Huawei Ascend
PulseAugur coverage of Huawei Ascend — every cluster mentioning Huawei Ascend across labs, papers, and developer communities, ranked by signal.
4 天有情绪数据
-
DeepSeek V4 adapts to Huawei Ascend chips, boosting China's AI stack
DeepSeek V4 has been fully adapted to run on Huawei's Ascend AI chips, a significant development for China's domestic AI infrastructure. This adaptation aims to reduce the country's dependence on foreign chip manufactur…
-
Guangdong plans to boost service trade, develop AI and semiconductor industries
Guangdong province in China has released a three-year plan to boost its service trade, aiming for over $310 billion in total trade volume by 2028 with an annual growth rate exceeding 10%. The plan emphasizes strengtheni…
-
Hongxin Electronics and Huawei launch China's first 'Token Factory'
Hongxin Electronics has partnered with Huawei's Ascend to establish China's first 'Token Factory' in Wuxi. This initiative is supported by a significant 16.45 billion CNY government project, aiming to leverage Huawei's …
-
DeepSeek-V4 launches with 1M context, Chinese hardware optimization
DeepSeek has officially released its latest flagship model, DeepSeek-V4, featuring a 1 million token context window and enhanced agent capabilities. The model comes in two versions, Pro and Flash, with the Pro version s…
-
Chinese AI industry focuses on domestic chips and LLMs
Dongwu Securities reports that the AI industry is undergoing a significant shift driven by intelligent agents and a push for domestic technological self-sufficiency. The training of DeepSeek V4 using domestic computing …
-
New MoE inference design uses pooled HBM to cut communication latency on Ascend
Researchers have developed a new communication design for Mixture-of-Experts (MoE) inference on Ascend systems, aiming to reduce bottlenecks in token exchange. This approach eliminates intermediate relay and reordering …
-
US and China diverge in AI race as tech ecosystems grow apart
Chinese AI development is increasingly diverging from the US, driven by a national policy of technological self-reliance. This trend was highlighted by the launch of DeepSeek's V4 AI models, which are specifically optim…
-
Huawei and USTC launch AI scientist platform with coordination engineering
Huawei and the University of Science and Technology of China have launched "Lingjing Zaowu," an intelligent research cloud platform designed to accelerate scientific discovery. This platform leverages a fully domestic s…
-
Triton language now runs efficiently on Huawei Ascend NPUs
A new compilation framework, Triton-Ascend 3.2.0, has been released to enable the Triton programming language to run efficiently on Huawei's Ascend hardware. This framework simplifies operator development by automating …
-
DeepSeek V4 First Release Adaptation Behind: Why does Ascend insist on not doing a CUDA compatibility layer?
Huawei's Ascend AI accelerators are forging a unique path by eschewing CUDA compatibility to build an independent ecosystem. This strategy focuses on deep architectural changes in their latest Ascend 950 chips to addres…
-
DeepSeek V4 release sparks surge in Chinese semiconductor stocks, boosting domestic AI computing power
DeepSeek V4's release has significantly boosted China's A-share semiconductor market, with sectors like GPU and semiconductor equipment experiencing a surge. This rally is attributed to V4's compatibility with Huawei's …
-
Meituan tests trillion-parameter AI model built on domestic compute
Meituan has reportedly initiated a private test of a trillion-parameter AI model, developed using only Chinese computing infrastructure. This model is said to rival GPT-4's performance and was likely trained using Huawe…
-
New specification language aims to formalize ML kernel contracts across silicon
Researchers have introduced a new specification language called Kernel Contracts, designed to formally define and verify the correctness of machine learning kernels across different hardware platforms. This language add…
-
DeepSeek V4 challenges US AI dominance with efficiency and Huawei integration
The AI model DeepSeek V4 has been released, showcasing significant advancements in performance and efficiency. It reportedly uses 9.5 times less memory than previous iterations, offering substantial cost reductions for …
-
DeepSeek V4 AI model undercuts GPT-5.5 on price, rivals performance
China's DeepSeek has released its V4 AI model, significantly undercutting competitors like OpenAI's GPT-5.5 in price. The V4 Pro model offers substantial discounts, with input costs reduced to a fraction of previous lev…
-
US warns global allies of Chinese AI firms stealing intellectual property
The U.S. State Department has issued a diplomatic cable warning global partners about Chinese companies, including DeepSeek, allegedly stealing intellectual property from U.S. AI labs. The cable highlights concerns over…
-
DeepSeek extends V4-Pro API discount, offers competitive performance at lower cost
DeepSeek has extended the promotional discount for its V4-Pro API until May 31, 2026. The V4-Pro model, featuring 1.6 trillion parameters and supporting a 1 million token context window, is optimized for Huawei Ascend A…
-
DeepSeek releases V4 Pro and Flash models with 1M context, runs on Huawei chips
DeepSeek has released its new V4 family of models, including V4 Pro and V4 Flash, which boast a 1 million token context window. These models were trained on 32 trillion tokens and feature a novel hybrid attention system…
-
DeepSeek V4 models offer high performance with reduced inference costs and NPU support
DeepSeek has released its V4 family of open-weight large language models, featuring a 1.6 trillion parameter model and a smaller 284 billion parameter Flash MoE model. These new models claim to rival top proprietary LLM…
-
DeepSeek previews new AI model that ‘closes the gap’ with frontier models
DeepSeek has released its V4 AI model, featuring two versions: V4-Pro and V4-Flash. These models boast a 1 million token context window and utilize a mixture-of-experts architecture for efficiency. While DeepSeek V4 aim…