Intel, Qualcomm, Google, and other tech giants are reportedly joining forces with over a hundred startups to challenge NVIDIA’s dominance in the market, as per a report from Reuters. Reportedly, their goal is to collectively penetrate the artificial intelligence (AI) software domain, guiding developers to migrate away from NVIDIA’s CUDA software platform.

NVIDIA’s CUDA is a parallel computing platform and programming model designed specifically to accelerate GPU computing. It allows GPU users to fully leverage their chip’s computational power in AI and other applications. As per a previous report from TrendForce, since 2006, NVIDIA has introduced the CUDA architecture, nearly ubiquitous in educational institutions. Thus, almost all AI engineers encounter CUDA during their academic tenure.

However, tech giants are now reportedly aiming to disrupt the current status quo. According to a report from Reuters on March 25th, Intel, Qualcomm, and Google are teaming up to challenge NVIDIA’s dominant position. They plan to provide alternative solutions for developers to reduce dependence on NVIDIA, encourage application migration to other platforms, and thereby break NVIDIA’s software monopoly and weaken its market influence.

The same report from Reuters further indicated that several tech companies have formed the “UXL Foundation,” named after the concept of “Unified Acceleration” (UXL), which aims to harness the power of acceleration computing using any hardware.

The project plans to leverage Intel’s oneAPI technology to develop software and tools supporting multiple AI accelerator chips. The goal is to reduce the technical barriers developers face when dealing with different hardware platforms, streamline the development process, enhance efficiency, and accelerate innovation and application of AI technology.

Vinesh Sukumar, Head of AI and Machine Learning Platform at Qualcomm, stated, “We’re actually showing developers how you migrate out from an NVIDIA platform.”

Bill Magro, Head of High-Performance Computing at Google, expressed, “It’s about specifically – in the context of machine learning frameworks – how do we create an open ecosystem, and promote productivity and choice in hardware.” The foundation is said to aim to finalize technical specifications in the first half of this year and strives to refine technical details by the end of the year.

However, CUDA software has established a solid foundation in the AI field, making it unlikely to be shaken overnight. Jay Goldberg, CEO of financial and strategic advisory firm D2D Advisory, believes that CUDA’s importance lies not only in its software capabilities but also in its 15-year history of usage. A vast amount of code has been built around it, deeply ingraining CUDA in numerous AI and high-performance computing projects. Changing this status quo would require overcoming significant inertia and dependency.

Read more

• [News] NVIDIA’s EULA Amendments Tighten Grip, Suppressing Third-Party CUDA Emulation
• [Insights] Microsoft Unveils In-House AI Chip, Poised for Competitive Edge with a Powerful Ecosystem

(Photo credit: NVIDIA)

According to a report from IJIWEI, there are recent indications that Huawei’s P70 series and smartphones with smaller foldable screens are set to be launched in the first half of this year, with the Kirin 5G chip making a comprehensive return.

Citing information from a blogger, Huawei’s return to the top spot in China’s market in the first week of 2024 is attributed to the outstanding performance of the Nova 12 series. In addition to this success, the first half of this year will witness the release of new models such as the P70 series flagship for imaging and compact foldable screen devices. The Kirin 5G platform is expected to make a full comeback.

In the first half of 2024, Huawei plans to unveil the P70 series and devices with smaller foldable screens, all powered by the Kirin 5G platform. It is reported that the P70 series will feature the Kirin 9010 chip. Looking at Huawei’s recent releases, the Kirin 5G platform is set to cover a broad spectrum, including the Mate series, P series, and nova series, among other mid-to-high-end product lines.

Analyst Ming-Chi Kuo from TF International Securities revealed that Huawei’s smartphone shipments are expected to reach 60 million units in 2024, making it the fastest-growing brand in the industry. Kuo also mentioned that due to Huawei adopting the new Kirin 5G platform, Qualcomm’s System-on-Chip (SoC) shipments to Chinese smartphone brands are anticipated to decrease by 50-60 million units in 2024 compared to 2023.

