2025-08-17 17:23:00
Table of Contents
Laptops with Intel Core Ultra processors gain flexibility for demanding tasks through a new graphics memory feature.
- Intel Core Ultra processors now support Shared GPU Memory Override.
- This feature allows integrated graphics to use a larger portion of system RAM.
- The enhancement aims to bridge the performance gap with discrete graphics cards.
- It’s beneficial for AI workloads and creative applications, not just gaming.
Intel has launched Shared GPU Memory Override, a new capability for integrated graphics on its Ultra Core processors. This feature allows the IGPU to tap into more system RAM, acting as additional graphics memory. Laptops with this upgrade can therefore handle more intensive applications with greater ease.
Can Intel’s new shared GPU memory feature improve laptop performance for demanding tasks like AI? Yes, Intel’s Shared GPU Memory Override allows integrated graphics to access more system RAM, boosting capabilities for AI and creative workloads.
This system is controlled via a slider within Intel Graphics Software. Users can adjust how much system memory is allocated. While the default setting is around 57%, Intel’s demonstrations have shown it can reach 87% on laptops equipped with substantial RAM. The company intends for this to narrow the performance distinction between integrated and dedicated graphics cards.
Bridging the Graphics Gap
This approach shares similarities with AMD’s variable memory graphics on its Ryzen platform. Intel highlights that the benefits extend beyond gaming. It specifically targets non-gaming tasks, such as running local AI models efficiently.
However, it’s important to note that increased VRAM doesn’t guarantee improved performance in every game. Some titles might load larger assets, leading to varied results. This feature is therefore most impactful for workloads that heavily rely on graphics memory, like local AI or certain creative applications.
To activate this feature, users must update their Intel drivers and restart their system. Minimum RAM and platform support are also required.
Future Implementations
Looking ahead, Original Equipment Manufacturers (OEMs) may offer similar memory allocation options within the BIOS. These would be guided by official specifications from each vendor.
This new capability provides IGPU laptops with more room for experimentation. It’s particularly significant for users looking to run on-device AI models without needing an internet connection. This feature presents a practical solution for such needs.
