News from the AI & ML world
DeeperML - #chinaai
|
Doug Black@insideAI News
//
NVIDIA is actively developing AI chips specifically designed to comply with U.S. export regulations for the Chinese market. This strategic move aims to allow NVIDIA to maintain its presence in China's significant AI market despite increasing restrictions. Concurrently, NVIDIA's CEO Jensen Huang has voiced support for Trump's tariff plan, describing it as "utterly visionary," signaling a complex navigation of both technological and political landscapes. These developments highlight NVIDIA's determination to balance its business interests with evolving geopolitical dynamics.
The new chips, reportedly based on the RTX Pro 6000-series, will have significantly reduced specifications to meet export control requirements. This includes forgoing advanced technologies like Taiwan Semiconductor’s CoWoS packaging and using standard GDDR7 memory instead of high-bandwidth memory. While the specifics of the chips, potentially named RTX Pro 6000D, are still emerging, these adjustments are essential for NVIDIA to continue offering competitive AI solutions in China, where a substantial number of AI developers are located. Challenges persist, as the company previously absorbed a $4.5 billion hit due to export restrictions, leading to a write-down on Chinese inventory and commitments. The emergence of strong domestic competitors, particularly Huawei, intensifies the pressure on NVIDIA. Huawei's Ascend 910C and 910B processors have gained traction among major Chinese tech firms, and their CloudMatrix 384 rack system directly rivals NVIDIA's Blackwell GB200 NVL72 configuration. Despite these obstacles, NVIDIA remains committed to the Chinese market, viewing it as crucial for maintaining its global leadership in AI technology. Recommended read:
References :
@felloai.com
//
Alibaba has launched Qwen3, a new generation of AI models designed to compete with Silicon Valley's leading AI technologies. Qwen3 represents a significant advancement in AI, with capabilities that directly challenge models from OpenAI, Google, and Meta. The unveiling has generated excitement and apprehension across the global tech industry, signaling a narrowing of the gap between China and the US in AI development.
The Qwen3 family includes models with varying parameter sizes, catering to different applications and offering exceptional capabilities in complex reasoning, mathematical problem-solving, and code generation. The models support 119 languages and have been trained on a massive dataset of over 36 trillion tokens, providing a broad understanding of global information. One key innovation is the "hybrid reasoning" approach, allowing the models to switch between "fast thinking" for quick responses and "slow thinking" for more analytical tasks. Benchmark results have begun to surface regarding the performance of the Qwen3 models. According to Alibaba, the Qwen3 models can match and, in some cases, outperform the best models available from Google and OpenAI. Some Qwen3 models are being released open-source, a move intended to boost China's AI ecosystem by encouraging wider adoption and collaborative development. Recommended read:
References :
@felloai.com
//
Alibaba has launched Qwen3, a new family of large language models (LLMs), posing a significant challenge to Silicon Valley's AI dominance. Qwen3 is not just an incremental update but a leap forward, demonstrating capabilities that rival leading models from OpenAI, Google, and Meta. This advancement signals China’s growing prowess in AI and its potential to redefine the global tech landscape. Qwen3's strengths lie in reasoning, coding, and multilingual understanding, marking a pivotal moment in China's AI development.
The Qwen3 family includes models of varying sizes to cater to diverse applications. Key features include complex reasoning, mathematical problem-solving, and code generation. The models support 119 languages and are trained on a massive dataset of over 36 trillion tokens. Another innovation is Qwen3’s “hybrid reasoning” approach, enabling models to switch between "fast thinking" for quick responses and "slow thinking" for deeper analysis, enhancing versatility and efficiency. Alibaba has also emphasized the open-source nature of some Qwen3 models, fostering wider adoption and collaborative development in China's AI ecosystem. Alibaba also introduced ZeroSearch, a method that uses reinforcement learning and simulated documents to teach LLMs retrieval without real-time search. It addresses the challenge of LLMs relying on static datasets, which can become outdated. By training the models to retrieve and incorporate external information, ZeroSearch aims to improve the reliability of LLMs in real-world applications like news, research, and product reviews. This method mitigates the high costs associated with large-scale interactions with live APIs, making it more accessible for academic research and commercial deployment. Recommended read:
References :
ashilov@gmail.com (Anton@tomshardware.com
//
References:
www.tomshardware.com
, blogs.nvidia.com
Nvidia's CEO, Jensen Huang, has stated that China is rapidly catching up to the U.S. in artificial intelligence capabilities. Huang emphasized that China isn't far behind, particularly in AI hardware development, where companies like Huawei are making significant strides. Huawei's advancements, including its Ascend 900-series AI accelerators and CloudMatrix 384 systems, demonstrate China's growing competitiveness. The CloudMatrix 384, featuring 384 dual-chiplet HiSilicon Ascend 910C interconnected via an optical mesh network, offers impressive computing power and memory bandwidth, rivaling Nvidia's offerings, though with lower efficiency. Huang acknowledged Huawei as a formidable technology company with incredible computing, networking, and software capabilities essential for advancing AI.
