Pangu Big Model 5.0 Released, Upgrading Multimodal Capabilities to Break Industry Challenges

At the Huawei Developer Conference 2024 held in Dongguan, Guangdong Province on June 21, Zhang Pingan, Managing Director of Huawei Technologies Co. and CEO of Huawei Cloud, released a new upgrade in the full range, multi-modal and strong thinking in three areasPangu LM5.0.
Zhang Pingan said, Pangu 5.0 launched to adapt to different business scenarios multiple specifications model, and combined with the physical world, accelerate the big model industry landing.
在全系列方面，盘古大模型5.0包含不同参数规格的模型，以适配不同的业务场景。十亿级参数的Pangu E系列可支撑手机、PC等端侧的smart (phone, system, bomb etc)应用；百亿级参数的Pangu P系列，适用于低时延、低成本的推理场景；千亿级参数的Pangu U系列适用于处理复杂任务，可以成为企业通用大模型的底座；万亿级参数的Pangu S系列超级大模型是处理跨领域多任务的超级大模型，能够帮助企业更好地在全场景应用AI技术。
In terms of multimodality, Pangu Big Model 5.0 is able to understand the physical world better and more accurately, including text, pictures, videos, radar, infrared, remote sensing and many more modalities. In terms of generation, Pangu 5.0 can generate multimodal content that conforms to the laws of the physical world, allowing innovation at will.
Complex logical reasoning is the key to making Big Model an industry assistant. In terms of strong thinking, Pangu Big Model 5.0 combines the deep combination of thought chain technology and strategy search technology, which greatly improves mathematical ability and complex task planning ability.
According to Pingan Zhang, Huawei Cloud Pangu Big Model focuses on the industry, and continues to overcome difficulties and forge ahead on the path of solving problems and doing difficult things to reshape thousands of industries. In the past year, Pangu Big Model has landed in more than 30 industries and 400 scenarios, playing a great value in government affairs, finance, manufacturing, pharmaceutical R&D, coal mining, steel, railroads, automated driving, industrial design, architectural design, meteorology and other fields.