Nexus Search
Search public articles and published connected research.
Use this focused workspace without learning the underlying database or content model.15published records
Explore the archive
Find the research behind your question
Top matched records in the current language. Counts describe this result set, not the entire archive.
Article-body matches use indexed words. Titles and summaries also support partial-word matches.
Asset
Open →Ascend 960DT
Automatically indexed from an explicit mention in article #43220. Open linked research for context.
EntityArticle
Open →Huawei Turns UnifiedBus Into China’s Nvidia-Bypass AI Factory Stack
Huawei says its Ascend 960DT will be ready in Q1 2027, with Ascend 960PR following in Q3 2027, while its UnifiedBus and Peerium architecture aim to scale AI systems from SuperPoDs to very large SuperClusters. Reuters also reports Huawei says domestic AI-compu…
AI Industrial Flywheel (⚙️🤖🔥)Article
Open →DeepSeek Turns Huawei Ascend Tools Into a CUDA-Bypass Stack
DeepSeek said on September 30, 2026 that it partnered with Huawei to build programming infrastructure for Huawei Ascend chips, including open-source compute and communication libraries. The release includes DeepGEMM-Ascend, DeepEP-Ascend, and Ascend support a…
AI Industrial Flywheel (⚙️🤖🔥)Asset
Open →Huawei Ascend
Automatically indexed from an explicit mention in article #45221. Open linked research for context.
EntityAsset
Open →Ascend 950
Automatically indexed from an explicit mention in article #45221. Open linked research for context.
EntityAsset
Open →DeepGEMM-Ascend
Automatically indexed from an explicit mention in article #45221. Open linked research for context.
EntityAsset
Open →DeepEP-Ascend
Automatically indexed from an explicit mention in article #45221. Open linked research for context.
EntitySource
Open →DeepSeek and Huawei release open-source Ascend AI programming tools to reduce reliance on Nvidia CUDA ecosystem - Tom's Hardware
www.tomshardware.com
Evidence registrySource
Open →DeepGEMM-Ascend: clean and efficient matrix multiplication kernel library for Huawei Ascend NPUs - DeepSeek AI GitHub
github.com
Evidence registrySource
Open →DeepEP-Ascend: A high-performance communication library for machine learning training and inference on Huawei Ascend NPUs - DeepSeek AI GitHub
github.com
Evidence registry