Latest news with #LianJyeSu


Emirates 24/7
13-02-2025
- Business
- Emirates 24/7
DeepSeek gives China's chipmakers leg up in race for cheaper AI
The rise of DeepSeek's artificial intelligence (AI) models is seen providing some Chinese chipmakers such as Huawei a better chance to compete in the domestic market against more powerful U.S. processors. Huawei and its Chinese peers have for years struggled to match Nvidia in building top-end chips that could compete with the U.S. firm's products for training models, a process where data is fed to algorithms to help them learn to make accurate decisions. However, DeepSeek's models, which focus on "inference," or when an AI model produces conclusions, optimise computational efficiency rather than relying solely on raw processing power. That is one reason why the model is expected to partly close the gap between what Chinese-made AI processors and their more powerful U.S. counterparts can do, analysts say. Huawei, and other Chinese AI chipmakers such as Hygon, Tencent-backed EnFlame, Tsingmicro and Moore Threads have in recent weeks issued statements claiming products will support DeepSeek models, although few details have been released. Huawei declined to comment. Moore Threads, Hygon EnFlame and Tsingmicro did not respond to Reuters queries seeking further comment. Industry executives are now predicting that DeepSeek's open-source nature and its low fees could boost adoption of AI and the development of real-life applications for the technology, helping Chinese firms overcome U.S. export curbs on their most powerful chips. Even before DeepSeek made headlines this year, products such as Huawei's Ascend 910B were seen by customers such as ByteDance as better suited for less computationally intensive "inference" tasks, the stage after training that involves trained AI models making predictions or performing tasks, such as through chatbots. In China, dozens of companies from automakers to telecoms providers have announced plans to integrate DeepSeek's models with their products and operations. "This development is very much aligned with the capability of Chinese AI chipset vendors," said Lian Jye Su, a chief analyst at tech research firm Omdia. "Chinese AI chipsets struggle to compete with Nvidia's GPU (graphics processing unit) in AI training, but AI inference workloads are much more forgiving and require a lot more local and industry-specific understanding," he said. NVIDIA STILL DOMINATES However, Bernstein analyst Lin Qingyuan said while Chinese AI chips were cost-competitive for inferencing, this was limited to the Chinese market as Nvidia chips were still better even for inference tasks. While U.S. export restrictions ban Nvidia's most advanced AI training chips from entering China, the company is still allowed to sell less powerful training chips that Chinese customers can use for inference tasks. Nvidia published a blog post on Thursday about how inference time was rising as a new scaling law and argued that its chips will be necessary to make DeepSeek and other "reasoning" models more useful. In addition to computing power, Nvidia's CUDA, a parallel computing platform that allows software developers to use Nvidia GPUs for general-purpose computing, not just AI or graphics, has become a crucial component of its dominance. Previously, many Chinese AI chip companies did not directly challenge Nvidia by asking users to abandon CUDA but instead, claimed their chips were compatible with CUDA. Huawei has been the most aggressive in its efforts to break away from Nvidia by offering a CUDA equivalent called Compute Architecture for Neural Networks (CANN), but experts said it faced obstacles in persuading developers to abandon CUDA. "Software performance of Chinese AI chip firms is also lacking at this stage. CUDA has a rich library and a diverse range of software capability, which requires significant long-term investment," said Omdia's Su. Follow Emirates 24|7 on Google News.


