英特尔CEO陈立武：DeepSeek等中国模型性能足够好 普及很快

据路透社报道，英特尔CEO陈立武周一表示，中国前沿AI模型的词元(token)成本一直较低，且普及速度一直非常快。在周一由星展银行组织的一场小组讨论会上，陈立武表示，DeepSeek、Kimi和GLM等中国AI模型的性能“已经足够好”。

陈立武

“我是开源的坚定支持者，这是提高效率的最佳途径。”陈立武表示。他补充说，市场将决定哪些模型最适合不同的应用场景。

他还表示，量子计算可能在未来三到五年内成为主流，而软件将在提高部分必要基础设施的使用效率方面发挥作用。

本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。

According to Reuters, Intel CEO Chen Liwu said Monday that the Chinese frontier AI model has been cost-effective and has been spreading very quickly. On Monday, during a panel organized by Star Fair Bank, Chen Li Mu said that the performance of the Chinese AI models DeepSeek, Kimi and GLM was “good enough”.

Chen Li Wu

“I am a strong supporter of the open source, and this is the best way to improve efficiency.” Chen Ziu indicated. He added that the market would determine which models were most suitable for different applications.

He also indicated that quantum calculations could be mainstreamed over the next three to five years and that software would play a role in improving the efficiency of some of the necessary infrastructure.

Auto-fetched and translated by AI Atlas. All rights belong to the original author.