According to a report from wccftech,  previous claims about Xiaomi abandoning the development of its smartphone processor due to high costs were incorrect. Instead, Xiaomi is expected to release its custom solution in the first half of 2025, with its performance rumored to be equivalent to Qualcomm’s Snapdragon 8 Gen 1, which was released two years ago. The chip is said to be manufactured with TSMC’s 4nm process.

The chip, as per wccftech citing sources, will be produced using TSMC’s N4P process,  which is a generation behind the Qualcomm’s Snapdragon 8 Gen 4 and MediaTek’s Dimensity 9400. However, as the shipment volume might be not as high, it is reasonable that Xiaomi might not need to opt for the most advanced manufacturing process.

Although TSMC has already introduced its 3nm process and is advancing towards 2nm, its 4nm N4P process is still competitive, as both the Snapdragon 8 Gen 3 and Dimensity 9300 are produced with N4P.

The sources also indicate that the performance of Xiaomi’s in-house chip is similar to that of Snapdragon 8 Gen 1’s, while the 5G modem chip will be supplied by another Chinese company, Unisoc. Xiaomi’s move towards developing its own smartphone chips is expected to its reduce reliance on Qualcomm and MediaTek.

Shanghai-based fabless chip firm Unisoc, is specialized in areas including 2G/3G/4G/5G, Wi-Fi, Bluetooth, TV FM, satellite communications and other related technologies, according to its website.

Per a previous report from wccftech, Qualcomm executives had hinted that the Snapdragon 8 Gen 4 will be more expensive than the Snapdragon 8 Gen 3, and they may also charge partners for the 5G modem chip. By developing its own chips, Xiaomi can gain valuable experience and gradually reduce its dependence on Qualcomm.

Read more

• [News] MediaTek and Xiaomi Opened a Joint Lab, the First Product with Stellar Performance Unveiled
• [News] China’s Xiaomi and Unisoc Claim Domestic 4nm Mobile Processors Coming Soon

(Photo credit: Xiaomi)

According to a report by 36Kr, China’s EV maker, XPeng Motors, has successfully completed the tape-out process for its self-developed intelligent driving chip.

Sources cited by the report further reveal that XPeng’s intelligent driving chip is specifically designed to meet AI demands, including end-to-end large models. The product is considered to be is a central computing architecture chip that supports integrated cabin and driving functionalities.

The AI computing power of this chip is said to be equivalent to that of three mainstream intelligent driving chips.

Additionally, the report mentions that on August 27th, during XPeng’s 10th anniversary and the launch event for the M03 model, XPeng Motors will officially release details about its self-developed chip.

In response to the rumors surrounding the unveiling of XPeng’s self-developed chip, as per the report, XPeng’s Chairman and CEO hinted on his personal social account that the company certainly won’t disappoint.

Previously, NIO, another automobile manufacturer in China, had also announced the successful tape-out of its 5nm autonomous driving chip, the NX9031.

The tape-outs of self-developed chips marks the beginning of a new phase in which automakers are further competing to enhance the efficiency of intelligent driving software and hardware.

Per a previous report by 36Kr, it was noted that XPeng began building its chip team in 2020. Initially, XPeng collaborated with the U.S. chip design company Marvell, but the partnership did not go smoothly.

Read more

• [News] China Reportedly Demands Automakers Increase Local Chip Procurement to 25%
• [COMPUTEX 2024] Mobility of Tomorrow: When a Car is More Than Just a Car

(Photo credit: XPeng)

China’s industrial profits continued to rise in July, as reported on August 27. According to  the National Bureau of Statistics, the cumulative year-on-year growth rate of profits for large-scale industrial enterprises reached 3.6% from January to July, an increase of 0.1% compared to the January to June period. The year-on-year growth rate for July alone was 4.1%, up 0.5 percentage points from June, marking the second consecutive month of accelerated growth.

This recovery was primarily driven by the high-tech manufacturing sector, which saw a profit growth rate of 12.8%, contributing over 60% to the overall growth. This reflects the Chinese government’s support for high-tech, high-efficiency, and high-quality products.

However, in a recent statement, Yu Wei Ning, an industrial statistician at the National Bureau of Statistics, pointed out that domestic consumer demand remains weak, which has compressed the profits of some industrial enterprises. This situation is also reflected in the July Producer Price Index (PPI), which declined by 0.8%, marking the 22nd consecutive month of negative growth in PPI.