New tools are emerging to empower artists to harness the power of AI in image generation. NVIDIA has introduced a 3D Guided Generative AI Blueprint that provides a workflow enabling artists to precisely control object placement and camera angles using a 3D scene in Blender. This tackles the common challenge of achieving the desired composition and layout in AI-generated images. This AI Blueprint is pre-optimized for NVIDIA and GeForce RTX GPUs, built on NVIDIA NIM microservices to maximize AI model performance. The process involves converting a 3D viewport to a depth map, which guides the image generator (FLUX.1-dev) along with a user prompt. For those looking to enter the AI job market, NVIDIA experts share several key tips. A diverse educational and professional background can be a valuable asset, enabling adaptability in the rapidly evolving AI field. Integrating AI into daily workflows, regardless of one's background, can help individuals stand out. It's also crucial to identify your passions within AI and gain experience in relevant domains such as autonomous vehicles, robotics, gaming, or healthcare. By aligning skills with specific AI applications, candidates can better position themselves for success. Recommended read:
References :
@blogs.nvidia.com
//
Nvidia is currently facing pressure from the U.S. government regarding AI GPU export rules. CEO Jensen Huang has been advocating for the Trump administration to relax these restrictions, arguing they hinder American companies' ability to compete in the global market. Huang stated at the Hill and Valley Forum that China is not far behind the U.S. in AI capabilities, emphasizing the need to accelerate the diffusion of American AI technology worldwide. He also acknowledged Huawei's progress in computing, networking, and software, noting their development of the CloudMatrix 384 system. This system, powered by Ascend 910C accelerators, is considered competitive with Nvidia's GB200 NVL72, signaling the emergence of domestic alternatives in China.
Despite Nvidia's pleas, the Trump administration is considering tighter controls on AI GPU exports. The administration plans to use chip access as leverage in trade negotiations with other nations. This approach contrasts with Nvidia's view that restricting exports will only fuel the development of competing hardware and software in countries like China. According to the AI Diffusion framework, access to advanced AI chips like Nvidia’s H100 is only unrestricted for companies based in the U.S. and "Tier 1" nations, while those in "Tier 2" nations face annual limits and "Tier 3" countries are effectively barred. Adding to the complexity, Nvidia is also engaged in a public dispute with AI startup Anthropic over the export restrictions. Anthropic has endorsed the Biden-era "AI Diffusion Rule" and has claimed there has been chip smuggling to China. An Nvidia spokesperson dismissed Anthropic's claims about chip smuggling tactics as "tall tales," arguing that American firms should focus on innovation instead of trying to manipulate policy for competitive advantage. As the May 15th export controls deadline approaches, the tensions continue to rise within the AI industry over the balance between national security, economic prosperity, and global competitiveness. Recommended read:
References :
Harsh Sharma@TechDator
//
Huawei is intensifying its challenge to Nvidia in the Chinese AI market by preparing to ship its Ascend 910C AI chips in large volumes. This move comes at a crucial time as Chinese tech firms are actively seeking domestic alternatives to Nvidia's H20 chip, which is now subject to U.S. export restrictions. The Ascend 910C aims to bolster China's tech independence, providing a homegrown solution amidst limited access to foreign chips. The chip combines two 910B processors into one package, utilizing advanced integration to rival the performance of Nvidia’s H100.
Huawei's strategy involves a multi-pronged approach. Late last year, the company sent Ascend 910C samples to Chinese tech firms and began taking early orders. Deliveries have already started, signaling Huawei's readiness to scale up production. While the 910C may not surpass Nvidia's newer B200, it is designed to meet the needs of Chinese developers who are restricted from accessing foreign options. The production of the Ascend 910C involves a complex supply chain, with parts crafted by China's Semiconductor Manufacturing International Corporation (SMIC) using its N+2 7nm process. Despite the challenges from Huawei, Nvidia remains committed to the Chinese market. Nvidia is reportedly collaborating with DeepSeek, a local AI leader, to develop chips within China using domestic factories and materials. This plan includes establishing research teams in China and utilizing SMIC, along with local memory makers and packaging partners, to produce China-specific chips. CEO Jensen Huang has affirmed that Nvidia will continue to make significant efforts to optimize its products to comply with regulations and serve Chinese companies, even amidst ongoing trade tensions and tariffs. Recommended read:
References :
Jaime Hampton@AIwire
//
China's multi-billion-dollar AI infrastructure boom is now facing a significant downturn, according to a new report. The rush to build AI datacenters, fueled by the rise of generative AI and encouraged by government incentives, has resulted in billions of dollars in idle infrastructure. Many newly built facilities are now sitting empty, with some reports indicating that up to 80% of China’s new computing resources remain unused.
