As more customers shift their business to the cloud, the company must ensure fast, seamless migration while maintaining strict data integrity and consistency. Any delays or inconsistencies could directly impact customer experience and service reliability.
As China Mobile is expanding into emerging businesses such as genomic computing, autonomous driving, and video rendering, its existing storage systems were unable to meet the performance requirements of these compute-intensive workloads.
To build next-generation AI compute centers capable of training large-scale models, delivering low-latency inference, and high GPU utilization, China Mobile needed ultra-high performance storage, and the ability to scale rapidly with evolving AI demands.
After evaluating numerous storage options, China Mobile selected TuringData as its partner to modernize its cloud-native and GPU-native infrastructure. By leveraging the Manila interface, TuringData seamlessly integrates its high-performance storage capabilities into China Mobile Cloud, providing a clear management interface that displays capacity and performance status for each directory. Combined with China Mobile’s native management components, this enables precise, reliable service delivery while ensuring data integrity and rapid onboarding of customer and enterprise workloads.
At the same time, TuringData provides the scalable data backbone required by the company’s GPU workloads. TuringData’s AI-optimized storage maximizes read/write performance, GPU utilization, and checkpointing efficiency—crucial for LLMs, genome sequencing, autonomous driving, and many other AI initiatives.
With TuringData’s high-performance storage system, China Mobile revamped its cloud services platform, achieving near-full infrastructure utilization while providing end users with fast, high-quality cloud and AI services.
TuringData significantly reduces cloud-based business processing times with its high-throughput and low-latency capabilities, accelerating business response times.
For cloud-native applications on the China Mobile Cloud platform, TuringData Platform effectively addresses I/O bottlenecks, enabling faster deployment and iteration of applications.
Automates GPU VM provisioning, workload scheduling, and tenant isolation, enabling developers and data teams to access AI compute resources instantly with cloud-like agility and efficiency.
Direct-storage GPU acceleration eliminates performance bottlenecks, minimizes idle time, and enhances ROI, delivering strong support for high-performance computing scenarios like AI model training.
Find out how much GPU utilization you're leaving on the table. Schedule a free 30-minute architecture review with our experts.