Browse Prior Art Database

Adaptive Video Memory Configuration for UMA Graphics

IP.com Disclosure Number: IPCOM000191586D
Original Publication Date: 2010-Jan-08
Included in the Prior Art Database: 2010-Jan-08
Document File: 2 page(s) / 85K

Publishing Venue

Lenovo

Abstract

This article describes the methodology how to allocate the graphics memory for the graphics controller included in the chipset which uses UMA (Unified Memory Architecture) to achieve the higher graphics performance. Generally, a graphics controller included in a chipset (Integrated Graphics) does not have dedicated memory devices, but uses a part of the system memory as the frame buffer or the work area of 3D graphics, which is called UMA. And in this case, the memory performance based on its bandwidth is very important for the graphics performance. Memory controller in recent chipsets support dual channel memory subsystem, which has two concurrent memory channels, and each channel has one memory slot. And generally, it supports the interleave mode, in which two memory modules work at the double bandwidth in the memory area of twice of the size of smaller memory module of the two. However, graphics memory is usually allocated in the higher address space of the system memory map, which is not in the interleave mode (i.e. non-interleave mode) when the size of two memory modules installed in a system is not identical. This article explains how graphics performance can be improved by allocating the high bandwidth memory area to the graphics frame buffer and 3D graphics work area communicating between the graphics chip and the system BIOS.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

AdaptiveVideoMemoryConfigurationfor UMA Graphics

This article describes the methodology how to allocate the graphics memory for the graphics controller included in the chipset which uses UMA (Unified Memory Architecture) to achieve the higher graphics performance.

Generally, a graphics controller included in a chipset (Integrated Graphics) does not have dedicated memory devices, but uses a part of the system memory as the frame buffer or the work area of 3D graphics, which is called UMA. And in this case, the memory performance based on its bandwidth is very important for the graphics performance. Memory controller in recent chipsets support dual channel memory subsystem, which has two concurrent memory channels, and each channel has one memory slot. And generally, it supports the interleave mode, in which two memory modules work at the double bandwidth in the memory area of twice of the size of smaller memory module of the two.

However, graphics memory is usually allocated in the higher address space of the system memory map, which is not in the interleave mode (i.e. non-interleave mode) when the size of two memory modules installed in a system is not identical.

This article explains how graphics performance can be improved by allocating the high bandwidth memory area to the graphics frame buffer and 3D graphics work area communicating between the graphics chip and the system BIOS.

The difference between interleave mode and non-interleave mode is explained below in figure-1.

 

                  Figure – 1 Interleave mode and non-interleave mode

The system configuration in which both of interleave mode and non-interleave mode has different memory bandwidth depending on the used address space,...