Microsoft announced the in-house AI chip, Azure Maia 100, at the Ignite developer conference in Seattle on November 15, 2023. This chip is designed to handle OpenAI models, Bing, GitHub Copilot, ChatGPT, and other AI services. Support for Copilot, Azure OpenAI is expected to commence in early 2024.

TrendForce’s Insights:

1. Speculating on the Emphasis of Maia 100 on Inference, Microsoft’s Robust Ecosystem Advantage is Poised to Emerge Gradually

Microsoft has not disclosed detailed specifications for Azure Maia 100. Currently, it is known that the chip will be manufactured using TSMC’s 5nm process, featuring 105 billion transistors and supporting at least INT8 and INT4 precision formats. While Microsoft has indicated that the chip will be used for both training and inference, the computational formats it supports suggest a focus on inference applications.

This emphasis is driven by its incorporation of the less common INT4 low-precision computational format in comparison to other CSP manufacturers’ AI ASICs. Additionally, the lower precision contributes to reduced power consumption, shortening inference times, enhancing efficiency. However, the drawback lies in the sacrifice of accuracy.

Microsoft initiated its in-house AI chip project, “Athena,” in 2019. Developed in collaboration with OpenAI. Azure Maia 100, like other CSP manufacturers, aims to reduce costs and decrease dependency on NVIDIA. Despite Microsoft entering the field of proprietary AI chips later than its primary competitors, its formidable ecosystem is expected to gradually demonstrate a competitive advantage in this regard.

2. U.S. CSP Manufacturers Unveil In-House AI Chips, Meta Exclusively Adopts RISC-V Architecture

Google led the way with its first in-house AI chip, TPU v1, introduced as early as 2016, and has since iterated to the fifth generation with TPU v5e. Amazon followed suit in 2018 with Inferentia for inference, introduced Trainium for training in 2020, and launched the second generation, Inferentia2, in 2023, with Trainium2 expected in 2024.

Meta plans to debut its inaugural in-house AI chip, MTIA v1, in 2025. Given the releases from major competitors, Meta has expedited its timeline and is set to unveil the second-generation in-house AI chip, MTIA v2, in 2026.

Unlike other CSP manufacturers, both MTIA v1 and MTIA v2 adopt the RISC-V architecture, while other CSP manufacturers opt for the ARM architecture. RISC-V is a fully open-source architecture, requiring no instruction set licensing fees. The number of instructions (approximately 200) in RISC-V is lower than ARM (approximately 1,000).

This choice allows chips utilizing the RISC-V architecture to achieve lower power consumption. However, the RISC-V ecosystem is currently less mature, resulting in fewer manufacturers adopting it. Nevertheless, with the growing trend in data centers towards energy efficiency, it is anticipated that more companies will start incorporating RISC-V architecture into their in-house AI chips in the future.

3. The Battle of AI Chips Ultimately Relies on Ecosystems, Microsoft Poised for Competitive Edge

The competition among AI chips will ultimately hinge on the competition of ecosystems. Since 2006, NVIDIA has introduced the CUDA architecture, nearly ubiquitous in educational institutions. Thus, almost all AI engineers encounter CUDA during their academic tenure.

In 2017, NVIDIA further solidified its ecosystem by launching the RAPIDS AI acceleration integration solution and the GPU Cloud service platform. Notably, over 70% of NVIDIA’s workforce comprises software engineers, emphasizing its status as a software company. The performance of NVIDIA’s AI chips can be further enhanced through software innovations.

On the contrary, Microsoft possess a robust ecosystem like Windows. The recent Intel Arc GPU A770 showcased a 1.7x performance improvement in AI-driven Stable Diffusion on Microsoft Olive, this demonstrates that, similar to NVIDIA, Microsoft has the capability to enhance GPU performance through software.

Consequently, Microsoft’s in-house AI chips are poised to achieve superior performance in software collaboration compared to other CSP manufacturers, providing Microsoft with a competitive advantage in the AI competition.

