As part of the IOWN initiative, NTT, in cooperation with Red Hat, NVIDIA, and Fujitsu, has developed technology that uses IOWN technology to realize real-time AI analysis using suburban data centers while saving power.
This AI analysis platform utilizes the data processing acceleration methods of the IOWN All Photonics Network (APN) and IOWN Data Centric Infrastructure (DCI).
Through demonstration experiments, they found that in AI analysis using a suburban data center, the delay time (the time from receiving data at the sensor installation site to completing AI analysis at the suburban data center) has been reduced to the maximum compared to conventional methods. Confirmed that it can be reduced by 60%. It was also confirmed that the power consumption required for AI analysis for each camera in a suburban data center can be reduced by up to 40%. These will make it possible to consolidate AI real-time analysis processing at suburban data centers and save power.
In the future, they will combine the AI analysis platform with opto-electronic convergence technology to further improve power efficiency and aim to achieve carbon neutrality. Additionally, these results will be applied to the NTT Pavilion at the Osaka/Kansai Expo in 2025 as part of IOWN Computing, with the aim of commercialization in 2026.