Zawya
13-02-2025
- Business
- Zawya
DeepSeek gives China's chipmakers leg up in race for cheaper AI
BEIJING: The rise of DeepSeek's artificial intelligence (AI) models is seen providing some Chinese chipmakers such as Huawei a better chance to compete in the domestic market against more powerful U.S. processors. Huawei and its Chinese peers have for years struggled to match Nvidia in building top-end chips that could compete with the U.S. firm's products for training models, a process where data is fed to algorithms to help them learn to make accurate decisions. However, DeepSeek's models, which focus on "inference," or when an AI model produces conclusions, optimise computational efficiency rather than relying solely on raw processing power. That is one reason why the model is expected to partly close the gap between what Chinese-made AI processors and their more powerful U.S. counterparts can do, analysts say. Huawei, and other Chinese AI chipmakers such as Hygon, Tencent-backed EnFlame, Tsingmicro and Moore Threads have in recent weeks issued statements claiming products will support DeepSeek models, although few details have been released. Huawei declined to comment. Moore Threads, Hygon EnFlame and Tsingmicro did not respond to Reuters queries seeking further comment. Industry executives are now predicting that DeepSeek's open-source nature and its low fees could boost adoption of AI and the development of real-life applications for the technology, helping Chinese firms overcome U.S. export curbs on their most powerful chips. Even before DeepSeek made headlines this year, products such as Huawei's Ascend 910B were seen by customers such as ByteDance as better suited for less computationally intensive "inference" tasks, the stage after training that involves trained AI models making predictions or performing tasks, such as through chatbots. In China, dozens of companies from automakers to telecoms providers have announced plans to integrate DeepSeek's models with their products and operations. "This development is very much aligned with the capability of Chinese AI chipset vendors," said Lian Jye Su, a chief analyst at tech research firm Omdia. "Chinese AI chipsets struggle to compete with Nvidia's GPU (graphics processing unit) in AI training, but AI inference workloads are much more forgiving and require a lot more local and industry-specific understanding," he said. NVIDIA STILL DOMINATES However, Bernstein analyst Lin Qingyuan said while Chinese AI chips were cost-competitive for inferencing, this was limited to the Chinese market as Nvidia chips were still better even for inference tasks. While U.S. export restrictions ban Nvidia's most advanced AI training chips from entering China, the company is still allowed to sell less powerful training chips that Chinese customers can use for inference tasks. Nvidia published a blog post on Thursday about how inference time was rising as a new scaling law and argued that its chips will be necessary to make DeepSeek and other "reasoning" models more useful. In addition to computing power, Nvidia's CUDA, a parallel computing platform that allows software developers to use Nvidia GPUs for general-purpose computing, not just AI or graphics, has become a crucial component of its dominance. Previously, many Chinese AI chip companies did not directly challenge Nvidia by asking users to abandon CUDA but instead, claimed their chips were compatible with CUDA. Huawei has been the most aggressive in its efforts to break away from Nvidia by offering a CUDA equivalent called Compute Architecture for Neural Networks (CANN), but experts said it faced obstacles in persuading developers to abandon CUDA. "Software performance of Chinese AI chip firms is also lacking at this stage. CUDA has a rich library and a diverse range of software capability, which requires significant long-term investment," said Omdia's Su. (Reporting by Liam Mo and Brenda Goh; Editing by Sam Holmes)
Yahoo
13-02-2025
- Business
- Yahoo
DeepSeek gives China's chipmakers leg up in race for cheaper AI
BEIJING (Reuters) - The rise of DeepSeek's artificial intelligence (AI) models is seen providing some Chinese chipmakers such as Huawei a better chance to compete in the domestic market against more powerful U.S. processors. Huawei and its Chinese peers have for years struggled to match Nvidia in building top-end chips that could compete with the U.S. firm's products for training models, a process where data is fed to algorithms to help them learn to make accurate decisions. However, DeepSeek's models, which focus on "inference," or when an AI model produces conclusions, optimise computational efficiency rather than relying solely on raw processing power. See for yourself — The Yodel is the go-to source for daily news, entertainment and feel-good stories. By signing up, you agree to our Terms and Privacy Policy. That is one reason why the model is expected to partly close the gap between what Chinese-made AI processors and their more powerful U.S. counterparts can do, analysts say. Huawei, and other Chinese AI chipmakers such as Hygon, Tencent-backed EnFlame, Tsingmicro and Moore Threads have in recent weeks issued statements claiming products will support DeepSeek models, although few details have been released. Huawei declined to comment. Moore Threads, Hygon EnFlame and Tsingmicro did not respond to Reuters queries seeking further comment. Industry executives are now predicting that DeepSeek's open-source nature and its low fees could boost adoption of AI and the development of real-life applications for the technology, helping Chinese firms overcome U.S. export curbs on their most powerful chips. Even before DeepSeek made headlines this year, products such as Huawei's Ascend 910B were seen by customers such as ByteDance as better suited for less computationally intensive "inference" tasks, the stage after training that involves trained AI models making predictions or performing tasks, such as through chatbots. In China, dozens of companies from automakers to telecoms providers have announced plans to integrate DeepSeek's models with their products and operations. "This development is very much aligned with the capability of Chinese AI chipset vendors," said Lian Jye Su, a chief analyst at tech research firm Omdia. "Chinese AI chipsets struggle to compete with Nvidia's GPU (graphics processing unit) in AI training, but AI inference workloads are much more forgiving and require a lot more local and industry-specific understanding," he said. NVIDIA STILL DOMINATES However, Bernstein analyst Lin Qingyuan said while Chinese AI chips were cost-competitive for inferencing, this was limited to the Chinese market as Nvidia chips were still better even for inference tasks. While U.S. export restrictions ban Nvidia's most advanced AI training chips from entering China, the company is still allowed to sell less powerful training chips that Chinese customers can use for inference tasks. Nvidia published a blog post on Thursday about how inference time was rising as a new scaling law and argued that its chips will be necessary to make DeepSeek and other "reasoning" models more useful. In addition to computing power, Nvidia's CUDA, a parallel computing platform that allows software developers to use Nvidia GPUs for general-purpose computing, not just AI or graphics, has become a crucial component of its dominance. Previously, many Chinese AI chip companies did not directly challenge Nvidia by asking users to abandon CUDA but instead, claimed their chips were compatible with CUDA. Huawei has been the most aggressive in its efforts to break away from Nvidia by offering a CUDA equivalent called Compute Architecture for Neural Networks (CANN), but experts said it faced obstacles in persuading developers to abandon CUDA. "Software performance of Chinese AI chip firms is also lacking at this stage. CUDA has a rich library and a diverse range of software capability, which requires significant long-term investment," said Omdia's Su.


