The Platform

Acquiring ATI amounts to a quick (but expensive) way of filling in the gaps in AMD's current business.  AMD has already proved that it can compete technologically with Intel, and is currently working on fixing the problems with being able to compete in terms of manufacturing ability as well.  By acquiring ATI, AMD will have the talented workforce necessary to produce its own chipsets/motherboards with integrated graphics and engineer some very unique hybrid CPU/GPU platforms using Torrenza.

There are other ways AMD could have gone about attaining the same goal, for example by building its own workforce and IP rather than spending the $5.4 billion dollars necessary to acquire ATI's, but the acquisition approach is arguably quicker and allows AMD to focus on reaping the benefits sooner, not to mention that it leaves AMD better prepared for the future if GPUs do grow closer to the CPU. 

Intel has created the perfect example of how to be a successful microprocessor manufacturer, and its platform focus is one key element of that example.  The ATI acquisition, in many ways, is about following Intel's example and improving wherever AMD can. 

Can't We All Just Get Along?

The one element of this acquisition that you don't read about in press releases, is what it takes to actually make it happen.  It's not always easy to get a bunch of people from varying backgrounds and with various interests to work well together, it's even more difficult to take two well established and fully operational companies and expect to combine the workforces into one.  While direction for the combined company will come from both AMD and ATI senior management, making that translate into a well oiled machine that can not only innovate but execute great products is quite difficult. 

Having each company operate entirely independently makes no sense, since we've already discussed that it's what these two can do together that makes this acquisition so interesting.  AMD and ATI have to work as one company, but getting from where both companies are today to the point where they are one single harmonious entity (or at least as much as Intel) is going to be a very long and difficult process. 

One fundamental hurdle is that neither AMD nor ATI have particularly strong marketing, at least compared to their competitors.  Both Intel and NVIDIA have arguably done a much better job at marketing their products, building brands and gaining mind share.  We are concerned about the marketing direction that the new AMD would take, especially considering that in many ways ATI has the stronger PR/marketing focus.  At least from our dealings with the two companies, ATI gives more importance to its PR/marketing teams than AMD does, which is cause for concern since it is AMD buying ATI and not the other way around.  Only time will tell if AMD will assimilate ATI into its way of thinking, or if both companies will be able to use this acquisition as an opportunity to learn from one another.  We all know what the sensible choice would be, but getting thousands of people to agree on the same thing tends to complicate things. 

The one perspective that is easiest to support right now is Intel's "let's wait and see what happens" view.  Even with all this talk about the potential merger, the benefits, the pitfalls, etc. the reality is that right now all we have is a proposed merger.  It's certainly big news to even have such an acquisition out in the open, but until the final ink is dry many people remain skeptical.  It really wouldn't be too shocking to see the whole merger evaporate and for ATI and AMD to just continue on their present, independent paths -- certainly no more surprising than the initial announcement.

Our Thoughts: The GPU Side
POST A COMMENT

61 Comments

View All Comments

  • HopJokey - Wednesday, August 2, 2006 - link

    quote:

    The food in Intel's cafeteria is actually quite good :)

    I beg to differ. It gets old after a while:(
    Reply
  • Regs - Tuesday, August 1, 2006 - link

    The distant future looks good. Though we yet to see any more green slides about new core technologies from AMD. It almost seems AMD will be making baby-steps for the next 5 or so years to try to compete with the performance Intel is now currently offering.

    For stock holders - lets just hope AMD can pull something off to gain revenue from other markets with the help of Dell and ATi. Their growing capital and recent acquisition need some definite profits to pay it off.
    Reply
  • AnandThenMan - Tuesday, August 1, 2006 - link

    I think it's fair to say the article has a very strong pro Intel and NVIDIA slant. For starters, it needs to be pointed out that ATI is actually the #2 graphic maker, not NVIDIA. Saying that NVIDIA is #1 in the desktop space is only part of the market, so why state it that way? Trying to make NVIDIA look good of course...

    And this:
    quote:

    It really wouldn't be too shocking to see the whole merger evaporate and for ATI and AMD to just continue on their present, independent paths -- certainly no more surprising than the initial announcement.

    This statement is just dumb. Unless the planet is destroyed by an asteroid, the deal is pretty much done. It is HIGHLY unlikely that the deal will not happen.
    Reply
  • defter - Wednesday, August 2, 2006 - link

    The desktop market is very important market since most of the profits are made in the high-end desktop market.

