Sam Altman, the CEO of OpenAI, the developer of the ChatGPT, is reportedly expected to visit Korea on January 26th. Altman may hold meetings with top executives from Samsung Electronics and SK Group to strengthen their collaboration on High-Bandwidth Memory (HBM).

According to sources cited by The Korea Times, Sam Altman is making a slight adjustment for the potential meeting details with Samsung Electronics’ Chairman Lee Jae-yong and SK Group’s Chairman Chey Tae-won. 

OpenAI is set to engage in discussions with Samsung Electronics and SK Group to collaboratively develop artificial intelligence (AI) semiconductors, as part of OpenAI’s strategy to reduce heavy reliance on the AI chip leader NVIDIA.

Reportedly, Altman visited Korea in June of last year, and this upcoming visit is expected to last only about six hours. Most of the time is anticipated to be spent in closed-door meetings with leaders of Korean chip companies or other high-profile executives. 

Altman is keen on strengthening relationships with Korean startups and chip industry players, as it contributes to OpenAI’s development of large-scale language models, powering ChatGPT. OpenAI unveiled its latest model, GPT-4 Turbo, at the end of last year and is currently proceeding with planned upgrades to related services.

Regarding this matter, The Korea Times also cited a spokesman at SK Group, indicating that SK Group also did not confirm whether Chey and Altman will meet.

“Nothing specific has been confirmed over our top management’s schedule with Altman,” an official at SK Group said.

Read more

• [News] Samsung and Naver Jointly Invest in New AI Chip, as Naver Reportedly Claiming an Eightfold Performance Boost Compared to NVIDIA H100

(Photo credit: OpenAI)

NVIDIA CEO Jensen Huang has reportedly gone to Taiwan once again, with reports suggesting a recent visit to China. Industry sources believe NVIDIA is planning to introduce downgraded AI chips to bypass U.S. restrictions on exporting high-end chips to China. Huang’s visit to China is seen as an effort to alleviate concerns among customers about adopting the downgraded versions.

Experts indicate that due to the expanded U.S. semiconductor restriction on China, NVIDIA’s sales in the Chinese market will decline. To counter this, NVIDIA might adjust its product portfolio and expand sales of high-end AI chips outside China.

The export of NVIDIA’s A100 and H100 chips to China and Hong Kong was prohibited in September 2022. Following that, the A800 and H800 chips, which were further designed with downgraded adjustments for the Chinese market, were also prohibited for export to China in October of the previous year.

In November 2023, the NVIDIA’s management acknowledged the significant impact of the U.S. restrictions on China’s revenue for the fourth quarter of 2023 but expressed confidence that revenue from other regions can offset this impact.

CEO Jensen Huang revealed in December in Singapore that NVIDIA was closely collaborating with the U.S. government to ensure compliance with export restrictions on new chips for the Chinese market.

According to reports in Chinese media The Paper, Jensen Huang recently made a low-profile visit to China. The market is closely watching the status of NVIDIA’s AI chip strategy in China and the company’s subsequent development strategies in response to U.S. restrictions. The fate of the newly designed AI chips, H20, L20, and L2, to comply with U.S. export regulations remains uncertain and will be closely observed.

Liu Pei-Chen, a researcher and director at the Taiwan Institute of Economic Research, discussed with CNA’s reporter about NVIDIA’s active planning to introduce a downgraded version of AI chips in China. 

The most urgent task, according to Liu, is to persuade Chinese customers to adopt these downgraded AI chips. Chinese clients believe that there isn’t a significant performance gap between NVIDIA’s downgraded AI chips and domestically designed AI chips.

Liu mentioned that this is likely the reason why Jensen Huang visited China. It serves as an opportunity to promote NVIDIA’s downgraded AI chips and alleviate concerns among Chinese customers. 

Read more

• [News] Chinese Companies Reportedly Disassembling NVIDIA Gaming Cards for AI Development
• [Insights] NVIDIA Maintains Confidence Despite Export Restrictions, Continues Unveiling New Products

(Photo credit: NVIDIA)

According to a report by TechNews, Taiwan has introduced its largest-ever investment deduction incentives under the “Statute for Industrial Innovation,” often referred to as the “Taiwanese Chip Act.” Articles 10-2 and 72 of the statute came into effect, and the Ministry of Economic Affairs announced that it would accept company applications from February 1 to May 31 this year.

The Ministry of Economic Affairs stated that applications for deductions would be accepted starting February this year. The provided tax incentives include a 25% deduction for research and development expenses and a 5% deduction for expenditures on acquiring new eqipment for advanced processes, all of which can be deducted from the current year’s corporate income tax.

Eligibility criteria include companies with research and development expenses of at least NT$6 billion, a research and development density of 6%, and expenditures of NT$10 billion for the purchase of equipment for advanced processes, with no restrictions on industry category.

