Cloud rendering breaks through the bottleneck of power grid 3D terminals, Dianliang Software has completed regional verification.
Against the backdrop of accelerated digital transformation of power grids, the 3D application of power grid GIS platforms is facing obvious experience bottlenecks: traditional solutions generally rely on client-side rendering via WebGL, and high-precision 3D models impose high requirements on terminal computing power, leading to stuttering and slow response when accessed by low-specification devices. Meanwhile, new services such as transparent substations and full-domain centralized management and control of power grids explicitly require multi-terminal access anytime and anywhere and support for high-concurrency access. The industry is in urgent need of a solution that can take into account performance, security and localization requirements. With more than 10 years of technical accumulation in the fields of real-time cloud rendering, low-latency transmission, and codec, the 3D cloud rendering solution for power grid GIS developed by Jinan Dianliang Software (hereinafter referred to as "Dianliang Software") targets this real pain point, and has thus entered the technical adaptation and implementation verification stage for energy power grid scenarios.
Terminal rendering hits bottlenecks, cloud rendering base cuts into power grid 3D scenarios
In response to the above pain points, Dianliang Software did not continue to "stack computing power" on the terminal side, but converted 3D rendering from a terminal capability to a cloud infrastructure capability. Its solution completes the whole process of 3D application rendering on the cloud-side GPU, and distributes it to multiple types of terminals such as PCs, mobile devices, and tablets in the form of encrypted video streams. Users can use the application immediately without installation, and no data will be stored locally throughout the process. At the same time, the solution improves the stability and resource utilization in high-concurrency scenarios through capabilities such as GPU resource pooling, multi-GPU load balancing, and multi-instance application isolation, striking a balance between rendering performance and deployment cost.
In terms of localization adaptation, Dianliang Cloud Stream has completed compatibility adaptation with mainstream domestic operating systems, CPUs and GPUs, including domestic operating systems such as Kylin, UnionTech, Euler, and Founder, domestic CPUs such as Hygon, Kunpeng, and Phytium, as well as domestic GPUs such as Moore Threads and Lisu. At the same time, the team has carried out full-stack information technology application innovation database adaptation and data reserve, providing data layer support for the subsequent large-scale implementation of 3D applications in power grids. This means that its solution can not only alleviate the stuttering problem of low-spec terminals accessing high-precision 3D scenarios, but also better meet the requirements of the energy industry for full-stack information technology application innovation and independent controllability.
From "Migrating to the Cloud" to Stable and Usable: Three Validations and Full-stack Information Technology Application Innovation
The core team of Dianliang Software has a core background in underlying software technology R&D, and has long been focused on fields including real-time cloud rendering, low-latency video transmission, GPU resource pooling scheduling, and full-stack information technology application innovation adaptation, and is good at polishing stable and reusable underlying technology bases. About two years ago, Dianliang Software launched the technical pre-research and solution preparation for this project, and then officially entered some regional power grid projects to carry out technical adaptation and practical verification.
The project has completed three core milestones so far: first, the adaptation verification is completed, realizing the implementation and adaptation of the real-time cloud rendering technical solution in regional power grid projects, focusing on supporting the upgrading of the 3D basic capability and data rendering capability of the power grid GIS platform; second, the core capability verification is completed, verifying the availability of capabilities such as GPU resource pooling, multi-instance application isolation, multi-GPU load balancing, and multi-terminal access in power grid scenarios, realizing that 3D applications are released in the form of encrypted video streams, users can use them immediately without installation and no data is stored locally; third, the full-stack information technology application innovation adaptation is completed, realizing compatibility and adaptation with mainstream domestic operating systems, CPUs, and GPUs, laying a foundation for subsequent large-scale implementation.
During the project promotion, the biggest difficulty for the team is to meet the four major requirements of performance, concurrency, security and localization at the same time. From "migrating 3D applications to the cloud" to "being stable and usable in the real power grid environment", the team gradually polished the solution through scenario-specific optimization and joint debugging with customers. At present, Dianliang Software has completed practical verification in a number of regional power grid projects, and has the foundation for multi-regional project collaboration, replication and promotion. The biggest challenge at present is how to realize engineering and large-scale replication under the superposition of multiple conditions, and the team is gradually solving the problem through phased promotion, establishing a standardized adaptation matrix, and accumulating templated delivery capabilities.
Do not replace business platforms, and promote large-scale replication through ecological cooperation
Different from most similar solutions on the market that focus on presentation layer functions, Dianliang Software places more emphasis on the capability of the underlying cloud rendering base. Its commercialization path is not to replace the original business platforms of customers, but to open cooperation to power grid construction and operation and maintenance units, GIS and system integration ecological partners, and provide underlying cloud rendering base capabilities. At the current stage, the work is mainly focused on regional power grid project adaptation verification and ecological cooperation. Its capabilities can also be extended to similar scenarios of China Southern Power Grid and other central energy enterprises, and the market expansion space depends on the progress of subsequent ecological cooperation and large-scale delivery.
At present, the energy and power grid industry is in the deep water zone of digital transformation, and the demand for 3D visualization in scenarios such as digital twin power grids, transparent substations, and full-domain management and control continues to grow, and the demand for cloud rendering bases is also released accordingly. Going forward, Dianliang Software will continue to follow up on existing regional power grid projects, cooperate with customers to complete scenario verification and launch guarantee, continuously polish core technical capabilities, accumulate standardized delivery capabilities, and support multi-regional large-scale replication through ecological cooperation. The team's deepest insight is that for underlying technology startups, there is no need to deliberately chase market trends. The key is to target the real pain points of the industry, deepen and stabilize the technologies you are good at, and you will naturally match the real industrial demands.