Read more at Datatrack

Amid the AI boom driving a surge in demand for advanced packaging, Samsung Electronics announced in March its ambition to achieve record-high revenue for the business this year, aiming to surpass the USD 100 million mark. The company, which is eager to catch up with TSMC not only on the foundry but also the advanced packaging business, was said to hire former TSMC deputy director Vic Lin as Vice President of the Advanced Packaging Business Unit in its semiconductor department. However, according to a report by ijiwei, the business unit has been disbanded recently, and rumor has it that Chinese semiconductor companies are attempting to recruit Lin.

It is worth noting that before joining TSMC, Lin worked at Micron Technology. Afterwards, during his 19-year tenure at TSMC from 1999 to 2017, Lin was responsible for the application of the semiconductor giant’s over 450 U.S. patents, the report notes. His major accomplishments included securing a major collaboration deal with Apple, as well as laying a solid foundation for TSMC’s expertise in 3D packaging technology.

Nowadays, the advanced packaging business has emerged as one of TSMC’s major growth momentum, with primary customer NVIDIA having the highest demand, occupying about half of the capacity, followed closely by AMD. As the demands for AI and HPC processors keep booming, TSMC revealed plans earlier to further expand its chip-on-wafer-on-substrate (CoWoS) capacity at a compound annual rate (CAGR) of over 60% until at least 2026, according to a report by AnandTech.

After leaving TSMC, Lin became the CEO of Skytech, where his extensive work experience helped him accumulate substantial expertise in packaging equipment manufacturing.

In 2022, Samsung established an Advanced Packaging Task Force, which was later transformed to its Advanced Packaging Business Team in 2023, of which Lin was said to join the team as Vice President, ijiwei notes.

However, industry insiders have revealed that the team was recently disbanded, and its members have returned to Samsung’s memory department and others, the report suggests. Additionally, Lin’s two-year contract with Samsung is said to be expire soon, and it seems unlikely that Samsung will renew it.

Being regarded as a “semiconductor packaging expert,” Lin’s next move is being closely watched. Certain Chinese semiconductor companies are rumored to get in contact with Lin, but it is expected that he will prioritize opportunities to collaborate with semiconductor companies in Taiwan, the report indicates.

The report notes that Samsung has confirmed that the team had been disbanded due to an internal organizational restructuring but declined to comment on personnel matters.

Earlier in May, as part of the restructuring process, the company has disbanded its Robot Business Team as well, which was responsible for developing its first wearable robot, “Bot Fit.”

Read more

• [News] Samsung to Intensify Efforts for Capturing Intel’s Advanced Packaging Orders Shifting from TSMC
• [News] Samsung Reportedly Commits to Advanced Packaging, Targets Over USD100 Million in Related Revenue This Year

(Photo credit: Samsung)

Kioxia has moved forward with its plans to go public, starting the process on August 23. This development follows Bain Capital, a U.S. private equity firm with a majority stake in Kioxia, submitting an application for the listing to the Tokyo Stock Exchange on the same day, according to Nikkei.

Kioxia, formerly Toshiba Memory Corporation, was spun off from Toshiba in 2018 and rebranded. Bain Capital spearheads a special purpose company that, along with South Korea’s SK Hynix, holds a 56% stake in Kioxia Holdings, making it the largest shareholder. Toshiba retains a 41% stake.

SK Hynix first invested in Kioxia in 2018, committing a total of 4 trillion won (around $2.9 billion). This investment was split between 2.7 trillion won into a private equity fund led by Bain Capital and 1.3 trillion won to acquire Kioxia convertible bonds issued by Toshiba. However, the latest report from Korean media BusinessKorea highlights that SK Hynix has faced difficulties recovering its investment due to Kioxia’s delayed IPO, failed merger attempts, and a weak semiconductor market.

Kioxia had planned to list on the Tokyo Stock Exchange in 2020, but escalating trade tensions between the U.S. and China led to a postponement. In 2023, Kioxia attempted to merge with Western Digital’s memory division to better compete with Samsung Electronics in the NAND flash market, but the effort was blocked by SK Hynix.

Despite these challenges, Kioxia posted a net profit of 69.8 billion yen in the second quarter of this year, its highest second-quarter earnings, as demand for memory in smartphones and PCs bottomed out. With signs of a semiconductor market recovery, Kioxia is pushing for a re-listing to enhance its financial flexibility.

According to Nikkei, after the listing, Bain Capital and Toshiba are expected to gradually reduce their stakes through share sales. SK hynix is also likely to sell part of its stake to recover its investment while maintaining strategic ties with Kioxia and Toshiba.

BusinessKorea believes that should Kioxia achieve a high valuation and successfully go public later this year, SK Group could recover its investment, enabling SK Hynix to reinvest in high-bandwidth memory (HBM) for AI applications.

(Photo credit: Kioxia)