Microsoft's Mu small language model partners with three major chip manufacturers to enhance user experience on PC platforms, attracting attention
Microsoft launched the Mu small-sized language model, designed for localized operation on personal computers to enhance user experience. This model will be applied to the intelligent agent module of the Windows system, supporting natural language processing. Microsoft collaborated with three major chip manufacturers: AMD, Intel, and Qualcomm, to optimize hardware performance. After the product launch, Microsoft's stock price rose slightly to $488.13 per share, with the market responding positively to its AI layout
According to Zhitong Finance APP, Microsoft (MSFT.US) officially launched the newly developed Mu small-sized language model on Monday local time. This AI tool, designed for localized operation on personal computers, has attracted market attention. According to Microsoft's official technical documentation, the model will first be applied to the intelligent agent module of the "Settings" feature in the Windows system, converting user input into system commands through natural language processing technology. It is currently open for testing to participants in the Insider program on the developer preview version of the Windows system equipped with the Copilot+ feature.
Notably, Microsoft emphasizes that the Mu model has achieved deep optimization at the hardware level, with all computing tasks completely offloaded to the Neural Processing Unit (NPU). The actual measured response speed has surpassed the processing threshold of 100 tokens per second, fully meeting the stringent requirements for instant interaction experience in system settings scenarios. In the model training phase, Microsoft built the training cluster using NVIDIA (NVDA.US) A100 data center GPUs, completing the full-process algorithm optimization relying on the Azure machine learning platform.
To ensure maximum performance of terminal devices, Microsoft has engaged in deep collaboration with three chip manufacturers: AMD (AMD.US), Intel (INTC.US), and Qualcomm (QCOM.US), to customize quantization computing solutions for different architectures of NPU units. This cross-industry collaborative strategy not only ensures the model's compatibility across various terminal devices but also achieves a significant improvement in energy efficiency through hardware acceleration technology.
Boosted by the news of the new product launch, Microsoft's stock price maintained a steady upward trend during the trading session. As of the time of writing, Microsoft’s after-hours trading rose slightly by 0.44%, reporting $488.13 per share, with the capital market responding positively to the tech giant's layout in the edge AI field
