New AI chip from MediaTek to be produced with NVIDIA

MediaTek is reportedly collaborating with NVIDIA to develop a new AI processor, which is expected to be manufactured using 3nm technology and is slated for tape-out later this month. This move comes alongside MediaTek’s recent launch of its Dimensity 9400 smartphone chipset, indicating the company’s ambition to diversify its product lineup. The new AI processor is projected to enter mass production by late 2025 and aims to fill a gap in the market for Windows-on-Arm devices, especially given the comparatively weaker graphics performance of Qualcomm’s Snapdragon X Elite.

The new chip is expected to work alongside NVIDIA’s GPU, aiming at major laptop brands like Dell, HP, Lenovo, and Asus as potential clients. Priced at approximately $300, this processor signifies MediaTek’s move into the AI PC sector, marking a significant shift from its usual collaborations with AMD. However, MediaTek has previously partnered with NVIDIA, notably in their recent unveiling of automotive SoCs that incorporate NVIDIA technology, highlighting their mutual interest in enhancing AI capabilities.

As the AI landscape evolves, MediaTek’s partnership with NVIDIA could bolster its position in a competitive market, especially given NVIDIA’s dominance in AI processors. This collaboration underscores MediaTek’s strategic shift and opens avenues for innovative products in the AI-driven technology sector.

OpenAI eyes ambitious plan for new wafer fabs

OpenAI, known for its AI models, appears to be exploring the semiconductor manufacturing sector, raising questions about the feasibility of building wafer fabrication plants. Reports recently surfaced about CEO Sam Altman’s discussions with executives from major chip manufacturers, including TSMC and Samsung, during his trip to Asia last year. Altman proposed an ambitious $7 trillion plan to construct 36 new wafer fabs and data centres, aiming to produce AI chips funded by the United Arab Emirates. He believes these facilities would support the burgeoning demand for AI capabilities.

The investment Altman suggested is staggering, amounting to a quarter of the annual output of the US economy. However, the timeline for establishing these fabs is lengthy, as it would take several years to meet OpenAI’s growing computing power requirements. TSMC, while approached for the project, found the proposal too aggressive and risky, noting that even a few additional wafer fabs would entail significant capital and risk.

Building a wafer fab is an enormous undertaking, often costing hundreds of billions of dollars due to various factors. The expenses stem from land acquisition, facility construction, equipment procurement, and ongoing operational costs. Advanced lithography machines and other essential equipment represent substantial financial commitments, while research, maintenance, and talent training add to the complexity. Current estimates place the cost of modern fabs in the billions; for instance, Intel’s factories in Arizona are expected to cost around $15 billion each, and Samsung’s Texas fab is projected at $25 billion.

Moreover, the cost of constructing a wafer fab varies by region. In Asia, established supply chains, available talent, and supportive policies contribute to lower costs, whereas building in Europe, the US, and the Middle East can be more expensive due to the necessity of importing technology and developing a comprehensive supply chain. Overall, OpenAI’s ambitions in the semiconductor space highlight the significant challenges and investments required to succeed in this critical industry.

Foxconn eyes massive Nvidia superchip plant in Mexico amid AI boom

Foxconn is constructing the world’s largest facility in Mexico for bundling Nvidia’s GB200 superchips, a crucial component of Nvidia’s upcoming Blackwell computing platform. The Taiwanese electronics giant, known for assembling Apple’s iPhones, has been capitalising on the surge in demand for AI technology by manufacturing servers designed for AI processing.

Nvidia has already started shipping samples of its Blackwell chips and expects substantial revenue from these in the coming months. Foxconn executives highlighted the enormous demand for Nvidia‘s Blackwell platform, with many industry players seeking to adopt this next-generation AI computing technology.

The manufacturing plant in Mexico will have substantial capacity, building on Foxconn’s existing investments in the country. The company’s production capabilities are equipped with advanced liquid cooling technologies necessary for supporting AI servers, ensuring they meet the growing needs of the AI sector.

In addition to its focus on AI, Foxconn is diversifying its operations by expanding into the electric vehicle market. The company plans to leverage its technological expertise to become a leading EV contract manufacturer, producing vehicles under its Foxtron brand while helping automakers with manufacturing.

Samsung apologises amid earnings miss and AI struggles

South Korean tech giant, Samsung, has issued an apology to its customers and investors after missing market expectations for its third-quarter earnings. The electronics giant reported a consolidated operating profit of KRW9.1 trillion ($6.64bn), falling short of the KRW10.33 trillion ($7.65bn) predicted by analysts. Despite this, the company’s sales saw a 6% rise compared to the previous quarter, totalling KRW79 trillion ($58bn).

