This is a guest blog from Christian Heidarson, a principal analyst with Gartner’s semiconductor industry team.
The tagline for the 2010 IDF event was “Visibly Smart”, communicating the major leap in graphics processing capabilities that will come with its new Sandy Bridge processor architecture. In 2007, Intel promised to put more emphasis on graphics, publicly aiming for a ten-fold improvement in performance by 2010. Intel claims that Sandy Bridge will actually deliver 25 times the 2008 performance. While the 2008 baseline was at that time below the demands of even mainstream media applications- and the products won’t hit the street until early next year- this is still a remarkable achievement. Since 2008, Intel has
- moved its graphics engine design from 90nm to 32nm,
- implemented its graphics engine with “real” GPU cores with appropriate 3D graphics-specific optimizations
- optimized its graphics engine for high-end media processing needs
- set the foundation for an architecture that can tackle the parallel processing problems of the future, with support for OpenCL and DirectCompute
- integrated the processor and graphics engine into a unified chip architecture, that delivers performance gains and power savings
- complemented its specialized graphics engine with effective extensions to the traditional processor core micro-architecture
As impressive as Intel’s achievement is, the performance of Intel’s integrated graphics should not be over-stated.
- Sandy Bridge does not support DirectX 11, Microsoft’s latest graphics framework that delivers spectacular new visual features, nor the latest version of OpenGL which is used by Apple.
- Intel has been frank (when prompted) in declaring that its graphic performance targets are limited to the growing requirements of the mainstream market- NOT at the traditional gaming enthusiasts or graphics intensive workstations. The mid-range and high-end of the discrete graphics market will remain even after low-end discrete graphics are displaced by integrated processor/graphics products.
Furthermore, Sandy Bridge processors will likely not hit the mainstream market until the second quarter of 2011, and will only transition gradually across segments through the year. Until the fourth quarter of 2011, both Nvidia and AMD will have plenty of opportunities to take advantage of Intel’s current graphics limitations. In conclusion, Intel has not killed discrete graphics, but rather protected the processor status quo against building threats from the graphics industry.
Two questions remain:
- To what extent will integrated processor/graphics products hit margins in the mid-range discrete graphics market segment?
- Does AMD have the marketing muscle to push mainstream graphics requirements beyond what Sandy Bridge can deliver?
The Gartner Blog Network provides an opportunity for Gartner analysts to test ideas and move research forward. Because the content posted by Gartner analysts on this site does not undergo our standard editorial review, all comments or opinions expressed hereunder are those of the individual contributors and do not represent the views of Gartner, Inc. or its management.