(Image: Huawei)

Both TSMC and Samsung plan to start production on the 2nm process in 2025, triggering an early battle for related orders. According to sources reported by the Financial Times (FT), Qualcomm intends to shift production of its next-generation high-end mobile chips from TSMC to Samsung’s 2nm process. Samsung, offering substantial discounts, is aggressively pursuing orders from major players like NVIDIA, posing a challenge to TSMC’s dominance.

According to reports citing insider information, TSMC has showcased its 2nm prototype test results to major clients like Apple and NVIDIA. Additionally, sources indicate that Samsung is not only planning to introduce its 2nm prototype but is also offering discounted prices, attracting interest from prominent clients, including NVIDIA.

The report highlights that Qualcomm is planning to use Samsung’s “SF2” (2nm) process for the next generation of high-end smartphone chips. Samsung, as the first company to globally mass-produce 3nm (SF3) chips last year, is also the pioneer in adopting the new Gate-All-Around (GAA) transistor architecture.

Samsung indicated that, “We have fully deployed and can mass-produce SF2 in 2025. Since we are the first company to enter and transform the GAA architecture, we hope that the progress from SF3 to SF2 will be relatively smooth.”

However, insiders have disclosed that Samsung’s yield for the most basic 3nm chips is only 60%, significantly below customer expectations. Moreover, when producing chips with complexity equivalent to Apple’s A17 Pro or NVIDIA graphics processing units (GPUs), the yield may further decrease.

Global giants such as Qualcomm and NVIDIA follow a diversified wafer foundry strategy, but they still rely heavily on TSMC for now. Previously, NVIDIA’s Chief Financial Officer, Colette Kress, hinted during the UBS Global Technology Conference that NVIDIA may consider Intel for the production of its next-generation chips, potentially breaking away from the exclusive partnership with TSMC for AI chips.

Now, Qualcomm is also exploring collaboration with Samsung for the 2nm process, intensifying the pressure on TSMC to address potential order losses from two major clients in advanced semiconductor manufacturing.

TSMC, on the other hand, informed the Financial Times that the development progress of its 2nm process is proceeding smoothly, and it is scheduled for production in 2025. When launched, it will represent the industry’s most advanced semiconductor technology in terms of density and energy efficiency.

In a previous investor conference, TSMC stated that there is significant customer interest and engagement in 2nm for high-speed computing and smartphone applications. It is expected that 2nm, when introduced in 2025, will be the most advanced semiconductor technology in the industry, comparable or even superior to 3nm at the same stage.

TSMC plans to launch the 2nm backside power rail solution in the second half of 2025, with mass production scheduled for 2026.

Besides TSMC and Samsung actively advancing towards 2nm and more advanced processes, Intel has also joined the competition. The Financial Times characterizes this race for the 2nm process as “shaping the future of a USD 500 billion industry.”

Intel is progressing along its previously set 5 process nodes in a four-year trajectory. The Intel 4 process is ready for mass production, and the Intel 3 process is planned for launch later this year. Intel CEO, Pat Gelsinger, has previously showcased the Intel 20A wafer, which is expected to enter the pre-production phase in the first half of next year. The Intel 18A process is scheduled for mass production in the second half of 2024.

Read more

• [News] Intense Competition in Advancing Processes at the 2nm by Samsung, Intel, and TSMC

(Photo credit: Samsung)

According to TechNews’ report, TSMC and Samsung fiercely compete in the semiconductor foundry sector. Earlier market reports suggested that Qualcomm’s Snapdragon 8 Gen 4 mobile processor might adopt a dual-foundry strategy with TSMC and Samsung manufacturing simultaneously.

However, according to the latest industry information, due to Samsung’s conservative expansion plan for next year’s 3nm production capacity and unstable yields, Qualcomm has officially canceled the plan to utilize Samsung for next year’s processors. The dual-sourcing model is now postponed until 2025.