The "DeepSeek Effect" is a major factor in this reversal. DeepSeek's AI models, particularly the Deepseek v3, have demonstrated impressive efficiency in training, reducing the demand for large-scale datacenter deployments. Smaller players are abandoning plans to pretrain large models because DeepSeek’s open-source models match ChatGPT-level performance at a fraction of the cost, leading to a collapse in demand for training infrastructure just as new facilities were ready to come online. Recommended read:
References :
Dashveenjit Kaur@AI News
//
Chinese AI startup DeepSeek is shaking up the global technology landscape with its latest large language model, DeepSeek-V3-0324. This new model has been lauded for matching the performance of American AI models, while boasting significantly lower development costs. According to Lee Kai-fu, CEO of Chinese startup 01.AI, the gap between Chinese and American AI capabilities has narrowed dramatically, with China even ahead in some specific areas.
DeepSeek-V3-0324 features enhanced reasoning capabilities and improved performance in multiple benchmarks, particularly in mathematics. The model scored 59.4 on the American Invitational Mathematics Examination (AIME), a significant improvement over its predecessor. Häme University lecturer Kuittinen Petri noted DeepSeek's achievements were realized with just a fraction of the resources available to competitors like OpenAI. This breakthrough has been attributed to DeepSeek’s focus on algorithmic efficiency and novel approaches to model architecture, allowing them to overcome restrictions on access to the latest silicon. This disruption is not going unnoticed, when DeepSeek launched its R1 model in January, America’s Nasdaq plunged 3.1%, while the S&P 500 fell 1.5%. While DeepSeek claimed a $5.6 million training cost, this represented only the marginal cost of the final training run. SemiAnalysis estimates DeepSeek's actual hardware investment at closer to $1.6 billion, with hundreds of millions in operating costs. The developments present opportunities and challenges for the. Recommended read:
References :
Dashveenjit Kaur@AI News
//
DeepSeek, a Chinese AI startup, is causing a stir in the AI industry with its new large language model, DeepSeek-V3-0324. Released with little fanfare on the Hugging Face AI repository, the 641-gigabyte model is freely available for commercial use under an MIT license. Early reports indicate it can run directly on consumer-grade hardware, such as Apple’s Mac Studio with the M3 Ultra chip, especially in a 4-bit quantized version that reduces the storage footprint to 352GB. This innovation challenges the previous notion that Silicon Valley held a chokehold on the AI industry.
China's focus on algorithmic efficiency over hardware superiority has allowed companies like DeepSeek to flourish despite restrictions on access to the latest silicon. DeepSeek's R1 model, launched earlier this year, already rivaled OpenAI's ChatGPT-4 at a fraction of the cost. Now the DeepSeek-V3-0324 features enhanced reasoning capabilities and improved performance. This has sparked a gold rush among Chinese tech startups, rewriting the playbook for AI development and allowing smaller companies to believe they have a shot in the market. Recommended read:
References :
Ryan Daws@AI News
//
DeepSeek, a Chinese AI company, has released DeepSeek V3-0324, an updated AI model that demonstrates impressive performance. The model is now running at 20 tokens per second on a Mac Studio. This model is said to contain 685 billion parameters and its cost-effectiveness challenges the dominance of American AI models, signaling that China continues to innovate in AI despite chip restrictions. Reports from early testers show improvements over previous versions and the model tops non-reasoning AI models in open-source first.
This new model runs on consumer-grade hardware, specifically Apple's Mac Studio with the M3 Ultra chip, diverging from the typical data center requirements for AI. It is freely available for commercial use under the MIT license. According to AI researcher Awni Hannun, the model runs at over 20 tokens per second on a 512GB M3 Ultra. The company has made no formal announcement, just an empty README file and the model weights themselves. This stands in contrast to the carefully orchestrated product launches by Western AI companies. Recommended read:
References :
Ryan Daws@AI News
//
DeepSeek, a Chinese AI startup, has emerged as a significant player in the artificial intelligence landscape, challenging the dominance of Western AI companies. Their release of the V3 large language model under the MIT open-source license marks a notable development, potentially shifting the global AI landscape. The DeepSeek-V3 model forms the foundation of DeepSeek-R1, showcasing innovation through Mixture of Experts (MoE) and efficient parameter activation system.