Vice-chairman Jun Young-hyun expressed concerns over the company’s performance, emphasising the need to regain its technological competitiveness. The statement acknowledged Samsung’s struggle to keep up with market demands, particularly in the rapidly expanding AI chip sector, where rivals like Nvidia have surged ahead. Samsung’s stock fell a further 1.3%, compounding its 20% decline earlier this year.

Samsung has held the title of the world’s largest memory chipmaker for over 30 years but faces increasing competition in both traditional and advanced chip sectors. The delay in the launch of its high-bandwidth memory (HBM3E) chips has been one of the challenges contributing to its weaker performance, with rival SK hynix making significant gains.

Industry analysts highlight that Samsung’s slower response to the AI chip boom has made it more dependent on lower-margin legacy chips. The company’s memory and contract chip manufacturing divisions have both struggled against competition from Chinese manufacturers and Taiwan Semiconductor Manufacturing Company (TSMC). More detailed earnings figures are expected later this month.

Japan investigates generative AI for fair competition

Japan’s Fair Trade Commission has launched an investigation into the rapidly expanding generative AI market. Concerns have been raised about the dominance of US tech companies, particularly in semiconductors and the specialist workforce needed for AI development.

The commission has invited businesses and users to provide input on antitrust risks, with a first report expected next spring. The study aims to identify challenges for new companies entering the AI market, which often depends on advanced semiconductors and vast data resources.

Nvidia’s dominance in the semiconductor market, controlling 80% of chips used for AI, is highlighted as a potential barrier to competition. The commission also noted risks related to monopolisation of specialists by large IT companies and prioritising their own AI products.

Other nations, including the US, European Union, and South Korea, are conducting similar investigations. Study in Japan intends to balance AI’s benefits with ensuring fair market access and competition.

Nvidia dominates AI hardware market amid growing demand

Jensen Huang, Nvidia’s CEO, has described demand for the company’s AI chips as ‘insane’, reflecting the increasing global interest in AI technology. His remarks came as Nvidia announced an expanded partnership with IT consultancy Accenture, aimed at scaling AI solutions for businesses worldwide.

The collaboration will see a new business group formed, focused on building custom AI systems using Nvidia’s cutting-edge technology. The partnership also involves Meta’s open-source AI models, Llama, further reinforcing Nvidia’s position as a major player in the growing AI ecosystem. Huang highlighted the role of the partnership in addressing global AI demand, marking the start of what he termed the ‘enterprise AI’ wave.

As corporations scramble to build AI infrastructure, Nvidia’s dominance in AI hardware, particularly in graphics processing units (GPUs), has been a key driver of the company’s success. Nvidia’s stock has surged, closing 1.6% higher, and more than doubling in value this year, while Accenture’s shares also rose by 1.2%.

Nvidia’s success is driven by widespread adoption of AI across industries such as healthcare, cloud computing, and finance. The partnership with Accenture represents the latest step in Nvidia’s strategy to secure its leadership in the enterprise AI market, which is poised for exponential growth in the coming years.

US to fund AI-driven semiconductor research with $100 million

The US Commerce Department announced its plan to allocate $100 million to promote the use of AI in developing sustainable semiconductor materials. This funding initiative is part of a broader effort overseeing $52.7 billion designated for US chip manufacturing and research, aimed at strengthening the country’s position in the semiconductor industry.

The new funding will support universities, national laboratories, and private sector companies in creating AI-driven autonomous experimentation methods. By harnessing the capabilities of AI, the initiative seeks to streamline and expedite the development of innovative semiconductor materials that are less resource-intensive, ultimately contributing to a more sustainable manufacturing process.

With the semiconductor industry facing increasing pressure to reduce environmental impact, this investment represents a significant step towards integrating advanced technologies to foster sustainable practices. The Commerce Department’s focus on AI in this sector underscores the potential for transformative advancements that can meet both economic and environmental goals, helping to secure a more resilient supply chain for the future.

TSMC activates robust typhoon protocols as typhoon Krathon approaches Taiwan

Taiwan Semiconductor Manufacturing Company (TSMC), the world’s largest contract chipmaker, has initiated its routine typhoon preparations ahead of Typhoon Krathon. The storm is expected to make landfall soon, but TSMC remains confident that its operations will face minimal disruption.

TSMC emphasised that it does not foresee significant impacts on production or business continuity. Its typhoon protocols, which have been developed over years of experience, ensure that any potential risks are effectively mitigated. Move like this one allows the company to continue providing services to its global clients.

Typhoon Krathon has prompted widespread preparation efforts across Taiwan, with residents and businesses securing their properties and essential supplies. The island frequently experiences typhoons, so firms like TSMC are well-versed in storm readiness.

Despite the potential severity of the typhoon, TSMC’s proactive approach and robust procedures highlight its resilience. The company remains focused on maintaining its leadership in the semiconductor industry while ensuring the safety of its workforce.