    For example ATI has much bigger overall marketshare than NVidia (27.6% vs 20.3%) and has lot of presense in other markets (consumer electronics, handhelds). Still, NVidia has bigger revenue, meaning that ASP of NVidia chips is much higher.

    If you look at profits, the difference is even bigger, during the last quarter, NVidia made three times as much profit as ATI. Thus high-end desktop market is definitely very important.

    Here are some GPU market share numbers for Q2:
    http://www.xbitlabs.com/news/video/display/2006073...">http://www.xbitlabs.com/news/video/display/2006073...
    Reply
  • PrinceGaz - Wednesday, August 2, 2006 - link

    quote:

    The desktop market is very important market since most of the profits are made in the high-end desktop market.


    Most of the profits are not made in the high-end desktop market, in fact the very high end probably struggles just to break even due to the relatively tiny number of units shipped compared to development costs. Most of the money in discrete graphics is actually made in the low-end discrete graphics segment, cards like the 7300 and the X1300.
    Reply
  • defter - Wednesday, August 2, 2006 - link

    This is like saying: "most of the revenue is made on $100 CPUs instead of FX/Opteron parts..."

    The revenue can be higher on the low end of the market. But GPUs like 7300/X1300 are selling at $20 or less, profit margins for those can't very high. High-end chips like 7900/X1900 are selling for about $100 and the margins are much higher. (Compare the die size between 7900 and 7300, the difference isn't THAT big).
    Reply
  • JarredWalton - Wednesday, August 2, 2006 - link

    Hey, I'm a skeptic and you can blame me for the comment. Still, until the deal is well and truly done we have a proposed merger. Government interference, cold feet, whatever other setback you want... these things can and do happen. Do I think the deal *won't* happen? Nope - no more than I think the deal *will* happen. If you had asked me three months ago when I first heard the rumors, I think I would have been about 90% sure it wouldn't happen, so obviously I'm less skeptical now than before.

    As for NVIDIA and Intel slant, the NVIDIA perspective is their view. That doesn't mean it's correct, any more than the ATI, AMD, or Intel perspectives. However, ATI is #2 for the same reason Intel is #1: integrated graphics, specifically on laptops, and again we're talking about the underpowered, mediocre kind that will choke on Vista's Glass GUI. Wipe out all of the low-end GPUs, and NVIDIA has a clear lead in the market. Not in performance, necessarily, but in mindset and brand recognition? Definitely. We are an enthusiast website, and so we're looking at the stuff that moves the market forward, not just what suffices to run office apps.
    Reply
  • AnandThenMan - Wednesday, August 2, 2006 - link

    quote:

    Intel is #1: integrated graphics, specifically on laptops, and again we're talking about the underpowered, mediocre kind that will choke on Vista's Glass GUI. Wipe out all of the low-end GPUs, and NVIDIA has a clear lead in the market.

    Being #1 in one market is not good enough anymore. NVIDIA NEEDS to be in the integrated graphics sector, the ultra thin mobile sector, the console market, the HD devices market etc. etc. This is where ATI is much more diverse than NVIDIA.

    The article is about the implications of AMD/ATI and how it affects Intel, NVIDIA, and the whole industry. I understand what you are saying about the discreet enthusiest market, and naturally this is the most interesting and desirable segment we all like to talk about. But the merger is about much more than that. IMO, NVIDIA has to re-invent itself to be capable of taking on AMD/ATI. NVIDIA has come out and bragged about how they are not the "last man standing" but this is marketing spin at best. NVIDIA is on the record years ago as saying they want to "be where ever there is a pixel" but honestly, AMD/ATI is far better positioned to deliver this than NVIDIA IMO.
    Reply
  • defter - Wednesday, August 2, 2006 - link

    quote:

    NVIDIA NEEDS to be in the integrated graphics sector, the ultra thin mobile sector


    Care to elaborate? NVidia is doing fine financially, why it NEEDS to be strongly present on those sectors?

    quote:

    the console market


    NVidia has been in the console market since 2001.
    Reply
  • Calin - Wednesday, August 2, 2006 - link

    NVidia IS in the integrated graphics sector - if you are referring to the "enthusiast" integrated graphic sector Reply

Log in

Don't have an account? Sign up now