According to the news from Commercial Times, in a recent press conference, the four major American cloud service providers (CSPs) collectively expressed their intention to expand their investment in AI application services. Simultaneously, they are continuing to enhance their cloud infrastructure. Apple has also initiated its foray into AI development, and both Intel and AMD have emphasized the robust demand for AI servers. These developments are expected to provide a significant boost to the post-market prospects of Taiwan’s AI server supply chain.

Industry insiders have highlighted the ongoing growth of the AI spillover effect, benefiting various sectors ranging from GPU modules, substrates, cooling systems, power supplies, chassis, and rails, to PCB manufacturers.

The American CSP players, including Microsoft, Google, Meta, and Amazon, which recently released their financial reports, have demonstrated growth in their cloud computing and AI-related service segments in their latest quarterly performance reports. Microsoft, Google, and Amazon are particularly competitive in the cloud services arena, and all have expressed optimistic outlooks for future operations.

The direct beneficiaries among Taiwan’s cloud data center suppliers are those in Tier 1, who are poised to reap positive effects on their average selling prices (ASP) and gross margins, driven by the strong demand for AI servers from these CSP giants in the latter half of the year.

Among them, the ODM manufacturers with over six years of collaboration with NVIDIA in multi-GPU architecture AI high-performance computing/cloud computing, including Quanta, Wistron, Wistron, Inventec, Foxconn, and Gigabyte, are expected to see operational benefits further reflected in the latter half of the year. Foxconn and Inventec are the main suppliers of GPU modules and GPU substrates, respectively, and are likely to witness noticeable shipment growth starting in the third quarter.

Furthermore, AI servers not only incorporate multiple GPU modules but also exhibit improvements in aspects such as chassis height, weight, and thermal design power (TDP) compared to standard servers. As a result, cooling solution providers like Asia Vital Components, Auras Technology, and SUNON; power supply companies such as Delta Electronics and Lite-On Technology; chassis manufacturers Chenbro; rail industry players like King Slide, and PCB/CCL manufacturers such as EMC, GCE are also poised to benefit from the increasing demand for AI servers.

(Source: https://ctee.com.tw/news/tech/915830.html)

OpenAI’s ChapGPT, Microsoft’s Copilot, Google’s Bard, and latest Elon Musk’s TruthGPT – what will be the next buzzword for AI? In just under six months, the AI competition has heated up, stirring up ripples in the once-calm AI server market, as AI-generated content (AIGC) models take center stage.

The convenience unprecedentedly brought by AIGC has attracted a massive number of users, with OpenAI’s mainstream model, GPT-3, receiving up to 25 million daily visits, often resulting in server overload and disconnection issues.

Given the evolution of these models has led to an increase in training parameters and data volume, making computational power even more scarce, OpenAI has reluctantly adopted measures such as paid access and traffic restriction to stabilize the server load.

High-end Cloud Computing is gaining momentum

According to Trendforce, AI servers currently have a merely 1% penetration rate in global data centers, which is far from sufficient to cope with the surge in data demand from the usage side. Therefore, besides optimizing software to reduce computational load, increasing the number of high-end AI servers in hardware will be another crucial solution.

Take GPT-3 for instance. The model requires at least 4,750 AI servers with 8 GPUs for each, and every similarly large language model like ChatGPT will need 3,125 to 5,000 units. Considering ChapGPT and Microsoft’s other applications as a whole, the need for AI servers is estimated to reach some 25,000 units in order to meet the basic computing power.

As the emerging applications of AIGC and its vast commercial potential have both revealed the technical roadmap moving forward, it also shed light on the bottlenecks in the supply chain.

The down-to-earth problem: cost

Compared to general-purpose servers that use CPUs as their main computational power, AI servers heavily rely on GPUs, and DGX A100 and H100, with computational performance up to 5 PetaFLOPS, serve as primary AI server computing power. Given that GPU costs account for over 70% of server costs, the increase in the adoption of high-end GPUs has made the architecture more expansive.

Moreover, a significant amount of data transmission occurs during the operation, which drives up the demand for DDR5 and High Bandwidth Memory (HBM). The high power consumption generated during operation also promotes the upgrade of components such as PCBs and cooling systems, which further raises the overall cost.