Egypt Independent
12-02-2025
- Business
- Egypt Independent
DeepSeek hasn't just disrupted OpenAI. Chinese tech giants are being upended too
Hong Kong CNN — DeepSeek's advances have roiled global stock markets and AI players. Now, its influence is spreading quickly at home, with some of China's biggest tech companies, many of which had been developing their own chatbots, racing to incorporate the open-source model into their own services. In early February telecoms giant Huawei said it would run DeepSeek on its own computing hardware composed of its Ascend computer processors, which are domestically produced. Some AI watchers have hailed this as a turning point, as it demonstrates that a high-performing model like DeepSeek no longer requires Nvidia's most powerful chips to operate. 'This partnership defies US sanctions by proving China can deliver globally competitive AI performance using domestically developed AI hardware and software stack, replacing Nvidia chips with Ascend chips,' analysts at Bernstein, an investment and research firm, wrote in a research note earlier this month. Starting in late 2022, the Biden Administration imposed several rounds of export controls on China in an effort to deprive the country of technology that Washington fears Beijing could use to make the next generation of weapons and AI systems. But the success of DeepSeek's latest R1 AI model, which is said to be trained at a fraction of the cost of established players like ChatGPT, challenged the assumption that cutting off access to advanced chips could successfully stymie China's progress. Such is DeepSeek's fame that leading Chinese AI chipmakers — including Moore Threads, Tencent-backed Enflame, Baidu's Kunlunxin and Hygon Information Technology — have also announced support for the firm, running the AI model on their own computing chips. Beyond chips Beyond chipmakers, the cloud arms of major Chinese technology companies have also rushed to incorporate DeepSeek's technology into their offerings. Alibaba, Tencent, Baidu, and TikTok parent Bytedance announced they had made Deepseek's service available through their cloud platforms. The country's three major telecom operators, along with electronics maker Lenovo and auto brand Geely, have also adopted DeepSeek into their cloud platforms and products. 'It serves as recognition for open-source large-model companies. If your model is good enough, AI giants will integrate it into their platforms,' said Lian Jye Su, chief analyst at Omdia, a technology research firm. These major Chinese tech firms, which have dominated the country's internet scene for more than a decade, began rolling out their own large language models over the past two years since the debut of ChatGPT in late 2022. Despite having competing products they have welcomed DeepSeek. Su said the primary reason for their adoption is to attract more people and businesses to use their platforms. 'In reality, making money directly from large models is difficult for AI giants — unless there are mature consumer and enterprise applications to support them,' he said. Their moves reflect a wider trend. American Big Tech — including Nvidia, Microsoft and Amazon — have similarly embraced DeepSeek. Topping the charts Deepseek, the Hangzhou-based startup founded in 2023, sent shock waves around the world last month when it launched its newest AI model. Its AI Assistant app quickly topped global download charts, surpassing ChatGPT in late January. Within 20 days of launch, its daily active users have exceeded 22 million, according to state media. Its success has catapulted its founder, Liang Wenfeng, into the ranks of national heroes. Despite the optimism, analysts caution that bottlenecks in China's AI chip development remain due to US export restrictions. 'Porting DeepSeek models to different chip architectures is a lot of low-level software work, and the fact they could do so quickly is amazing, but it doesn't solve the chip shortage problem,' said Linghao Bao, senior analyst at Trivium China, a research and advisory firm. The ramp-up of advanced chip production continues to be hampered by the absence of advanced chipmaking equipment, which the US has banned from being exported to China, he said. And outside China, DeepSeek's rapid rise has drawn scrutiny from various governments. Last week, Taiwan and Australia banned their government officials from using the Chinese AI service over data security risks. Some South Korean ministries and government departments have also issued bans. Last month, Italy imposed a blanket block on DeepSeek's app after the company failed to address privacy concerns raised by the authorities. More countries have since raised concerns over the firm's data practices.