The Ministry of Economic Affairs emphasized that as the parent law already specifies an effective tax rate of 12% for the fiscal year 112 and a threshold of 15% from the fiscal year 113 onwards, this measure aims to encourage businesses that do not meet these tax rate qualifications to strive for them and become eligible for tax incentives.

A review panel will be formed to assess whether applying companies meet the criteria for a critical position in the international supply chain and other qualification requirements.

The Ministry of Economic Affairs shared that the application period for Article 10-2 of the Statute for Industrial Innovation is from February 1 to May 31 this year. Companies are required to provide explanatory documents and supporting evidence, including data on products, international market share, rankings, import-export trade, and other statistics, serving as indicators for the assessment of technological innovation and critical positions.

According to the financial reports of publicly listed companies in 2022, including TSMC, MediaTek, Realtek, Novatek, Delta Electronics, Nanya Technology, Phison and Winbond, their research and development expenses and research and development density all meet the application thresholds.

(Image: TSMC)

According to recent reports, Huawei is expected to unveil its flagship P70 series later this year, alongside the introduction of the new Kirin 9010 chipset. However, there are indications that the older Kirin 9000S might be utilized in a specific model.

Wccftech suggests that the P70 series will include the P70, P70 Pro, and P70 Art, followed by the Mate 70 series. Notably, not all P70 models will feature the new Kirin 9010.

As per insights from the Weibo account Smart Pikachu, the P70 series will boast a custom curved display that is easy on the eyes and power-efficient but lacks a 2K resolution, and the standard version of the P70 is tested with the Kirin 9000S. This may potentially impact the motivation for users who have already purchased the Mate 60 and might not find sufficient reasons to upgrade to the P70.

Wccftech suggests that the adoption of the 9000S in some models could be attributed to the limited supply of the Kirin 9010. The Kirin 9000S, produced by SMIC using a 7nm process, faces production challenges due to the use of older-generation DUV equipment, resulting in a time-consuming and costly manufacturing process with lower yields.

Despite this, there is a glimmer of hope for Huawei’s pricing competitiveness, as the production cost of the Kirin 9000S is expected to be lower than that of the Kirin 9010. This cost advantage could potentially contribute to Huawei’s goal of reaching an estimated shipment volume of 100 million smartphones in 2024, especially considering the company’s historical strength in offering competitive pricing for its base models.

(Image: Huawei)

In the intense battle of AI chips between NVIDIA and AMD this year, AMD’s MI300 has entered mass production and shipment 1H24, gaining positive adoption from clients. In response, NVIDIA is gearing up to launch upgraded AI chips. TSMC emerges as the big winner by securing orders from both NVIDIA and AMD.

Industry sources have revealed optimism as NVIDIA’s AI chip shipment momentum is expected to reach around 3 million units this year, representing multiple growth compared to 2023.

With the production ramp-up of the AMD MI300 series chips, the total number of AI high-performance computing chips from NVIDIA and AMD for TSMC in 2024 is anticipated to reach 3.5 million units. This boost in demand is expected to contribute to the utilization rate of TSMC’s advanced nodes.

According to a report from the Economic Daily News, TSMC has not commented on rumors regarding customers and orders.

Industry sources have further noted that the global AI boom ignited in 2023, and 2024 continues to be a focal point for the industry. A notable shift from 2023 is that NVIDIA, which has traditionally dominated the field of high-performance computing (HPC) in AI, is now facing a challenge from AMD’s MI300 series products, which have begun shipping, intensifying competition for market share.

Reportedly, the AMD MI300A series products have commenced mass production and shipment this quarter. The central processing unit (CPU) and graphics processing unit (GPU) tile are manufactured using TSMC’s 5nm process, while the IO tile use TSMC’s 6nm process.

These chips are integrated through TSMC’s new System-on-Integrated-Chip (SoIC) and Chip-on-Wafer-on-Substrate (CoWoS) advanced packaging technologies. Additionally, AMD’s MI300X, which does not integrate the CPU, is also shipping simultaneously.

Compared to NVIDIA’s GH200, which integrates CPU and GPU, and the H200, focusing solely on GPU computation, AMD’s new AI chip performance exceeds expectations. It offers a lower price and a high cost-performance advantage, attracting adoption by ODMs.

In response to strong competition from AMD, NVIDIA is upgrading its product line. Apart from its high-demand H200 and GH200, NVIDIA is expected to launch new products such as B100 and GB200, utilizing TSMC’s 3nm process, by the end of the year.

Read more

• [News] Nvidia vs. AMD, the Expensive Duel as Two Major Buyers Opt for AMD’s Latest AI GPU
• [News] TSMC’s CoWoS Demand Surges from NVIDIA, Apple, AMD, Broadcom, Marvell, Monthly Capacity Up 120% in 2024

(Photo credit: NVIDIA)