DeepSeek V3-0324 has achieved the position of highest-scoring non-reasoning model on the Artificial Analysis Intelligence Index. This open-source model outperforms proprietary counterparts like Google's Gemini 2.0 Pro and Meta's Llama 3.3 70B in real-time use cases. While DeepSeek models demonstrate strong performance, especially in mathematics and reasoning tasks, concerns have been raised regarding intellectual property, government connections, and security vulnerabilities. Recommended read:
References :
Ryan Daws@AI News
//
References:
venturebeat.com
, AI News
,
DeepSeek, a Chinese AI startup, is making waves in the artificial intelligence industry with its DeepSeek-V3 model. This model is demonstrating performance that rivals Western AI models like those from OpenAI and Anthropic, but at significantly lower development costs. The release of DeepSeek-V3 is seen as jumpstarting AI development across China, with other startups and established companies releasing their own advanced models, further fueling competition. This has narrowed the technology gap between China and the United States as China has adapted to and overcome international restrictions through creative approaches to AI development.
One particularly notable aspect of DeepSeek-V3 is its ability to run efficiently on consumer-grade hardware, such as the Mac Studio with an M3 Ultra chip. Reports indicate that the model achieves speeds of over 20 tokens per second on this platform, making it a potential "nightmare for OpenAI". This contrasts sharply with the data center requirements typically associated with state-of-the-art AI models. The company's focus on algorithmic efficiency has allowed them to achieve notable gains despite restricted access to the latest silicon, showcasing that Chinese AI innovation has flourished by focusing on algorithmic efficiency and novel approaches to model architecture. Recommended read:
References :
@tomshardware.com
//
References:
Jon Keegan
, www.tomshardware.com
,
Ant Group has announced a significant breakthrough in AI, achieving a 20% reduction in AI costs by training models on domestically produced Chinese chips. According to reports, the company utilized chips from Chinese tech giants Alibaba and Huawei, reaching performance levels comparable to those obtained with Nvidia's H800 chips. The AI models, named Ling-Plus and Ling-Lite, are said to match or even outperform leading models, with Ant Group claiming its AI models outperformed Meta’s in benchmarks and cut inference costs.
This accomplishment signals a potential leap forward in China's AI development efforts and a move towards self-reliance in semiconductor technology. While Ant Group still uses Nvidia hardware for some tasks, it is now relying more on alternatives, including chips from AMD and Chinese manufacturers, driven in part by U.S. sanctions that limit access to Nvidia's advanced GPUs. This shift could lessen the country’s dependence on foreign technology. Recommended read:
References :
Ryan Daws@AI News
//
DeepSeek V3-0324, the latest large language model from Chinese AI startup DeepSeek, is making waves in the artificial intelligence industry. The model, quietly released with an MIT license for commercial use, has quickly become the highest-scoring non-reasoning model on the Artificial Analysis Intelligence Index. This marks a significant milestone for open-source AI, surpassing proprietary counterparts like Google’s Gemini 2.0 Pro, Anthropic’s Claude 3.7 Sonnet, and Meta’s Llama 3.3 70B.
DeepSeek V3-0324's efficiency is particularly notable. Early reports indicate that it can run directly on consumer-grade hardware, specifically Apple’s Mac Studio with an M3 Ultra chip, achieving speeds of over 20 tokens per second. This capability is a major departure from the typical data center requirements associated with state-of-the-art AI. The updated version demonstrates substantial improvements in reasoning and benchmark performance, as well as enhanced Chinese writing proficiency and optimized translation quality. Recommended read:
References :
Ryan Daws@AI News
//
DeepSeek V3-0324 has emerged as a leading AI model, topping benchmarks for non-reasoning AI in an open-source breakthrough. This milestone signifies a significant advancement in the field, as it marks the first time an open weights model has achieved the top position among non-reasoning models. The model's performance surpasses proprietary counterparts and edges it closer to proprietary reasoning models, highlighting the growing viability of open-source solutions for latency-sensitive applications. DeepSeek V3-0324 represents a new era for open-source AI, offering a powerful and adaptable tool for developers and enterprises.
DeepSeek-V3 now runs at 20 tokens per second on Apple’s Mac Studio, presenting a challenge to OpenAI’s cloud-dependent business model. The 685-billion-parameter model, DeepSeek-V3-0324, is freely available for commercial use under the MIT license. This achievement, coupled with its cost efficiency and performance, signals a shift in the AI sector, where open-source frameworks increasingly compete with closed systems. Early testers report significant improvements over previous versions, positioning DeepSeek's new model above Claude Sonnet 3.5 from Anthropic. Recommended read:
References :
|
BenchmarksBlogsResearch Tools |