Read more

• [News] Microsoft First In-House AI Chip “Maia” Produced by TSMC’s 5nm

Rumors swirl around AMD’s upcoming chip architecture, codenamed “Prometheus,” featuring the Zen 5C core. As reported by TechNews, the chip is poised to leverage both TSMC’s 3nm and Samsung’s 4nm processes simultaneously, marking a shift in the competitive landscape from process nodes, yield, and cost to factors like capacity, ecosystem, and geopolitics, are all depends on customer considerations.

Examining yields, TSMC claims an estimated 80% yield for its 4nm process, while Samsung has surged from 50% to an impressive 75%, aligning with TSMC’s standards and raising the likelihood of chip customers returning. Speculation abounds that major players such as Qualcomm and Nvidia may reconsider their suppliers, with industry sources suggesting Samsung’s 4nm capacity is roughly half of TSMC’s.

Revegnus, a reputable X(formerly Twitter) source, unveiled information from high-level Apple meetings, indicating a 63% yield for TSMC’s 3nm process but at double the price of the 4nm process. In the 4nm realm, Samsung’s yield mirrors TSMC’s, with Samsung showing a faster-than-expected yield recovery.

Consequently, with Samsung’s significant improvements in yield and capacity, coupled with TSMC’s decision to raise prices, major clients may explore secondary suppliers to diversify outsourcing orders, factoring in considerations such as cost and geopolitics. Recent reports suggest Samsung is in final negotiations for a 4nm collaboration with AMD, planning to shift some 4nm processor orders from TSMC to Samsung.

Beyond AMD, the Tensor G3 processor in Google’s Pixel 8 series this year adopts Samsung’s 4nm process. Samsung’s new fabs in Taylor, Texas, sees its inaugural customer in its Galaxy smartphones, producing Exynos processors.

Furthermore, Samsung announced that U.S.-based AI solution provider Groq will entrust the company to manufacture next-generation AI chips using the 4nm process, slated to commence production in 2025, marking the first order for the new Texas plant.

Regarding TSMC’s 4nm clients, alongside longstanding partners like Apple, Nvidia, Qualcomm, MediaTek, AMD, and Intel, indications propose a potential transition to TSMC’s 4nm process for Tensor G4, while Tensor G5 will be produced using TSMC’s 3nm process. Ending the current collaboration with Samsung, TSMC’s chip manufacturing debut is anticipated to be delayed until 2025.

Last year, rumors circulated about Tesla, the electric vehicle giant, shifting orders for the 5th generation self-driving chip, Hardware 5 (HW 5.0), to TSMC. This decision was prompted by Samsung’s lagging 4nm process yield at that time. However, with Samsung’s improved yield, industry inclination leans towards splitting orders between the two companies.

Google is set to follow the lead of industry giants like Apple and Samsung by manufacturing its Pixel 8 series smartphones in India, with plans to supply the market starting in 2024.

“Today we see an even greater opportunity to make Pixel smartphones available to more people in India, and are very excited to announce our plan to manufacture Pixel smartphones in India. We intend to start with the Pixel 8, and will partner with international and domestic manufacturers to produce Pixel smartphones locally. We expect these devices to start to roll out in 2024, joining India’s ‘Make in India’ initiative,” said Rick Osterloh, Senior Vice President of Devices and Services at Google, emphasizing the significant opportunity for Google in serving Indian consumers with Pixel phones.

However, Google has not disclosed specific production quantities or the proportion of Pixel phones that will be manufactured in India, nor have they revealed information about their manufacturing partners’ factory locations.

According to insiders, India’s largest contract manufacturing company, Dixon Technologies, and Foxconn’s Indian subsidiary are among the competitors in this effort to produce Pixel phones.

Under the policies promoted by Prime Minister Narendra Modi, India has attracted increased investments from Apple. This year, Apple has not only opened two Apple-owned stores in India but has also moved more of its iPhone production from China to India. The recently launched iPhone 15 is the first iPhone manufactured in India.

