NVIDIA GPUs Now Support Copilot+! Is AI PC Set for a Revolution?

🌈 Abstract

The article explores the impact of Microsoft's latest AI service, Copilot+, and NVIDIA's support on the future of personal computing. It discusses the capabilities of Copilot+, including its "PC time travel" feature, the hardware requirements to run it, and the potential implications for the Windows AI ecosystem.

🙋 Q&A

[01] Copilot+ and its Features

1. What are the key features of Microsoft's Copilot+ AI service?

• Copilot+ is a highly integrated, system-level AI service that can deliver proactive services.
• One of its key features is "Recall", which allows it to locate and deliver present and past material from your computer desktop.
• This "PC time travel" feature raises questions about security and privacy, but Copilot+ is designed to address these concerns by running as an on-line, locally-run AI model that does not rely on cloud computing power.

2. What are the hardware requirements for running Copilot+?

• Copilot+ requires a PC with a minimum of 40 TOPS of computing power, 16GB RAM, and 256GB storage.
• This level of computing power can only be achieved with either a PC with a recent NPU or by using an RTX series dedicated graphics card, as the latest mobile CPUs from Intel and AMD do not have sufficient computing power.

3. What are the implications of Copilot+ for the Windows AI ecosystem?

• The availability of Copilot+ on RTX 40 series laptops, in addition to mobile PCs with NPU chips, suggests that the AI service does not always need an NPU to be operational.
• This could be a significant boost for the Windows AI ecosystem, as more than 50% of the fastest computers already meet the minimum requirements to run Copilot+.

[02] Microsoft and NVIDIA's Collaboration

1. What is the relationship between Microsoft and NVIDIA in the context of Copilot+?

• Microsoft and NVIDIA have a mutual interest in their ecosystems becoming bigger, and they are not rivals from a hardware requirement perspective.
• NVIDIA is expected to account for a significant portion of Microsoft's results for FY 2023, making it NVIDIA's largest profit generator.
• Microsoft has also announced several large data center projects, many of which are likely to use NVIDIA hardware.

2. Why did NVIDIA initially mention that Copilot+ would not support its graphics cards?

• NVIDIA had earlier mentioned that Copilot+ would not support its graphics cards due to the more power-thirsty appetite of NVIDIA graphics cards, which were not suitable for being AI computing cores on mobile PCs.
• However, NVIDIA has now explicitly stated that its RTX series graphics cards are capable of Copilot+ support, indicating a change in their original plan.

3. What are the potential implications of Copilot+ being available on RTX graphics cards?

• The availability of Copilot+ on RTX graphics cards could be a significant boost for the Windows AI ecosystem, as these cards have a high level of AI computing power.
• This could open up new possibilities for Copilot+ in areas such as gaming, game design, special effects rendering, video editing, and photo editing, where the RTX cards' computational power could be leveraged.