CNN
11-02-2025
- Business
- CNN
DeepSeek hasn't just disrupted OpenAI. Chinese tech giants are being upended too
DeepSeek's advances have roiled global stock markets and AI players. Now, its influence is spreading quickly at home, with some of China's biggest tech companies, many of which had been developing their own chatbots, racing to incorporate the open-source model into their own services. In early February telecoms giant Huawei said it would run DeepSeek on its own computing hardware composed of its Ascend computer processors, which are domestically produced. Some AI watchers have hailed this as a turning point, as it demonstrates that a high-performing model like DeepSeek no longer requires Nvidia's most powerful chips to operate. 'This partnership defies US sanctions by proving China can deliver globally competitive AI performance using domestically developed AI hardware and software stack, replacing Nvidia chips with Ascend chips,' analysts at Bernstein, an investment and research firm, wrote in a research note earlier this month. Starting in late 2022, the Biden Administration imposed several rounds of export controls on China in an effort to deprive the country of technology that Washington fears Beijing could use to make the next generation of weapons and AI systems. But the success of DeepSeek's latest R1 AI model, which is said to be trained at a fraction of the cost of established players like ChatGPT, challenged the assumption that cutting off access to advanced chips could successfully stymie China's progress. Such is DeepSeek's fame that leading Chinese AI chipmakers — including Moore Threads, Tencent-backed Enflame, Baidu's Kunlunxin and Hygon Information Technology — have also announced support for the firm, running the AI model on their own computing chips. Beyond chipmakers, the cloud arms of major Chinese technology companies have also rushed to incorporate DeepSeek's technology into their offerings. Alibaba, Tencent, Baidu, and TikTok parent Bytedance announced they had made Deepseek's service available through their cloud platforms. The country's three major telecom operators, along with electronics maker Lenovo and auto brand Geely, have also adopted DeepSeek into their cloud platforms and products. 'It serves as recognition for open-source large-model companies. If your model is good enough, AI giants will integrate it into their platforms,' said Lian Jye Su, chief analyst at Omdia, a technology research firm. These major Chinese tech firms, which have dominated the country's internet scene for more than a decade, began rolling out their own large language models over the past two years since the debut of ChatGPT in late 2022. Despite having competing products they have welcomed DeepSeek. Su said the primary reason for their adoption is to attract more people and businesses to use their platforms. 'In reality, making money directly from large models is difficult for AI giants — unless there are mature consumer and enterprise applications to support them,' he said. Their moves reflect a wider trend. American Big Tech — including Nvidia, Microsoft and Amazon — have similarly embraced DeepSeek. Deepseek, the Hangzhou-based startup founded in 2023, sent shock waves around the world last month when it launched its newest AI model. Its AI Assistant app quickly topped global download charts, surpassing ChatGPT in late January. Within 20 days of launch, its daily active users have exceeded 22 million, according to state media. Its success has catapulted its founder, Liang Wenfeng, into the ranks of national heroes. Despite the optimism, analysts caution that bottlenecks in China's AI chip development remain due to US export restrictions. 'Porting DeepSeek models to different chip architectures is a lot of low-level software work, and the fact they could do so quickly is amazing, but it doesn't solve the chip shortage problem,' said Linghao Bao, senior analyst at Trivium China, a research and advisory firm. The ramp-up of advanced chip production continues to be hampered by the absence of advanced chipmaking equipment, which the US has banned from being exported to China, he said. And outside China, DeepSeek's rapid rise has drawn scrutiny from various governments. Last week, Taiwan and Australia banned their government officials from using the Chinese AI service over data security risks. Some South Korean ministries and government departments have also issued bans. Last month, Italy imposed a blanket block on DeepSeek's app after the company failed to address privacy concerns raised by the authorities. More countries have since raised concerns over the firm's data practices.