A majority of Samsung smartphones sold in India are produced at Samsung’s Noida facility, the largest smartphone manufacturing plant globally, which manufactures the Galaxy A and M series. With the introduction of the Galaxy S23 series earlier this year, Samsung has also confirmed that the Galaxy S23 series for the Indian market will be produced locally. In addition to Apple and Samsung, Chinese Android smartphone manufacturers have also established partnerships with local Indian manufacturers.

Read more

• [Insights] Google Teams Up with HP to Produce Chromebooks in India

(Photo credit: Pixabay)

On October 2, 2023, the CEO of Alphabet announced that Google, under its umbrella, is partnering with the leading laptop brand, HP, to manufacture Chromebook laptops in India. This move comes at a time when Dell, Asus, and other laptop manufacturers have begun production in India. HP is leveraging its existing local factories and government incentives, aiming to capitalize on the vast population and educational opportunities in India. In a similar vein, Samsung and Google have collaborated in Vietnam to launch the Galaxy Chromebook Go, targeting the education sector. The Chromebook is gradually expanding its production and introduction to various regions.

TrendForce’s Insights:

1. Google Teams Up with HP to Launch Chromebook, Chromebook Plus Aims to Attract New Consumer Segments

Through the collaboration of HP, a dominant player in the global laptop market, and Google, renowned for its software expertise in the Chrome operating system, Chromebooks are set to be manufactured in India for the first time. This partnership leverages the power of two industry giants to produce Chromebooks locally, offering a cost-effective alternative to the long-standing dominance of white label tablets in India’s education market. Chromebooks, equipped with the Chrome OS, known for easy management and robust security, are set to benefit from government incentives and manufacturer collaborations, working together to create affordable, secure, and high-quality devices. This effort aims to enhance the learning experience for Indian students.

HP, since 2020, has been utilizing the Flex Ltd. factory in Chennai, India, for the production of laptops and desktops. On October 2, 2023, they extended their production line to include Chromebooks. The factory will assemble various types of devices, including laptops, desktops, and Chromebooks, all designed to cater to the local market’s needs. This move is expected to enhance HP’s brand value and market share in India.

In addition, Google and various leading brands launched the Chromebook Plus in the North American market on October 8, 2023. The new models come in screen sizes of 14 inches and 15.6 inches, boasting high-end hardware configurations and AI features. The base model starts at $399.99, with a price difference of less than $200 compared to conventional laptops currently sold in India. This suggests the possibility of local production for the Google Chromebook Plus in India, aiming to attract a diverse range of consumers.

2. Apart from India’s policies and incentives, Vietnam also Attracts Foreign Investment with EU Tariff Benefits and Low Labor Costs

India, with its population of 1.4 billion, presents a dual advantage of a vast labor force and a significant domestic market. After the successful local assembly of iPhones in India, the nation is taking strides toward its goal of local manufacturing. The Indian government has initiated the “Make in India” policy, aiming to entice the production, assembly, and shipment of more electronic end-products in the country. In August 2023, the Indian government announced a delayed implementation of import restrictions on computer products, with a decision pending a year later. Currently, many brand manufacturers can still import fully assembled products into India, temporarily avoiding the impact of high tariffs on imported materials and equipment. Brand manufacturers may also utilize this period to actively collaborate with the government in planning local production initiatives in the region.

Beyond India, numerous Taiwanese assembly plants have made investments in Vietnam. In contrast to India’s policy-driven approach to local manufacturing, Vietnam offers advantages such as EU tariff preferences and low labor costs. For example, Samsung has previously established its dominance in Vietnam by assembling panel modules and laptops. As a result, in collaboration with Google, they introduced the Galaxy Chromebook Go in Vietnam, targeting local markets with simplified, lower-end laptop configurations. It is speculated that other Taiwanese manufacturing plants in Vietnam will follow this operating model.

As the overall technological and industrial capabilities in Vietnam continue to improve, they are poised to take on mid to high-end laptop models by 2024-2025. While American brands continue to rely on China for the production of consumer and business laptops, the collaboration between Google Chromebook and other brand manufacturers in regions like India and Vietnam is expected to become increasingly close.

