Industry positioning
For terrestrial telecom operators, we support new-technology R&D validation, test environment development for large-scale scientific facilities, and intelligent computing center testing.
Industry challenges
Operators’ technology evolution is accelerating: the integration of 5G-A and AI, integrated space-ground networking, and computing networks all require repeatable validation before large-scale commercial deployment. At the same time, major scientific research and experimental facilities place higher demands than commercial testing on the scale, programmability and reproducibility of test environments.
Application scenarios
① New-technology R&D testing
- Validation of new technologies related to 5G-A / AI-RAN (see “Research institutes and universities · AI-RAN testing”)
- Interoperability testing between NTN and terrestrial networks, and validation of integrated space-ground networking
- Functional and performance evaluation of new technology directions such as integrated communication and computing, deterministic networking, and network energy saving
- Simulation of 4G / 5G / NTN / 6G core network functions (EPC / 5GC / NTN / 6G core networks) and fully/partially surrounded testing, enabling a complete core network to be built without real network functions
② Large-scale scientific facilities
- Programmable and customizable end-to-end test environments for major national science and technology infrastructure and large-scale scientific research and experimental facilities
- Complete environment development spanning terminals, base stations, core networks (4G / 5G / NTN / 6G) and service traffic generators
- Protocol field-level editing and construction of non-standard procedures for frontier technology research scenarios
③ Intelligent computing center testing
- Unified computing-network simulation and testing: joint simulation across training, inference and network transmission to rehearse intelligent computing cluster designs without relying on physical clusters of thousands or tens of thousands of accelerators
- Training parallelism strategy selection: enumerate TP / PP / DP / EP combinations and overlay network topologies such as Spine-Leaf and Rail-Optimized to compare wall-clock time to convergence, MFU and construction costs
- Inference capacity planning: given peak-period request traffic, sweep replica counts and scheduling strategies to evaluate TTFT / TPOT / SLO attainment
- Lossless network configuration validation: validate the stability of PFC / ECN / DCQCN parameter combinations under high-volume training traffic
- Interference testing for mixed computing-network workloads: evaluate interference when training and inference share the same network, and determine the upper limit of training bandwidth that preserves inference SLOs
- Import topology and configuration from real clusters for digital twin simulation and design regression testing
Verifiable facts
- Supports large-scale 5G core network networking tests (historical project)
- Participated in the development of an industry-standard testing platform (App IPv6 Ready certification testing platform)
- Intelligent computing center network testing: a single instrument supports 100GE~400GE ports and covers all collective communication patterns (see the product pages)
Success stories
| Case | Content |
|---|---|
| China Telecom Group Research Institute | Test instrument framework cooperation and testing capability development over three consecutive phases |
| China Telecom Group Research Institute | 4G/5G core network centralized procurement network admission testing |
