Server opportunity
Interim CEO Thomas Seifert recently spoke at a Goldman Sachs conference. He was quizzed extensively on tablets, and was careful to pay them the appropriate lip-service, but the real opportunity he identified for AMD and Fusion is in the server sector, where AMD's market-share has shrunk to an embarrassing level.
I've copied some key parts of the transcript below, but the long and short of it is that AMD has direct competitive advantage over Intel for the first time in years thanks to the superior performance of the graphics it integrates into its Fusion chips, compared to the graphics in Intel's Sandy Bridge architecture. This advantage will manifest itself not just in games, video, 3D, etc, but in the use of the GPU in general computing, where it's often more efficient and powerful than the CPU.
Thomas Seifert: Well, we were really pleased to show a demo today of our four-core Llano product that is going to ship in the second quarter,1.8 gigahertz, full DX11 capability, and we let it run against some I7 Sandy Bridge for architecture. And the visual graphic compute performance of our chip is just stunning. We outperform our competition by a factor of two to three in graphic performance.
And the applications that run today that provide the richness of your vivid experience, video streaming and all the applications that require massive parallel compute capability are letting us shine. And I think those are the applications that will matter on the consumer side moving forward. We talk about high definition, videoconferencing on your platform, cleaning up video streams. I mean, there are so many neat applications out there. And we have been investing a lot of money and time and resources to make sure that those applications are going to be ready for the launch and can be preloaded on our customers' systems.
So we have been starting to sample the Bulldozer product, too. It's a 32 nanometer product. It's all new server architecture.
It will be available in the beginning of the summer for the high end desktop segment. It will be available for the server segment and server applications in the late summer. And we have been building good product momentum and the performance benchmarks that we see today on workloads and power consumption will be stunning we think.
We have been disappointed, however. I am not going to hide that with the [server] market share development recently seen at the end of last quarter. We have been putting a significant amount of initiatives in place to correct that. And we are confident that the combination of the new products and what we have started to work on is going to pay dividends in the later part of this year.
However, I think over the last couple of months we have seen certain trends that make-that move usage models in our direction. So people have been playing with tablets now for a certain period of time. Expectations in terms of performance are going up. You want to deliver a 3-D game capability, you want to be able to watch high definition video streams, and resolution expectations are going up, driving pixel count up. And those are all goodness for us because it requires massive graphic capability and it requires good parallel compute capability.
So especially with the products--with the second generation APU products that are on the roadmap for next year, we feel confident that we can also address a significant amount of those form factors out in the tablet space.
But I also have to be honest that in terms of our priorities and where we see the big profit pools are for us, the server segment, and whatever mirrors those trends on the thin client side, whatever mirrors those trends in the infrastructure is significant for us. And we have to make sure that we will not get distracted from where those opportunities are, but it's an exciting form factor.
Right now Apple owns the tablet market and maybe it always will, but the market is still very young and may present new opportunities as it matures. AMD's own roadmap doesn't really have it producing sufficiently low power chips to play in this market for another year or two, and by then we may well see a version of Windows that supports ARM. Rather than being a bad thing for AMD, this may create a whole new segment defined by x86 and ARM chips coexisting in the same device.
AMD will want to be seen to be active in the tablet market, as indicated by product marketing VP Leslie Sobon's exclusive blog for us back in January, but for the sake of everyone who values competition in the x86 CPU market, I implore the AMD board to resist getting too distracted by new shiny things and focus on the massive opportunities still open to AMD in its core markets.