Intelligent World 2030

ICT Services & Software 2030

New technologies, applications, and even paradigms are emerging as we approach an era of intelligence. Intelligent transformation will leverage new ICT infrastructure, intelligent digital platforms, operations, and O&M. AI-powered assistants and agents will make it easy for individuals, businesses, and industries to harness the power of inclusive AI. The future is here, and it is time to embrace it.

Download Report

Indicator prediction

Continuous innovation in computing power supply and breakthroughs in challenging resource constraints will become the main theme of data center development in the future

Copilot coverage: 100%

Agent coverage: 45%

Digital twin coverage: 30%

Key features

The service model is to be reconstructed using a new platform architecture, human-machine collaboration, and service-oriented effectiveness

Model-driven

Network service LLMs and foundation models will be built oriented to ICT service & software:

Network service LLMs: We pre-train foundation models to equip them with basic ICT service knowledge. This enables further industry upgrades and addresses the challenges of processing and training unstructured knowledge.
Network service foundation models: We construct foundation models that can interpret ICT protocols, signaling language, the lifecycle of planning, construction, maintenance, optimization, and operations, and OSS/BSS to solve more complex business issues. Foundation models can accurately interpret the chain of thought (COT) of complex tasks with an accuracy of over 99%. They implement a knowledge system of structured data equivalent to an undergraduate level of education.

Show Less

ICT Synergy Delivery

System clusters and O&M services will be combined to solve high MFU issues and facilitate ICT synergy delivery through:

1. Multi-level global load scheduling.
2. Availability of deterministic and competitive clusters: Faults and sub-health can be detected in advance for proactive maintenance. Resumable training after breakpoint is supported, and single-card exceptions have no impact on jobs.
3. Cost-effective linear network, AI computing, and storage performance optimization.

Show Less

Service-centric

Service-oriented experience management: We will specify the driving forces of reward and penalty for each indicator through modeling and analysis. Through the cross-analysis of the satisfaction matrix, we will be able to evaluate the NPS.
Business-oriented ODA O&M: Newly developed applications are managed as cloud components compliant with ODA standard interfaces and specifications. A platform for automatic software version verification and release will be provided.

Show Less

Intelligent Platform

Model enablement and agent platforms based on the ICT service and software industry are needed to deal with massive agents in the future. They help all customers to rapidly develop and operate the foundation models and agents of their industries based on a framework, enabling out-of-the-box capabilities. Model training assets can be replicated rapidly based on different scenarios.
Model enablement services include model selection, knowledge management, model training/fine-tuning, model evaluation, as well as model compression and inference.

Show Less

Digital Twin

Digital twin systems based on real-time spatiotemporal and network data allow for operations and O&M to extend from networks to business:

1. Spatiotemporal digital twins provide real-time data for network service models based on TAZ scenario identification and business type distribution differences oriented to users, businesses, and networks. The collection can be done in minutes rather than hours.
2. Real-time network awareness twins measure the impact of network issues on the business. Connecting the physical world with the digital world, they review KQIs and KPIs based on the relationship between the two worlds and business associations, and visualize and manage the direct impact of each operation and change on services.

Show Less