With Microsoft Windows 7 set to be released in October 2009, rumors are flying rampant in the video card industry as DirectX 11 graphics card will also be launching at the same time. As many of our readers are already aware AMD has already been pushing DirectX 11 as the next be step for graphics. Those who read Legit Reviews will remember AMD started promoting DirectX 11 all the way back to August 2008 when they discussed plans to add full support for DirectX 11 in their Stream Software Development Kit (SDK). More recently on June 2nd, 2009, AMD actually demonstrated tessellation on the worlds first DirectX 11 graphics processor! AMD says they are on track to launching their ‘Evergreen’ series of DirectX 11 cards around the time Microsoft Windows 7 launches, which makes sense due to all the marketing hype they are trying to build.
Some of the guys up in Toronto sent us an e-mail this morning and explained that DirectX 11 is important and we should take NVIDIA downplaying it with a grain of salt. It seems they are worried that the tried and true NVIDIA marketing machine will possibly start pumping out a lot of information to deflect and sway consumers. This is likely going to be the case as NVIDIA is rumored to be late to the game with DirectX 11 graphics cards and is going to have a rough time to get them released this calendar year. To help avoid this looming marketing battle, AMD has released a whitepaper that has been put together by AMD to showcase the importance of DirectX 11 and why it matters. They also gave us quick breakdown explaining that DX 11 is about efficiency.
Many of the DX 11 compute features are very technical and programming oriented, so they released this white paper to make sure the public understands the features and their importance going forward. That being said we will let you read the white paper and make your own conclusions on the importance of DirectX 11. Just looking at the leaked DirectX 11 screen shots, we would think that many would be impressed.
DirectX11 is the latest industry standard programming interface from Microsoft that provides accessto the advanced capabilities of next generation GPUs. It will be a key component of the new Windows 7 operating system, and Microsoft plans to eventually support it on Windows Vista as well with a software update.
One of the key new features of DirectX 11 is support for DirectX Compute, which enables developers to utilize the massive parallel processing power of modern GPUs to accelerate a much wider range of applications that were previously only executable on CPUs. Accessed via programs called Compute Shaders that are executed on the GPU, they can be used to enable new graphical techniques (such as order independent transparency, ray tracing, and advanced post-processing effects), or to accelerate a wide variety of non-graphics applications (such as video transcoding, video upscaling, game physics simulation, and artificial intelligence).
Today’s GPUs offer orders of magnitude more raw processing power than today’s CPUs for a given cost or power budget. However, as application-specific processors, GPUs lack some of the flexibility of CPUs, which can make it challenging for developers to access their full potential. More recent GPUs include new features designed to improve their flexibility and make these challenges easier to overcome. Another challenge has been coming up with a unified programming model for DirectX Compute that works across a range of different GPU architectures with varying capabilities. The DirectX 11 programming interface handles this by including a set of profiles called Shader Models. Each Shader Model includes a superset of the features in lower versions. Benefits of higher shader models include:
One advantage of Compute Shaders over other programming models for parallel processors is that they share a unified instruction set with other shader types used for graphics programming, such as Pixel Shaders and Vertex Shaders. So although Compute Shaders are a new feature of DirectX 11, some Shader Models with reduced features sets can run on earlier hardware, as follows:
This allows developers to choose between maximizing compatibility by choosing a lower Shader Model,or simplifying development and maximizing performance by using a higher Shader Model.
The AMD White Paper is continued on the next page…