Google focuses on AI and sensor upgrades with Pixel Watch 2, while Charge 6 smart wristbands may stand out in the market through AI applications.

TrendForce’s Insights:

In early October 2023, Google held its annual fall product launch event, introducing new smartphones, smartwatches, earphones, and AI assistant services. Regarding the Pixel Watch 2, Google opted for a relatively conservative upgrade compared to its competitors, with more significant changes expected in the smart wristband market through the application of generative AI.

1. Google’s New Products Align with Industry Trends, Focusing on AI – Pixel Watch 2 Highlights Sensor Upgrades

The third quarter of the year is an important period for major tech companies to hold product launch events. Following Apple’s event in September and the recently concluded Meta event, Google also hosted its Made By Google fall event in early October. Much like these other companies, Google focused on AI as the central theme of the event, enhancing its range of products and services. The highlights of this event include the Pixel 8 series smartphones, featuring AI-powered image editing, the Pixel Watch 2 with new sensors and monitoring capabilities, and the Pixel Buds Pro smart headphones that can adjust audio settings based on the user’s conversations. Additionally, Google, being a leading player in Generative AI, introduced an AI version of its Google Assistant service called “Assistant with Bard,” which is expected to enhance its app offerings through quick organization, predictive capabilities, and content generation, with deployment on a wider range of devices in the future.

A closer look at the specifications of the Pixel Watch 2 reveals that, compared to its predecessor released just a year ago, the hardware upgrades are not substantial. The primary differences include changes in the watch case material to align with environmental concerns, an upgraded processor, and a slight increase in battery capacity. However, the most significant improvements are in the sensors. The Pixel Watch 2 builds upon its existing features by introducing an electronic sensor for continuous skin conductance monitoring and a skin temperature sensor. Additionally, it has replaced the previous optical heart rate sensor with a multi-path sensor, enabling more precise measurements in different physiological conditions. This allows the watch to provide users with more accurate data, including temperature, heart rate variability (HRV), and other metrics.

Enhanced by advanced AI algorithms, the Pixel Watch 2 can analyze users’ sleep quality, stress levels, activity duration, and calorie expenditure. This allows the watch to provide all-day body response tracking, stress notifications, and guided breathing exercises, offering a range of new features.

2. Google Struggles in the Smartwatch Market; Smart Wristbands Poised for Market Differentiation through AI

Google’s latest release, the Pixel Watch 2, follows a strategy similar to that of other major players like Samsung and Apple in the smartwatch market. It represents a modest upgrade with little change in external appearance, and hardware specifications closely align with existing products in the market. On the software front, besides the new stress management features, other additions such as security checks, emergency sharing, and fall detection closely resemble those of the previous generation.

The primary reason for this conservative approach lies in the challenging global economic conditions and a declining industry landscape. Google’s smartwatch products face relative weakness in the market, lacking the extensive user base and brand loyalty enjoyed by giants like Apple, Samsung, and Huawei. Google also lacks the niche market segmentation seen in players like Garmin. These factors collectively position Google in a cautious stance regarding the development of smartwatches.

While Google has displayed a bit of hesitation in the smartwatch sector, its recent introduction, the Fitbit Charge 6 smart band, has garnered significant attention. Although its price exceeds that of the Huawei Band 8 and Xiaomi 8 Active, released later in 2023, by several folds, the Charge 6 leverages Google’s resources and hardware-software upgrades to emphasize its value. For instance, it incorporates a new machine learning algorithm derived from the Pixel Watch, ensuring more accurate heart rate monitoring. This marks Google’s first smart band integrated with Google Apps, offering seamless integration with widely used applications such as Google Maps, Google Wallet, YouTube Music, and more.

Furthermore, Google has hinted at the development of generative AI features during the Made by Google event. These AI-driven features are anticipated to analyze fitness trends and provide insights through chatbots, with potential integration into smartwatches and smart bands. With these advancements, Google aims to elevate its smart bands’ reputation through AI applications, setting it apart in the market, even in comparison to competitors like Huawei and Xiaomi.