SenseCore SenseTime AI Infrastructure is an integrated AI cloud infrastructure and deep learning platform launched by SenseTime. It is positioned to build efficient, low-cost, and large-scale AI cloud infrastructure, create a professional deep learning platform and algorithm model system, and help industry and academia explore the boundaries of AI.
In terms of use, SenseCore SenseTime AI Infrastructure is used to support large-scale AI model training and application deployment. The computing power it manages enables unified nationwide networked scheduling and has expanded new computing nodes in Shanghai, Shenzhen, Guangzhou, Fuzhou, Jinan, Chongqing, and other places. The platform supports both SenseTime's own large model research and development and external customers' large model training and application deployment.
Main capabilities include: powerful computing power can support the simultaneous training of more than 20 large models with over 100 billion parameters and support the full lifecycle generation of large models with trillions of parameters; provide end-to-end AI expert services covering the full stack of services such as business consulting, scenario design, model training, and inference deployment; provide two modes, productized license delivery and cloud service delivery, with professional personnel providing product usage training and technical support; provide security protection services such as real-time monitoring, analysis, and early warning to protect the security of users' cloud resources; gather AI industry experts and form an after-sales team to provide technical support and after-sales assurance.
In terms of versions, SenseCore 2.0 has been newly released, featuring an open ecosystem, ultimate elasticity, Region-level fully interconnected network, and providing the next generation of integrated training and inference AI Infra. The platform has also released Token Plan, providing developers and enterprises with high-value, high-frequency, and scalable productivity capability packages; released cloud server ECS, supporting full lifecycle management of cloud servers and providing high-performance GPU and CPU computing power. Related services also include managed Kubernetes service ECP and DeepLink heterogeneous hybrid scheduling solution.
Applicable scenarios: for industry and academia, supporting large model training, inference deployment, application deployment, etc. Specific usage methods, fees, and pricing information are not yet provided. Contact information: phone 400 001 4090, email business@sensetime.com.
