ATI and AMD merger...

http://www.forbes.com/markets/economy/2006/05/31/amd-ati-technologies-0531markets10.html

This is rather upsetting.

Any comments?

I'm not sure why AMD would do it. Their bigger need is to break into the Dell and HP consumer market. They already do well in the gamer and hobbyist segment, and picking up ATI isn't going to help them increase their share of that segment. Most people building their own PCs are either in the NVidia camp (which will hurt AMD) or in the ATI camp, so it's likely a wash. If AMD does anything to give ATI an inside edge, it'll alienate the NVidia camp, and if AMD can't keep ATI competitive, they'll lose the ATI-fanboys eventually, too.

It would hurt the overall graphics market, as it would likely reduce ATI's competitiveness with NVidia in the short-term. NVidia will be loving it, as it should give them a nice window while AMD tries to digest ATI and start taking advantage of their technology.

With the next-gen Intel chips looking to finally win them back the performance edge, this would not be a good time for AMD to lose their focus. AFAIC, ATI can disappear. Too many years of buggy and closed-source drivers -- I've pretty much written them off now.

When I first drew a fellow sys admin's attention to it he literally screamed. "NOOooooooooooo....."

And then went on a rant rather similiar to yours.

Sigh ... beyond me. Maybe this is some sort of secret conspiracy dreamed up by some intel insiders in amd somewhere to break up the company somewhat so intel can get back the falling market share?!?

Rodrigo Istalindir said:

I'm not sure why AMD would do it. Their bigger need is to break into the Dell and HP consumer market. They already do well in the gamer and hobbyist segment, and picking up ATI isn't going to help them increase their share of that segment. Most people building their own PCs are either in the NVidia camp (which will hurt AMD) or in the ATI camp, so it's likely a wash. If AMD does anything to give ATI an inside edge, it'll alienate the NVidia camp, and if AMD can't keep ATI competitive, they'll lose the ATI-fanboys eventually, too.

It would hurt the overall graphics market, as it would likely reduce ATI's competitiveness with NVidia in the short-term. NVidia will be loving it, as it should give them a nice window while AMD tries to digest ATI and start taking advantage of their technology.

With the next-gen Intel chips looking to finally win them back the performance edge, this would not be a good time for AMD to lose their focus. AFAIC, ATI can disappear. Too many years of buggy and closed-source drivers -- I've pretty much written them off now.

Click to expand...

I couldn't disagree with this more.

The current ATI video cards are doing quite well, and in fact, the top end cards the ATI card looks like it will over the long term be a better buy (as in the x1900 will outperform the Nvidia 7900GTX in 2 years with most games) dispite being older. Nvidia has a slightly broader midrange market than ATI, and a few more suppliers, so their market share is slightly higher, but not as much as you might think. (This is coming from the owner of a 7800 GT btw).

What ATI will give AMD is it's own chipset manufacturer, and the ATI x200 chipset showed it has good potential, even if it was slightly behind the Nvidia counterpart in performance (less than 2%). This has been considered AMD's achilies heel for a while, and this is probably more of what this aquisition will address.

Also, there is some overlap in GPU and CPU research, particularly in memory management, so with AMD's current architecture, this seems like an ideal match.

I'm not sure why this would make either company less competitive, and in fact should make ATI more competitive, particularly in the chipset market as they will have first access to the newer AMD CPUs. Meanwhile, it will make AMD a more viable company overall, as they won't be as reliant on their CPU sales anymore (I believe they still do other chip manufacture, much like Intel).

The aftermarket sales of video cards are a drop in the bucket compared to the pre-installed and integrated market. ATI and NVidia are for all intents and purposes in a dead heat, with no sign that things will tip either way anytime soon. In the CPU and MB market, Intel dominates, with AMD so far behind it's not even funny. (And actually, I seem to recall that the CPU and MB market for Intel is pretty small compared to their embedded market.)

In the short term, this will eliminate the use of ATI chipsets in any Intel-based motherboard, which is a loss. This will cut them out of the vast majority of the market, since most (but not all) of HPs offerings are Intel-based, and Dell is still Intel-only in the desktop. This will likely also hurt both of them in the laptop market, which (a) is growing much faster than the desktop, and (b) is a part of the market where AMD doesn't have a credible offering already. Making yourself less attractive in the growth part of the market doesn't make much sense. AMD+Nvidia would have made more sense, since at least the nForce chipset is well-established and reasonably mature (although the networking in the series4 was buggy as hell).

The real risk in the next year or so is in digesting the acquisition. Its not uncommon for a company to stumble during this period, and right now, neither can afford a delay in their next generation technology. Any synergies (and I'm skeptical, as there isn't a lot of overlap right now between GPU and CPU design) are going to be a long time coming. ATI doesn't have their own fab plants, so AMD isn't addressing one of their other shortcomings, either.

AMD might see some long-term benefit in an all-in-one MB offering -- it might help them crack into Dell's, for example -- but there are two roadblocks there as well. First, they still won't be able to manufacture their own stuff, so high-volume availability could still be a concern. Secondly, they'll need to keep their price advantage over Intel, which doesn't give them a lot of wiggle-room in terms of feature-set. You won't magically see x1800s in $400 PCs just because of this.

I don't see how it will limit ATI usage in integrated chipsets, and many laptops now use a standardized connection for video cards so you can swap them out.

It may be an initial struggle, but that depends on how much they force integration. If they simply conect their Chipset research with CPU research, and leave the rest fairly well alone, it shouldn't have too much of an impact.

It is an interesting move, that has some benifits and some potential pitfalls. We'll see if it happens. Last rumor I heard about that sort of thing was Dell and Alienware, and that happened (and that was just suposition by some tech writer)

Another of my friends thought that this would actually be a catalyst for a lessening in overall quality for AMD ... he thought they would spend time integrating the ati stuff into thier boards and lose time and "neck and neck"'nitude in the Intel race.

I have, of course, seen that ATI has even ventured to release chipsets for motherboards with integrated graphics ... although now that I think of it I never bothered to look at any reviews for them, never considering ATI as a real possibility for muscling into the moboard chipset market with Via, SiS and now NVidia.

Do you think ATI might be for this possible merger to run "neck and neck" with NVidia in the chipset realm?

Hmmm interesting. I use ATI & AMD but I have no "loyalty" to either. I was reading the AMD roadmap and it looks pretty good for CPU's next year with the K8L looking like it will be a very nice chip. Of course they didn't mention an ATI merger and any possible impact so who knows. Maybe AMD will have a chip come out with a GPU core on the CPU? I don't know if that would be feasable though.