Not to mention the technical hurdles posed by the complex design architecture – for example, a new approach for heterogeneous computing architecture is urgently required to enhance the overall computing efficiency.

The high cost and complexity of AI servers has inevitably limited their development to only large manufacturers. Two leading companies, HPE and Dell, have taken different strategies to enter the market:

• HPE has continuously strengthened its cooperation with Google and plans to convert all products to service form by 2022. It also acquired startup Pachyderm in January 2023 to launch cloud-based supercomputing services, making it easier to train and develop large models.
• In March 2023, Dell launched its latest PowerEdge series servers, which offers options equipped with NVIDIA H100 or A100 Tensor Core GPUs and NVIDIA AI Enterprise. They use the 4th generation Intel Xeon Scalable processor and introduce Dell software Smart Flow, catering to different demands such as data centers, large public clouds, AI, and edge computing.

With the booming market for AIGC applications, we seem to be one step closer to a future metaverse centered around fully virtualized content. However, it remains unclear whether the hardware infrastructure can keep up with the surge in demand. This persistent challenge will continue to test the capabilities of cloud server manufacturers to balance cost and performance.

(Photo credit: Google)

Smart watch shipments reached 107 million units in 2021, surpassing the shipments of smart bands for the first time. Smart bands declined in 2021 with only 70.33 million units delivered. As branded manufacturers strengthen their investment in the smart watch market, market growth has accelerated. Due to the proximity of functional applications, the smart band market has been gradually supplanted by low-cost smart watch products. Therefore, driven by new products in 2022, the current forecast indicates that smart watch shipments will continue to grow in 2022, reaching 119 million units, for an annual growth rate of 11.4%. Smart band shipments are expected to decline to 55.18 million units and the gap between the two continues to widen. It is worth noting, although branded manufacturers are optimistic regarding the development of the smart watch market, in light of risks such as the COVID-19 pandemic, war, and inflation, continuing downward revisions of smart watch shipment volume in 2022 cannot be ruled out and annual growth rate may even fall below 5%.

The top five major manufacturers of smart watches are Apple, Samsung, Huawei, Xiaomi, and Garmin with a combined market share of 63%. Apple Watch shipments reached 36.6 million units in 2021, with an annual growth rate of 20%. Apple is optimistic regarding the market in 2022. On the one hand, the company believes that the Watch Series 8 can maintain the same consumer interest as the Watch Series 7. On the other hand, a wave of entry-level product replacements is expected with the launch of the second-generation Watch SE. Coupled with the launch of new Pro products, Apple believes that market demand in 2022 will surpass that in 2021 and shipment volume is expected to reach 43 million to 46 million units.

However, considering the negative economic impacts following the wake of the COVID-19 pandemic and inflation, 4Q22 holiday sales may not pan out as expected. Throwing in the wildcard of China’s lockdown strategy, production of new Apple Watches may be affected if lockdown measures reappear in 2H22, which will lead to a risk of further delays in shipments of new products, upsetting shipments in 2022. Therefore, shipments of Apple Watches are conservatively forecast to reach 39.1 million units in 2022.

Although Samsung began exploring the wearable device market very early, due to the small market size and the rapid growth of the smartphone market, smart watches were not the focus of Samsung’s product development at the time. This was true up until recent years. As the growth of smartphones slowed, Samsung has once again strengthened its distribution of other consumer electronics products which has led to continuous growth in Samsung Galaxy Watch shipments. In addition, since Huawei has been flattened by Sino-US trade friction, Samsung leaped to claim the number two spot in the smart watch market in 2021. Samsung smartwatch shipments are estimated to grow from 11.09 million in 2021 to 14.1 million units in 2022.

