罕见！两大AI巨头DeepSeek、寒武纪同步重磅发布，有何影响

Core Insights - DeepSeek and Cambricon have simultaneously released significant updates regarding their AI models, indicating a strong collaboration within the Chinese AI industry [1][2][7] Group 1: DeepSeek's Model Release - DeepSeek officially launched the DeepSeek-V3.2-Exp model on September 29, 2025, which introduces a new sparse attention mechanism aimed at improving training and inference efficiency for long texts [1] - The new model version is available across DeepSeek's app, web, and mini-program platforms, with a significant reduction in API costs by over 50% for developers [2][6] Group 2: Cambricon's Adaptation and Support - Cambricon announced the adaptation of DeepSeek-V3.2-Exp on the same day, providing an open-source inference engine, vLLM-MLU, for developers to experience the new model's features [3][5] - The company emphasizes its commitment to building a robust software ecosystem for large models, supporting mainstream open-source models like DeepSeek [6] - Cambricon has achieved industry-leading computational efficiency levels through extensive hardware-software optimization for the DeepSeek series, enhancing deployment performance and reducing costs [6]