• GenAI tools like Copilot and Adobe Firefly process tasks in remote data centres, requiring significant resources. Intel Meteor Lake chips and AMD’s NPUs enhance AI efficiency, with MSI’s AI engine automatically adjusting tasks.
  • Local AI is gaining popularity on Google’s Pixel 8 and Pixel 8 Pro smartphones, potentially enhancing cybersecurity by running extensive AI models locally.

Microsoft’s Copilot AI could soon run locally on PCs, using neural processing units capable of exceeding 40 trillion operations per second (TOPS). Intel expects these AI-enabled PCs to handle more elements of Copilot directly, potentially boosting performance and privacy by reducing delays caused by cloud processing.

Using Copilot on a PC

Copilot on Windows 11 and other genAI tools like ChatGPT and Adobe Firefly consume significant resources and power in a remote data center, whereas local applications like Stable Diffusion and Llama language modeling can be run locally. High-quality results require a PC with powerful processing power, particularly a high-speed GPU, which was once the focus of cryptocurrency mining.

Also read:Windows is under new management due to Microsoft AI reorganization

Recent hardware advancements, including neural processing units (NPUs) in Intel Meteor Lake chips and AMD’s offerings, have sparked discussions about AI-powered PCs. These low-power components aim to enhance AI processing efficiency and enable genAI tasks to operate seamlessly on battery power.

MSI’s AI engine optimizes laptop performance by automatically adjusting battery, fan speed, and screen settings based on user activities, enhancing gaming performance, and restoring it for Word documents.

Also read:Intel aims for over 100 million AI PCs globally by 2025

AI in Development

Google’s Pixel 8 and Pixel 8 Pro smartphones feature the Tensor G3 chip, which supports on-device generative AI. However, this hardware is not yet capable of running extensive AI models like Bard AI, Copilot, or ChatGPT locally, instead running more compact models. The local AI trend extends beyond PCs.

Local AI processing can enhance cybersecurity by reducing the risk of data flow and access to intellectual property tasks. Cybersecurity consultant John Bambenek highlights the potential for data loss in third-party cloud service breaches. By implementing Microsoft’s Copilot AI locally, organizations can maintain control over their data, removing the largest barrier to adoption as CISOs feel more secure.