After Google officially acquired Fitbit, although Fitbit remains nominally an independent brand, internal reorganization is bound to be carried out considering this affords more efficient resource allocation, integrating the companies’ procurement, product development, etc. Google will have two smart watch brands, Fitbit and Pixel. In terms of high-end watches, the Pixel Watch is a flagship product positioned as an extension of the Pixel mobile phone and is primarily used in synergy with smart phone functionality. Therefore, the Pixel Watch uses the Samsung Exynos chip and it is expected to highlight the features of various apps in the Google Play Store. Since Fitbit was acquired by Google, it has become necessary to adjust unified strategy and integrate and reorganize resources. Although smart watch shipments resumed growth in 2021, there is some distance between leading brands. Fitbit Is expected to launch new products in 2H22. Google Pixel Watch is also expected to be launched in 2H22, so Google/Fitbit smart watch shipments are estimated to grow to 3.65 million units in 2022.

（Image credit: Pixabay）

(TechNews) Google confirmed on May 4th that it has acquired Raxium, a start-up company with Micro LED display technology, which is expected to become key in Google’s mission to create a new generation of AR displays.

Google senior vice president of devices and services, Rick Osterloh, who leads the development of Google’s hardware products, stated that Raxium has spent five years creating a small, cost-effective, and energy-efficient high-resolution display that lays the foundation for future display technologies, adding, this company’s technology in this field could play a key role in Google’s hardware investments. Raxium, headquartered in Fremont, California, will be merged into Google’s devices and services group in the future but he did not disclose the purchase price or other details.

According to Raxium’s official website, the pixel pitch of s Super AMOLED screen on a mobile phone is approximately 50 microns but the company’s Micro LED technology can achieve approximately 3.5 microns and it claims to be able to create unprecedented display efficiency.

When foreign media, The Information, reported last month and first exposed Google’s plan to acquire Raxium, it pointed out that Micro LED technology can create AR displays that are more energy-efficient than other solutions while retaining vivid colors. In addition, Raxium is working on the monolithic integration of Micro LEDs, which is expected to significantly reduce costs.

This move makes Google’s plans for subsequent AR hardware products increasingly clear. Google acquired glasses startup North in 2020 and is reportedly recruiting engineers to develop an operating system for AR displays. It was revealed by foreign media in January this year that Google’s laboratory is developing a head-mounted AR device code-named “Project Iris” which is under the same management as “Project Starline” shown at the Google I/O 2021 developer conference last year.

(Source: https://technews.tw/2022/05/05/google-acquires-raxium/)

In 2021, notebook panel shipments reached a record high of 282 million units, with an annual growth rate of 25.1%, according to TrendForce’s research. In the first half of the year, demand was driven by the pandemic and primarily focused on consumer notebooks and Chromebooks while, in the second half of the year, as Europe and the United States gradually lifted lockdowns and work returned to normal, demand largely shifted to commercial models, which continued to support the demand for notebook panels throughout the year.

It is worth noting that TrendForce believes shipment totals of notebook panels from 1Q22 to 2Q22 may be corrected. Notebook panel shipments in 1Q22 are estimated at approximately 67.9 million units, a QoQ decline of 9.7% while 2Q22 shipments are expected to drop to 61.4 million units, down 9.5% QoQ. In addition to the impact of the traditional off-season, there are two reasons for this correction. One is that inventory on the brand-side has increased. Due to the shortage of panels in the past two years, the brand-side continued to purchase panels in 2021 to avoid supply chain disruption. Normally, notebook brands hold 4 to 8 weeks of inventory but some brands have already stocked up to 8 weeks. Two, since a whole notebook requires numerous components, it cannot be assembled and shipped if even one is missing. Limitations impose by incomplete materials lists caused the growth rate of notebook computer shipments to fall behind that of panel shipments, shifting notebook computer panels into oversupply.

Despite this, TrendForce has specifically mentioned, since the profit margin of notebook panels still beats LCD monitor panels and TV panels, panel makers will still desire an increase in the supply of notebook panels. However, in the face of a possible correction in notebook panel shipments, panel makers may accumulate more inventory and deepen the downward pressure on notebook panel pricing.

Looking forward to 2022, panel shipment performance and price trends will be adversely affected by adjustments in notebook brand inventories in 1H22. In 2H22, notebook brands will continue to focus on sales plans for whole notebook computers. The sales performance of these brands during the peak season is still worth looking forward to and the restocking momentum of notebook panels is expected to recover. Current estimates put the shipment of notebook panels at 265 million in 2022, a decrease of 6.0% YoY.