Intel and AMD

Intel

Let me put this boldly: Intel is lucky to be where they are right now. They bet heavily on NetBurst, killed off the Pentium III architecture on the desktop, and no doubt would have done so for notebooks if P4s weren't so heat and energy inefficient. If not for the rising interest in notebooks and the incredible achievements of their Israeli lab, Intel could still be a year or two away from taking back the lead from AMD. This is a company that stumbled, more than led, the way forward. Its great size shielded it from the consequences of its own decisions. If they had been the size of AMD, with only the potential to run a single processor line and fewer engineering teams, the bet on NetBurst would have been far more damaging.

That said, Intel is in a fine position now. With Conroe putting them back on track, AMD's acquisition of ATI isn't that damaging. As with NVIDIA, if this had happened two years before, the consequences could have been more painful, but not now. ATI will still need a market for their products and Conroe looks to be the enthusiast platform of choice for at least the near future. In the short run, Intel's position may actually be stronger - they can deny bus licenses to ATI/AMD to prevent their chipsets being made, and AMD has spent a lot of capital and stock by acquiring ATI. The danger to Intel comes in the long run, to see what AMD can do.

As it is, however, Intel isn't affected much by the deal in the short run. Conroe is going to give Intel back its dominance in its key market, and ATI will have to wait and see if they can continue making chipsets - especially those important CrossFire chipsets - for Intel processors. Intel isn't as dependent on ATI as NVIDIA is on AMD, especially not with the new processors out, but it isn't necessarily in Intel's interest to deny enthusiasts access to a larger variety of graphics solutions. However, given that Conroe is currently putting pressure on AMD proper, Intel may choose to interfere with ATI's offerings, in order to decrease the revenues of the company as a whole.

Intel is also probably looking at a deal with, or acquisition of NVIDIA. They'll no doubt put a lot of thought into it, and try to avoid simply reacting by grabbing the asset. After all, right now NVIDIA is more dependent on Intel than vice-versa, and NVIDIA is an aggressive, motivated company in its own right. Being acquired by Intel could stunt NVIDIA's growth due to mismanagement, and possibly make Intel dependent on AMD's ATI division for graphics, a distasteful prospect. Remember, Intel has traditionally been terrible at graphics. It is therefore doubtful that Intel will seek an acquisition, unless it feels NVIDIA is on shaky ground, or AMD's acquisition of ATI begins paying off in unexpected ways and attacking Intel from new directions.

AMD

AMD took the biggest risk in the deal with an unclear payoff. A purchase of NVIDIA would make more sense on paper, since nForce is more important to AMD, but NVIDIA would likely drive a harder bargain and perhaps not integrate as well. ATI, on the other hand, seems less aggressive and determined - certainly their marketing is a weak point which at least suggests that more passive attitude. However, AMD has spent $5.4 billion on this acquisition, 80% of it cash, and their plan is rather uncertain. They have more debt now, but some tantalizing assets.

What AMD clearly gains is some variety. Just as ATI will benefit from having an outside revenue stream not tied to the volatile graphics card market, AMD gets revenue from a company that was competing against a target that was its own size, rather than a behemoth like Intel. AMD should finally have first-party high-quality chipsets for its processors, and they are now able to produce what OEMs want with on-board video and audio solutions. AMD made chipsets before, but that division was shut down by Ruiz. Maybe they belatedly realized it's a mistake. That's about it for obvious and clear gains by AMD.

As for the medium-turn possibilities, AMD has several ways to take advantage of the ATI deal. Graphics are advancing at blistering pace and are the bottleneck at the highest resolutions and settings. AMD may see itself able to take over the graphics market (Brandon strongly disagrees and will have his own theories posted later). Another option is that AMD is seeking to create its own platform - still Windows-compatible, but a more unified architecture like a Macintosh, or something along the lines of a desktop Centrino. Finally, AMD might simply seek to hurt Intel by denying ATI graphics to the company, but this is unlikely. For starters, it will cut about half the market or more for ATI cards and chipsets. Secondly, it may induce Intel to retaliate by acquiring NVIDIA just to deny their products to AMD.

Unlike ATI, which would have gained a lot to see this deal go through two years ago, or Intel and NVIDIA, which would have lost out more if it happened then, AMD isn't as affected itself by the timing. Two years ago AMD didn't have the financial clout to pull this off. It's a risky deal now, but it may have very well been impossible back then.

In the long term, however, a plan that AMD would no doubt love to execute to perfection, is the creation of a unified x86 platform. One of the major issues for the PC industry, whether office, home, or gaming, is support and compatibility. Various driver revisions, graphics combinations, motherboards, BIOS versions, processors - all combine to create a mess for programmers and end-users to work through. The amazing thing about it isn't what a disastrous minefield it is, but the fact that it actually works relatively well. "Relatively well" isn't good enough, however. Now imagine a PC that is almost like a console - key parts (chipset, graphics, CPU) from a single company, designed to work together, tested together, providing the kind of stability that everyone from server farm operators to game developers crave. It's still Windows, it's just the set of homogenous hardware which makes it special.