It just hit me. I was reading Raul Sood's blog, as linked from here (don't remember by whom, sorry I can't give credit). The original post even quoted the relevant part, but I don't think the poster quite made the connection to AMD's stated strategy: Become a Complete Fool
ATi is the best third party notebook chipset manufacturer in the world. They understand power management like no ones business, and they have the capabilities to build killer platforms for mobile platforms. (...) ATi + AMD under one roof = mobility platforms that will rival anything that anyone else can put out.
Do you remember AMD's priority list?
- servers first (check)
- notebooks second (working on it)
- desktops third (right now, using parts from either servers or notebooks)
Well, I think it's hard to disagree that server parts are a very healthy part of AMD's product line and so I'd put a checkmark on the first step of AMD's strategy. Not that they can rest on their laurels, but the next market for AMD will be notebooks.
So, while we can talk to death about integrating GPUs with CPUs, socket GPUs, memory strategies, I think the answer is far simpler.
But why not just design new in-house chipsets, you might ask? Well, I remember comments about designing chipsets for the original Athlons. AMD needed their own reference chipset, and according to the comments, designing, testing and manufacturing of those chipsets cost AMD a billion dollars. That's a large amount of money. And I'm sure that chipset revenue wasn't nearly that high. It was necessary though, because AMD needed a chipset and no 3rd party was selling one at the time.
Now, it's true that once you have a chipset, designing the next version of it is much cheaper. And AMD has current chipsets for their microprocessors, so it shouldn't be big deal to design a low-power one for notebooks, right? I don't think that's the case here. Designing low-power logic is quite complicated, and requires specific experience in the field to get it right. ATI has a design team that knows how to do this, so AMD is acquiring such a design team now. To be sure, some of AMD's designers also know how to do this, but they are busy building the next low-power notebook or server CPU.
Finally, why not just let their partners build 3rd party chipsets, like they always have? I think INTC has demonstrated that especially in the heavily embedded products like notebooks, creating a single platform and marketing it together can have huge benefits. I'm not even sure who makes mobile chipsets for AMD right now, but I suspect that outside of ATI, demand for such low-power mobile chipsets isn't very high. So they're not really shooting any of their partners in the foot, and nobody should get overly upset.
So, I think that with this acquisition, we're just seeing AMD's long-stated strategy go to the next step. The GPU business, integration, and reduced licensing costs are all just (very expensive, but probably worth it in the long term anyway) frosting on the cake. I expect AMD to make a very strong push for thin'n'light notebooks in the next year. Especially with K8L coming out soon!
Join the best community on the web! Becoming a full member of the Fool Community is easy, takes just a minute, and is very inexpensive.
It just hit me. I was reading Raul Sood's blog, as linked from here (don't remember by whom, sorry I can't give credit). The original post even quoted the relevant part, but I don't think the poster quite made the connection to AMD's stated strategy:
Become a Complete Fool