese companies are switching from TSMC to local manufacturers like SMIC, potentially affecting Samsung Foundry’s customer base.
    5. Revenue from trailing-edge nodes is vital for funding future R&D; without it, Samsung may struggle to compete with TSMC and Intel, leading to possible business divestitures.

    Last year, Samsung Foundry had to halt operations at its 4 nm, 5 nm, and 7 nm fabs due to a lack of demand. The company decided to concentrate on its newer 3 nm (and more advanced) products. However, it’s still facing significant challenges. Currently, there is new optimism for the struggling business.

    Production Resumption

    According to a report from Chosun (H/T @Jukanlosreve on X), the 4 nm line is set to restart full production in June. The orders are expected to come from Samsung LSI, most likely for mid-range Exynos chips and HBM4 logic dies. Additionally, cryptocurrency mining ASIC manufacturers from China are also among the customers. It’s noteworthy that one of Samsung’s initial 3 nm customers was also from that sector.

    Potential Challenges Ahead

    However, this situation may soon be at risk. The U.S. sanctions against China could compel Samsung to eliminate these clients. Some Chinese fabless companies are already shifting from TSMC to local options like SMIC, and it’s feasible that some have now turned to Samsung Foundry. Nevertheless, it’s just a matter of time before Samsung Foundry encounters such a departure.

    Future Implications

    This will add more difficulties for the chipmaker, as trailing-edge nodes are crucial for generating revenue that supports future R&D efforts. Without this financial boost, Samsung’s already struggling advanced nodes could lag even further behind TSMC and Intel, potentially leading Samsung to consider selling off parts of its business.

    Source:
    Link

    Tags: , ,
  • TSMC to Manufacture OpenAI’s AI Chips, Not Samsung Foundry

    TSMC to Manufacture OpenAI’s AI Chips, Not Samsung Foundry

    Key Takeaways

    1. AI-powered tools are attracting significant investor interest, but developing them can be costly.
    2. OpenAI is seeking to reduce operating expenses by creating its own hardware instead of relying on Nvidia.
    3. Samsung Foundry was considered for producing AI hardware, but OpenAI has chosen TSMC’s 3nm process instead.
    4. OpenAI plans to mass-produce its AI chips by 2026, with designs nearing completion.
    5. OpenAI is investing $500 million in its proprietary AI chip, aiming for long-term savings in operational costs.

    Development of AI-powered tools appears to be a profitable venture. An increasing number of investors are keen on backing these kinds of projects. Nonetheless, creating AI products can be pretty costly if one aims to compete with top players in the market. OpenAI, which is the parent organization of ChatGPT, is fully aware of these financial challenges, prompting them to seek ways to reduce operating expenses.

    OpenAI’s Strategy for Cost Reduction

    As per various reports, one of the strategies OpenAI intends to adopt for long-term savings involves crafting its own hardware to manage AI services. At present, the company relies on Nvidia, which holds a dominant position as the leading supplier of AI hardware globally. Nevertheless, Nvidia’s stronghold allows it to dictate prices that some companies, including OpenAI, find excessive.

    Samsung Foundry vs. TSMC

    Samsung Foundry has surfaced as a key potential producer for OpenAI’s AI hardware after a dialogue took place between Samsung Electronics Chairman Jay Y. Lee and OpenAI CEO Sam Altman during the week of February 3 to February 9. Some insiders hinted that the production of OpenAI’s AI chips using Samsung’s 3nm process was among the matters discussed. However, a recent report from Reuters suggests that OpenAI has opted for TSMC’s 3nm process instead.

    Future Plans and Investments

    The AI-centric firm won’t be the first significant player to part ways with Samsung after encountering issues and dissatisfaction with its wafer performance. Other companies like Qualcomm and Nvidia have also transitioned to TSMC for their needs. OpenAI is said to aim for the mass production of its AI chips by 2026. In the near future, TSMC might receive designs from OpenAI to kick off production tests, with reports indicating that the hardware design is nearing completion.

    OpenAI plans to invest approximately $500 million in the creation of its proprietary AI chip. Though this initial expense seems steep, the long-term savings in operational costs could be substantial. Apple previously took a similar path by moving away from Intel in favor of its own ARM chips for the Mac lineup.

    Source:
    Link

    Tags: , ,
  • Snapdragon 8 Elite 2 Production by TSMC Confirmed in Report

    Snapdragon 8 Elite 2 Production by TSMC Confirmed in Report

    Qualcomm has unveiled its newest top-tier processor, the Snapdragon 8 Elite, which is made using TSMC’s advanced third-generation 3nm (N3E) technology. It looks like Qualcomm is sticking with a trusted ally for its upcoming chip, the Snapdragon 8 Elite 2, as whispers indicate it will be crafted with TSMC’s even more cutting-edge 3nm N3P process.

    Confirmation of Production

    A recent report from Gamma’s Hardware info (via Jukanlosreve) has confirmed that the Snapdragon 8 Elite 2 will indeed be produced with TSMC’s 3nm (N3P) tech. This implies that Qualcomm is still not ready to switch over to Samsung Foundry for its chip manufacturing needs.

    GOOD NEWS: THE PRODUCTION OF SNAPDRAGON 8 ELITE 2 BY TSMC HAS BEEN CONFIRMED. PIC.TWITTER.COM/EN0SDOR60G
    — Jukanlosreve (@Jukanlosreve) January 21, 2025

    Specifications of the Chip

    The Snapdragon 8 Elite 2 is said to carry the model number SM8850 and bears the codename Kaanapali. The report also suggests that Samsung made attempts to create the 8 Elite 2 using its own process, which was supposedly referred to as “KaanapliS.”

    Nonetheless, the report indicates that “KaanapliS” is no longer appearing in logistics data. This hints that the anticipated Samsung-made SM8850 is unlikely to come to fruition. Consequently, the production has been redirected to TSMC, with the new codename

    Source:
    Link

    Tags: , ,