Optoelectronic integrated chips are expected to cut Token costs by 50%; domestic computing clusters could become a key factor in cost reduction

robot
Abstract generation in progress
Golden Finance reported that on July 19, as AI agents are being applied more deeply across industries, the number of Token calls has continued to grow at a rapid pace. In the future, how to further reduce computing power costs has become a key focus for all parties. Industry insiders say they are coordinating the use of multiple mainstream large models through computing power platforms, comprehensively reducing token costs. At the same time, they are also accelerating the construction of large-scale computing power clusters nationwide. Moreover, ultra-large-scale computing power clusters supported by domestically produced computing chips are viewed by industry as the key to cost reduction in the next step. In addition, in the next three to five years, the deployment of newer technologies such as optical chips is also expected to be another direction for exploring how to lower computing power costs. “Compared with electric chips, chips for optical-electronic integration have, at their core, lower computation latency and also lower power consumption. The cost per unit Token will correspondingly drop by 50%, or even more.” (Jin Ten)
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pinned