A naukri.com initiative
Windows
4w
219
Image Credit: Windows
Available today: DeepSeek R1 7B & 14B distilled models for Copilot+ PCs via Azure AI Foundry – further expanding AI on the edge
- Microsoft continues to push the boundaries of AI innovation by introducing DeepSeek R1 7B & 14B distilled models for Copilot+ PCs via Azure AI Foundry.
- The availability of these models reinforces the commitment to providing fast, efficient AI capabilities for real-world applications.
- Copilot+ PCs powered by Qualcomm Snapdragon X, Intel Core Ultra 200V, and AMD Ryzen will support the new models.
- Running 7B and 14B parameter reasoning models on NPUs signifies a milestone in AI democratization, enabling users to leverage powerful machine learning models directly from their PCs.
- NPUs in Copilot+ PCs are designed for efficient AI inference, allowing CPUs and GPUs to handle other tasks while maintaining superior results.
- Efficient inferencing has become crucial for language models, enabling improved response quality across various tasks through enhanced computational power.
- Microsoft's research investments have led to innovations like Phi Silica, enabling low-bit inference on NPUs for advanced scenarios like model fine-tuning.
- Developers can access and run DeepSeek models by downloading the AI Toolkit VS Code extension, allowing experimentation with different model variants on Copilot+ PCs.
- Copilot+ PCs offer local compute capabilities alongside Azure services, facilitating training and fine-tuning of models on-device while leveraging the cloud for intensive workloads.
- This integration of cloud and edge computing signifies a new paradigm of continuous compute, empowering developers to create innovative AI solutions.
Read Full Article
13 Likes
For uninterrupted reading, download the app