Samsung began mass production of its first-generation 3nm GAA (SF3E) process at the end of June last year, marking Samsung’s initial use of the innovative GAA architecture for transistor technology. The second-generation 3nm process, 3GAP (SF3), will utilize the second-generation MBCFET architecture, optimizing it based on the foundation of the first-generation 3nm SF3E. It is expected to enter mass production in 2024.

The dual-foundry strategy for Qualcomm was initially leaked by the reputable source Revegnus via the X platform (formerly Twitter). It was mentioned that the Snapdragon 8 Gen 4 processor would adopt TSMC’s 3nm (N3E) process, while Samsung’s 3GAP process would be used for the Snapdragon 8 Gen 4 supplying Samsung’s Galaxy series smartphones. Other sources suggested that due to limited capacity at TSMC’s 3nm production, Qualcomm had to seek Samsung as an alternative chip foundry.

As a result, Qualcomm originally anticipated dual-foundry production with both TSMC and Samsung in 2024, with hopes of being the first customer for the 3GAP process. However, considering Samsung’s conservative 3nm production capacity plan for next year and the instability in yields, Qualcomm decided to scrap the plan and exclusively rely on TSMC, pushing the dual-foundry strategy to 2025.

Currently, TSMC’s 3nm process technology capacity is on the rise, with expectations that by the end of 2024, monthly production capacity will reach 100,000 wafers, and the revenue contribution will increase from the current 5% to 10%.

(Photo credit: Qualcomm)

Read more

• [News] IC Design Companies Seek Advanced Process Second Source, Overview of Competition Between TSMC and Samsung
• [News] Samsung Reportedly Lands a 4nm Mega Order – Why is AMD Switching to “Dual Foundry Mode” for Its Next-Gen Chips?

According to TechNews’ report, Apple, NVIDIA, AMD, Qualcomm, and MediaTek all utilize TSMC’s semiconductor processes for manufacturing their latest chips, with some potentially employing Samsung’s foundry, though typically not for flagship products.

With Samsung’s improved yield rates in recent months, the company is eager to secure a portion of the orders, particularly for the 3-nanometer GAA (Gate-All-Around) process.

Earlier market reports suggested that Qualcomm’s Snapdragon 8 Gen 4 might adopt a dual-foundry strategy, simultaneously utilizing TSMC’s N3E process technology and Samsung’s SF3E process technology.

However, both Qualcomm and MediaTek currently plan to employ TSMC’s second-generation 3-nanometer process technology (N3E) for manufacturing chips like the Snapdragon 8 Gen 4 and Dimensity 4, without pursuing a dual-foundry strategy at this time.

As of the end of June 2022, Samsung announced the commencement of production for 3-nanometer process chips at its Hwaseong Industrial Complex in South Korea. These chips incorporate a new GAA transistor architecture technology, rumored to be more energy-efficient compared to TSMC’s 3-nanometer FinFET technology. Despite this, in the realm of 3nm, Samsung has yet to secure substantial orders from major clients.

Interestingly, the company has seen more success in the 4nm domain. It is reported that Samsung has gradually addressed yield and various issues in the 4-nanometer process technology domain. The third generation of 4-nanometer process technology has seen improvements in performance, reduced power consumption, increased density, and achieved yields close to TSMC’s level. Market sources indicate that Samsung has gained recognition from companies like AMD and Tesla, securing new orders.

Currently, TSMC’s 3-nanometer process technology production capacity is ramping up, with an expected monthly capacity of 100,000 wafers by the end of 2024. The revenue contribution is projected to increase from the current 5% to 10%.

Meanwhile, Samsung plans to introduce the second generation of its 3-nanometer process technology, named SF3 (3GAP), in 2024. Building upon the existing SF3E, it aims for further optimization, and Samsung’s in-house Exynos 2500 is expected to be one of the first high-performance chips to adopt this new process technology.

Read more

• [News] Samsung Reportedly Lands a 4nm Mega Order – Why is AMD Switching to “Dual Foundry Mode” for Its Next-Gen Chips?
• [News] A battle on 4nm: AMD Teams Up with Samsung, while Google Weighs Supplier Split