Recently, according to the FT, citing the latest forecast data of the market research institute SemiAnalysis, AI chip giant NVIDIA will ship more than 1 million new NVIDIA H20 acceleration chips to the Chinese market this year, and it is expected that the cost of each chip is between $12,000 and $13,000. This is expected to generate more than $12 billion in revenue for Nvidia. Affected by the United States export control policy, Nvidia's advanced AI chip exports to China have been restricted, H20 is Nvidia based on H100 specifically for the Chinese market to launch the three "castration version" GPU among the strongest performance, but its AI performance is only less than 15% of H100, some performance is even less than the domestic Ascend 910B. When Nvidia launched the new H20 in the spring of this year, there were reports that due to the large castration of H20 performance, coupled with the high price, Chinese customers' interest in buying is insufficient, and they will turn more to choose China's domestic AI chips. Then there are rumors that Nvidia has lowered the price of the H20 in order to improve its competitiveness. However, the latest news shows that due to supply issues caused by the low yield of the Ascend 910B chip, Chinese manufacturers in the absence of supply and other better options, Nvidia H20 has started to attract new purchases from Chinese tech giants such as Baidu, Alibaba, Tencent and Bytedance. Analysts at both Morgan Stanley and SemiAnalysis said the H20 chip is now being shipped in bulk and is popular with Chinese customers, despite its performance degradation compared to chips Nvidia sells in the United States.

An artificial intelligence (AI) team at Stanford University apologized for plagiarizing a large language model (LLM) from a Chinese AI company, which became a trending topic on the Chinese social media platforms, where it sparked concern among netizens on Tuesday. We apologize to the authors of MiniCPM [the AI model developed by a Chinese company] for any inconvenience that we caused for not doing the full diligence to verify and peer review the novelty of this work, the multimodal AI model Llama3-V's developers wrote in a post on social platform X. The apology came after the team from Stanford University announced Llama3-V on May 29, claiming it had comparable performance to GPT4-V and other models with the capability to train for less than $500. According to media reports, the announcement published by one of the team members quickly received more than 300,000 views. However, some netizens from X found and listed evidence of how the Llama3-V project code was reformatted and similar to MiniCPM-Llama3-V 2.5, an LLM developed by a Chinese technology company, ModelBest, and Tsinghua University. Two team members, Aksh Garg and Siddharth Sharma, reposted a netizen's query and apologized on Monday, while claiming that their role was to promote the model on Medium and X (formerly Twitter), and that they had been unable to contact the member who wrote the code for the project. They looked at recent papers to validate the novelty of the work but had not been informed of or were aware of any of the work by Open Lab for Big Model Base, which was founded by the Natural Language Processing Lab at Tsinghua University and ModelBest, according to their responses. They noted that they have taken all references to Llama3-V down in respect to the original work. In response, Liu Zhiyuan, chief scientist at ModelBest, spoke out on the Chinese social media platform Zhihu, saying that the Llama3-V team failed to comply with open-source protocols for respecting and honoring the achievements of previous researchers, thus seriously undermining the cornerstone of open-source sharing. According to a screenshot leaked online, Li Dahai, CEO of ModelBest, also made a post on his WeChat moment, saying that the two models were verified to have highly similarity in terms of providing answers and even the same errors, and that some relevant data had not yet been released to the public. He said the team hopes that their work will receive more attention and recognition, but not in this way. He also called for an open, cooperative and trusting community environment. Director of the Stanford Artificial Intelligence Laboratory Christopher Manning also responded to Garg's explanation on Sunday, commenting "How not to own your mistakes!" on X. As the incident became a trending topic on Sina Weibo, Chinese netizens commented that academic research should be factual, but the incident also proves that the technology development in China is progressing. Global Times

SEOUL, July 9 (Reuters) - Samsung Electronics (005930.KS), opens new tab said on Tuesday it won an order from Japanese artificial intelligence company Preferred Networks to make chips for AI applications using the South Korean firm's 2-nanometre foundry process and advanced chip packaging service. It is the first order Samsung has revealed for its cutting-edge 2-nanometre chip contract manufacturing process. Samsung did not elaborate on the size of the order. The chips will be made using high-tech chip architecture known as gate all-around (GAA) and multiple chips will be integrated in one package to enhance inter-connection speed and reduce size, Samsung said in a statement. South Korea's Gaonchips Co (399720.KQ), opens new tab designed the chips, Samsung said. The chips will go toward Preferred Networks' high-performance computing hardware for generative AI technologies such as large language models, Junichiro Makino, Preferred Networks vice president and chief technology officer of computing architecture, said in the statement.

On the afternoon of July 8, local time, the Vietnamese Ministry of Health issued a notice stating that an 18-year-old girl in the country died of diphtheria. The Ministry of Health asked Nghe An Province and Bac Giang Province to take urgent action to control the epidemic. Diphtheria is an acute respiratory infectious disease caused by Corynebacterium diphtheriae, which is mainly transmitted through droplets and can also be indirectly transmitted by contact with objects containing Corynebacterium diphtheriae. Severe cases may show symptoms of poisoning throughout the body, complicated by myocarditis and peripheral nerve paralysis.

HONG KONG, July 9 (Reuters) - Nornickel (GMKN.MM), opens new tab is in talks with China Copper to form a joint venture that would allow the Russian mining giant to move its entire copper smelting base to China, four sources with knowledge of the matter told Reuters. If the move goes ahead, it would mark Russia's first uprooting of a domestic plant since the U.S. and Britain banned metal exchanges from accepting new aluminium, copper and nickel produced by Russia. It also means Nornickel's copper will be produced within the country where it is most consumed. Nornickel said in April it planned to close its Arctic facility and build a new plant in China with an unnamed partner. Executives at China Copper, owned by the world's largest aluminium producer Chinalco (601600.SS), opens new tab, flew to Moscow in June to discuss a possible joint venture, one of the sources said, adding that details of the structure and investment are still under discussion. Nornickel declined to comment. Chinalco and China Copper did not respond to requests for comment via email and phone. Sites being considered in China include Fangchenggang and Qinzhou in the Guangxi region, the two sources said, with another source saying Qingdao in Shandong province was also possible. A decision on a joint venture will be made over the next few months, a fifth source said, adding that Nornickel's Chinese output is likely to be consumed domestically. The new facility will have capacity to produce 450,000 tonnes of copper annually, two of the sources said, amounting to around 2% of global mined supplies estimated at around 22 million metric tons this year. Nornickel, which according to its annual report produced 425,400 tonnes of refined copper last year, processed all of its concentrates in 2023 at the Arctic plant, its only operation producing finished copper suitable for delivery to exchanges.