On November 27, the "Optics Valley 2026 AI Innovation Conference & the 3rd Forbes China Top 50 AI Tech Companies Launch Ceremony" was held in Wuhan Optics Valley. Government officials, research institutions, global AI leaders, investment firms, and international organizations gathered to discuss cutting-edge topics such as "AI+" technology, industrial integration, regional collaboration, and capital empowerment. Cui Li, Chief Development Officer of ZTE, delivered a keynote speech, sharing insights on ZTE's vision and practices for the ubiquitous AI era.
**Evolution, Challenges, and Opportunities in the AI Wave** Cui Li reflected on the rapid evolution of large models over the past three years, noting their transformative impact. She highlighted milestones like ChatGPT (2022), Llama2's open-source breakthrough (2023), Sora's multimodal potential (late 2023), OpenAI's O1 (2024), and DeepSeek-R1's cost efficiency (late 2024). She emphasized that while AI is advancing into physical spaces via agents and embodied intelligence, achieving Artificial Superintelligence (ASI) remains the ultimate challenge, requiring stable, goal-aligned self-evolution mechanisms.
**Reshaping Industries with Intelligent Computing** Cui outlined key trends: surging demand, structural shifts in training-to-inference ratios, and greener, heterogeneous computing infrastructures leveraging liquid cooling and modular designs. "Compute-network integration" is emerging as a new foundation, enabling global resource orchestration. However, challenges persist—compute supply-demand imbalances, low utilization rates, fragmented standards, and the need for deeper technical adaptation to deliver value efficiently.
**Four Pillars of the AI-Driven Future** 1. **AI for Network**: Enhancing hardware/software intelligence and advancing autonomous networks (L4+) through Agentic AI and digital twins. 2. **Network for AI**: Optimizing connectivity from chip-to-chip (Die 2 Die) to data-center levels, supporting cloud-edge-device collaboration. 3. **Compute Innovations**: In-network computing, memory-compute fusion, and sustainable tech like liquid cooling and waste-heat recovery. 4. **Intelligent Systems**: Pursuing ASI via scaling laws, reinforcement learning, and world models while accelerating domain-specific applications.
**Connectivity as the Backbone** ZTE drives digital-physical integration through wireless (integrated sensing-communication, ultra-reliable 5G for industry) and wired (50G PON, FTTR) innovations, enabling immersive experiences and smart manufacturing.
**Full-Stack Solutions for Inclusive AI** ZTE’s portfolio spans AI servers, storage, and modular data centers, coupled with open standards for high-performance interconnectivity. Algorithm tools streamline model training/deployment, while AI terminals (smart screens, robots, phones) democratize access.
**Collaborative Ecosystem for Real-World Impact** ZTE partners with 1,000+ industry leaders across 20 sectors, focusing on data utilization, OICT convergence, and green digital infrastructure to elevate the real economy.
**From Digital to Physical: ZTE’s Commitment** Since 2016, ZTE’s digital/Intelligent transformation has yielded tangible results: - **Nebula Coder**: Quadrupled coding efficiency, earning SuperCLUE’s推理模型 gold. - **Communication AI**: Expanded from 8 chatbot agents to 20+ expert assistants, topping GAIA’s开源框架 rankings. - **Industrial AI**: Evolving from static to adaptive agents in smart factories like Nanjing Binjiang.
In closing, Cui Li affirmed, "We’re not just witnessing history—we’re shaping it. ZTE is ready to collaborate globally for an efficient, green, and inclusive AI future."
Comments