Microsoft’s Copilot-focused Surface Pro and Laptop PCs ship with Snapdragon and Intel options

In This Article:

At an event in New York City on Thursday, Microsoft unveiled two additions to the Surface line of PCs. The new Surface Pro and Surface Laptop both arrive with Copilot+ PC certification. Announced in May 2024, the program aims to identify business-focused devices that are designed specifically for the company’s AI platform, Copilot.

It's a bid to integrate Microsoft's AI play more deeply into the Windows 11 experience. The new Surfaces arrive amid a post-CES flood of Copilot+ PCs, including those from Acer, Asus, Dell, HP, and Lenovo. Copilot+ specifically refers to the inclusion of a neural processing unit (NPU), which allows for many features like Live Translation, Super Res, Recall, and Studio Effects to be performed on device.

Both devices ship with either an Intel Core Ultra (Series 2) or Snapdragon X processor. Earlier this month at CES, Qualcomm revealed plans to bring its 4nm Snapdragon chip to mid-range systems. Extended battery life and a lower price point are the processor’s primary advantages over the Intel option. In fact, Qualcomm promises “multi-day” use on a single charge compared to Intel's 22 hours. The primary tradeoff, however, is a reduction in system performance -- though Surface corporate VP Brett Ostrum tells TechCrunch that Copilot performance should remain largely unaffected.

Microsoft Copilot logo with laptop as backdrop
Image Credits:Microsoft

Copilot is a major focus for both configurations. For the Snapdragon, that means an eight-core processor with a devoted NPU specifically designed to manage AI workloads. The Series 2 Intel Core chips, meanwhile, are said to offer up to a 28% CPU performance and 98% graphics performance boost.

<span class="wp-block-image__credits"><strong>Image Credits:</strong>Microsoft</span>
Image Credits:Microsoft

Later this year, the Surface Laptop for business will get a 5G option for the first time. That connectivity is key to a play that combines local AI compute with cloud-based processing.

Most current generative AI platforms require more resources than consumer devices can provide, requiring that workload to be offloaded to remote servers. But this week’s DeepSeek reveal has turned the spotlight back to lighter-weight and more focused models that can run on-device. The Chinese AI startup’s strategy involves “distilled” versions of its R1 AI model capable of running on a laptop. Similarly, Apple has based its Apple Intelligence platform on smaller models that can run locally.

Microsoft is clearly interested in offering more on-device AI experiences, too. Much like Apple Intelligence, however, many of Copilot’s processes require more powerful hardware than can currently be crammed into a laptop. The addition of 5G is a bid to create a constant level of connectivity to assure that Copilot queries don’t come up blank over an inability to access remote servers.