Recently, SenseTime has successfully achieved full compatibility between its SenseCore Large Model Infrastructure and Huawei's Ascend 384 super node, fulfilling both functional and performance requirements. The Ascend 384 super node, a large-scale supercomputing solution introduced by Huawei, facilitates seamless resource interconnection and efficient pooling. Through collaborative efforts, the teams from SenseTime and Huawei have introduced a series of innovative enhancements. These improvements encompass scheduling optimization, ensuring cross-POD (Point of Delivery) training stability, and implementing multi-dimensional fault detection and recovery mechanisms. Looking ahead, both parties are committed to exploring additional application scenarios, with the aim of expediting the deployment and maximizing the utilization of SenseCore and the Ascend 384 super node.