There has been a lot of talk about AMD Fusion APUs (Accelerated Processing Unit) for client products here at AMD, and often the question comes up, “what about AMD Fusion APUs for servers?”  AMD Fusion APUs will become a reality for notebooks and desktops in the first half of 2011 and as we recently demonstrated at Computex, there is likely to be an immediate gain in application performance because the mainstream applications are there today to take advantage of AMD Fusion APUs.

Let’s step away for a minute from APUs for servers and focus for a minute on the concept of Fusion or heterogeneous computing for servers. Today in the server world, CPUs do all of the heavy lifting in the vast majority of applications and typically only mathematical operations are optimized for GPUs today.  Case in point, the current #7 cluster in the Top 500 Supercomputer list is run on AMD GPUs. Clusters such as this have specialized software for the GPU to run parallel math operations that assist the CPU in decoding instructions and solving large computing problems. This is an example of where an application has been recompiled, likely with a language/framework like OpenCL or DirectCompute to take advantage of the GPGPU environment.

The need to recompile the application is one reason why you have yet to see heterogeneous computing for servers widely adopted thus far and why I believe the push for an APU within servers is still to come.  It’s something AMD is committed to delivering in the years ahead and when the time is right, more info on Fusion APUs for servers will come.  For now, AMD is focused on client APUs where GPU acceleration in mainstream applications is happening rapidly, Microsoft Office and Internet Explorer 9 are two of the most recent examples. Optimizing mainstream server applications is not even close to this level yet.

So how do we start bringing the concept of Fusion to servers and ultimately lay the groundwork for APUs in servers?  I think there are a number of things to consider in the march, but a couple of things stand out to me that can happen sooner rather than later: 

• OS-level enablement will likely grease the wheels for more rapid adoption among ISVs 
• Finding a way for the CPU to speak more directly to the GPU  

Future silicon advances, including our next generation of AMD FireStream™ solutions, planned for released later this month, will go a long way towards making this a reality. However, we shouldn’t assume that there is only one way to accomplish this level of integration. The key to the whole discussion is that this solution is driven by the ecosystem and will be software-led.  We’re seeing great progress in that area and the OpenCL/DirectCompute initiatives are helping further this movement.

As you begin to see and hear more buzz about AMD Fusion and AMD Fusion APUs on the client side, just know that there is a plan to bring this level of integration to servers. It’s just a bit more complicated, so we’ll continue to work out the details and expect to see more from us over time.

What other challenges do you see for Fusion/heterogeneous computing in server environments?

John Fruehe is the Director of Product Marketing for Server/Workstation products at AMD. His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied