Core Innovations

Data Center Scene Empowered TCF (Industrial Manufacturing) Innovative Solution

Differentiation Advantages

• Disruptive to traditional flow control modes:

  Replaces traditional target-type flow controllers that are bulky and non-transparent. It can be directly applied to transparent observation windows/protective panels of data center server cabinets and 5G base station transmission equipment, without affecting visualized operation and maintenance of equipment while realizing full-link flow monitoring.

  The packet loss rate is reduced by 15% compared with traditional solutions, meeting the stringent requirements of 5G core networks for millisecond-level response and zero-interruption transmission.
• Intelligent dynamic adaptation capability:

  Through deep learning of traffic characteristics of 5G services (e.g., HD video, industrial IoT) and cloud computing tasks (e.g., distributed computing, big data analysis), the LSTM neural network can accurately identify sudden traffic (e.g., DDoS attack precursors, service peak fluctuations) and automatically adjust bandwidth allocation of conductive paths.

  It ensures priority transmission of core services and dynamic adaptation of non-core services, improving network resource utilization by 30%.
• High compatibility and low power consumption:

  Light transmittance ≥ 88%, sheet resistance ≤ 25 Ω/sq, without affecting equipment heat dissipation and visual observation.

  Chip power consumption ≤ 0.3 W, supporting PoE power supply. It can be directly connected to the existing SDN management platform of data centers without additional independent control equipment, reducing installation cost by 40%.

  Compatible with network equipment from mainstream manufacturers including Huawei, H3C and Cisco.

Technical Support

• Micron-scale circuit fabrication:

  Using UV laser etching technology combined with a high-precision visual positioning system, distributed electrode arrays are fabricated on TCF substrates with line width strictly controlled to < 10 μm and uniform electrode spacing (500 μm per group), ensuring accurate flow signal collection without compromising light transmittance.
• Intelligent algorithm integration:

  The LSTM neural network chip integrates a built-in traffic feature recognition model, which can quickly distinguish normal service traffic from abnormal fluctuations with recognition accuracy ≥ 98%.

  It links with the SDN controller in real time via the SPI interface, realizing closed-loop control of “flow monitoring – signal analysis – conductivity adjustment” with response time stably within 50 ms.
• Stable transmission guarantee:

  The TCF substrate adopts PET/PI composite material, and customized conductive adhesive is used to firmly bond electrodes and the substrate. The electrode conductivity remains 100% after 1000 times of 180° bending.

  The film surface is coated with an antistatic coating (surface resistance: 10⁶–10⁹ Ω), avoiding electrostatic interference to flow monitoring in data centers and ensuring long-term stable operation.