Huawei Ascend
PulseAugur coverage of Huawei Ascend — every cluster mentioning Huawei Ascend across labs, papers, and developer communities, ranked by signal.
- developed by Huawei 100%
- used by Meituan 90%
- used by LongCat-2.0 90%
- partners with University of Science and Technology of China 90%
- instance of Ascend 950 90%
- used by Cann 90%
- developed by Cann 70%
- affiliated with Moore Threads 70%
- used by Shenzhen 70%
- used by Tsinghua University 70%
- instance of Xian Y-20 70%
- used by University of Science and Technology of China 70%
- 2026-06-24 product_launch Huawei will announce new Ascend AI chip solutions at the 2026 World Artificial Intelligence Conference. source
15 day(s) with sentiment data
-
New proxy models slash LLM debugging costs by up to 33x
Researchers have developed a method to create "proxy models" for large language models (LLMs) that significantly reduce the computational cost and time required for debugging and reproducing failures during reinforcemen…
-
openJiuwen launches enterprise AI agent swarm architecture, lands first financial deployment
openJiuwen has released an enterprise-grade distributed swarm architecture for its AI Agent platform, JiuwenSwarm. This new architecture addresses key challenges in scaling AI agents for enterprise use, including managi…
-
Mingxin FX100 storage solution accelerates video inference, reducing latency
Mingxin's FX100 storage solution addresses latency bottlenecks in real-time video inference, which are often caused by storage and data path limitations rather than GPU compute. The system employs a tiered KV cache appr…
-
Huawei chip scientist details strategy for AI processor self-reliance
Liao Heng, chief scientist of Huawei's semiconductor business and architect of its Ascend AI processors, shared insights into the company's strategy for achieving technological self-reliance amidst US sanctions. He disc…
-
China's AI Chip Ambitions Challenge NVIDIA with Huawei Ascend
China has made significant advancements in AI, particularly with the development of its own AI chips. The Huawei Ascend series is highlighted as a key competitor to NVIDIA's offerings, signaling a growing technological …
-
Huawei open-sources openPangu-2.0-Pro MoE model with 505B parameters
Huawei has open-sourced its new Mixture-of-Experts (MoE) model, openPangu-2.0-Pro. This model boasts a total of 505 billion parameters with 18 billion activated, and supports a context length of 512,000 tokens. It was t…
-
New benchmarks and optimization tools emerge for LLM-generated hardware kernels
Two new research papers introduce benchmarks and optimization frameworks for large language models (LLMs) generating code for hardware accelerators. The first paper, KernelGenBench, offers a unified benchmark to evaluat…
-
Clos Network Architecture: Cost and Selection Framework for AI Inference Clusters
The Clos (or Fat-Tree) network architecture is a popular choice for large-scale AI inference clusters due to its scalability and high bandwidth. This article analyzes the cost components of Clos networks, including swit…
-
China pushes domestic AI chip adoption amid US import bans · 5 sources tracked
China is accelerating its domestic AI chip development, aiming for self-sufficiency to overcome US import restrictions. Vice Premier Ding Xuexiang has reportedly urged companies to adopt Chinese AI chips, even labeling …
-
Huawei Ascend chips show efficiency gains with DeepSeek V4, but questions remain
Huawei has released benchmarks for its Ascend chips demonstrating efficiency gains in the post-training phase of the DeepSeek V4 model. These benchmarks indicate that the Huawei 910C chips achieve approximately 60% of t…
-
Meituan trains 1.6T coding model on Chinese ASICs, bypassing Nvidia
Meituan has released LongCat-2.0, a 1.6-trillion-parameter Mixture-of-Experts model designed for agentic coding tasks. Notably, the model was trained entirely on domestic Chinese AI ASICs, avoiding Nvidia hardware. Long…
-
Chinese AI lab Z.ai launches 1GW data center using domestic chips, bypassing Nvidia
Chinese AI developer Z.ai has reportedly completed and partially activated a 1-gigawatt data center utilizing exclusively domestic chips, aiming to train its GLM model family. This facility is among the largest AI data …
-
Alibaba open-sources AI software stack to challenge Nvidia's CUDA
Alibaba's chip design unit, T-Head, is open-sourcing its proprietary software stack to challenge Nvidia's CUDA ecosystem. This move aims to simplify the adoption of Alibaba's Zhenwu AI computing architectures for develo…
-
China's AI Chips and Infrastructure Reach Tipping Point at WAIC 2026
The World Artificial Intelligence Conference (WAIC) 2026 in Shanghai is highlighting China's advancements in AI hardware and infrastructure. The event features over 100 Chinese chip companies, including Huawei Ascend an…
-
Baidu AI Cloud leads China's financial large model market, IDC reports
Baidu AI Cloud has secured the top position in China's financial large model market, holding a 16.6% market share according to a report by International Data Corporation (IDC). The overall Chinese financial generative A…
-
Google Gemini 3.5 Pro launches July 17; Chinese LLMs surge ahead · 1 source tracked
Google is set to release Gemini 3.5 Pro on July 17, featuring a 2 million token context window, aiming to compete with GPT-5.6. Concurrently, Chinese AI companies are making significant advancements; Zhipu AI has open-s…
-
Huawei's Atlas 950 SuperPoD and first AI agent phone to debut at China AI summit
Huawei is set to unveil its Atlas 950 SuperPoD computing cluster at the upcoming World Artificial Intelligence Conference (WAIC) in Shanghai. This cluster, previously revealed last year, is designed for AI and machine l…
-
Meituan open-sources LongCat-2.0 models; Solstice Advanced Materials in merger talks
Solstice Advanced Materials is reportedly in merger negotiations with Element Solutions, with a potential deal possibly concluding this week. Separately, Meituan has open-sourced its LongCat-2.0 models, with domestic ch…
-
New Causal ASCEND framework accelerates multi-omics data discovery
Researchers have developed Causal ASCEND, a novel framework for causal discovery in high-dimensional multi-omics data. This method leverages the inherent two-tiered structure of biological systems to infer ancestral rel…
-
China's AI chip market faces price surge and grassroots boom in Shenzhen
Shenzhen's Huaqiangbei market is experiencing a dual AI chip situation: a domestic price surge due to speculative hoarding, and a grassroots AI revolution fueled by local chipmakers. While global attention is on high-en…