Core Insights - Fourth Paradigm launched the "ModelHub XC" platform, along with a community and value-added services for model adaptation, aiming to address compatibility issues between domestic AI models and chips [1] - The introduction of EngineX, an AI engine system specifically designed for domestic computing power, facilitates batch model support and reduces deployment time, promoting a "plug-and-play" model approach [1] - Fourth Paradigm's value-added services, supported by hundreds of engineers, ensure model adaptation to specified domestic computing power, with over a hundred certified models currently and a target of reaching hundreds of thousands within a year [1] Industry Context - Compatibility issues between models and chips in the domestic computing ecosystem often lead to prolonged deployment cycles and increased redundancy in work [1] - The launch of "ModelHub XC" aims to bridge the gap between computing power and models, potentially providing critical support for the large-scale implementation of AI in the industry [1]
第四范式携手昆仑芯、摩尔线程等头部芯片厂商，推进国产芯片与模型适配难的问题解决