| Mobile Web

LG Uplus speeds AI token efficiency research to cut service costs

LG Uplus is accelerating development of AI token efficiency technology to lower operating costs while maintaining service quality. It is applying model routing to match tasks with suitable AI models and researching quantisation, knowledge distillation and speculative decoding to speed token generation. The company is also pursuing chip-mix technology to use GPUs and NPUs by workload. It plans to expand token optimisation into the B2B market from next year.