ATIC moves to buy AMD’s stake in GlobalFoundries

AMD currently owns a chunk of GlobalFoundries and enjoys equal voting rights with Advanced Technology Investment Company, but that could soon change. As Reuters reports, ATIC has filed an application with Germany’s Bundeskartellamt, or Federal Cartel Office, to take over GlobalFoundries entirely.

We tracked down a reference to the application on the Bundeskartellamt website, and indeed, even the automated Google translation leaves little to the imagination:

12.01.2010 B7-11/10 Company:

ATIC LLC, Abu Dhabi, acquisition of sole control over GLOBALFOUNDRIES (USA)

Product markets:

Semiconductor production

ATIC, which is owned by the government of Abu Dhabi in the United Arab Emirates, made no effort to downplay the move. Reuters quotes the firm as saying, "This action is simply consistent with the long-announced plan for AMD to gradually become fab-less, which was one of the key strategies behind the creation of Globalfoundries."

The creation of GlobalFoundries from AMD’s former manufacturing arm left AMD with a 34.2% stake in the new entity. Since then, ATIC has merged GlobalFoundries with Singapore-based Chartered Semiconductor, and AMD has struck a new cross-licensing agreement with Intel. Thanks to that agreement, AMD can divest itself of GlobalFoundries entirely and have all of its x86 microprocessors made by third parties without running the risk of legal issues.

Comments closed
    • zagortenay
    • 10 years ago

    Hmmm, a lot of wasted talents here…trying to give advice to multi billion Dollar investors.
    Die-hard Intel suckers feeling soooo sorry for AMD…now that’s really bizarre!

      • MadManOriginal
      • 10 years ago

      Wealthy investors don’t make the right call each and every time. What they do is make lots of calls and the calls they make that do well do i[

    • MadManOriginal
    • 10 years ago

    Well, that didn’t take long. Somehow I don’t see this ending well for AMD.

      • NeelyCam
      • 10 years ago

      Agreed. AMD will have to pay a high-price for costly SOI manufacturing, and will either 1) lose more money, or 2) hike prices and lose market share.

      I would bet on 2), since 1) would be the end of the company. 2) will result in higher CPU prices – something Intel will be very happy to play along with.

      And the CPU consumer loses.

    • green
    • 10 years ago

    l[

    • Stranger
    • 10 years ago

    Personally I’m suprised ATIC would want the entire portion. Primarily because once AMD is not tied to GF it is totally able to skip foundries as desired. The biggest incentive in bringing ATI to GF was because AMD owned it but if that’s no longer the case AMD is just as free to move their CPU buisness over to TSMC as well.

      • The Dark One
      • 10 years ago

      Considering how their 40nm process turned out, why would they?

      • UberGerbil
      • 10 years ago

      GF is scheduled to have the only non-Intel SOI/high-k fabs at the nodes AMD needs. If they want to be competitive, AMD needs to keep fabing CPUs (and eventually GPUs in the form of “fusion”) with GF.

      However, GF could now offer its bulk fabs to nVidia as a true neutral alternative to TSMC.

        • NeelyCam
        • 10 years ago

        ^ +1 Yes. Just wanted to point out (to others; I’m sure Ubergerbil knows this) that Intel does not have an SOI process.

    • wira020
    • 10 years ago

    I wonder if it’s possible for this plan to backfire on AMD?

    – AMD have to buy the wafer from GF now, right?
    -GF could make profit.
    – Why didnt they open the fab to others while they own it? Legal issues?

      • AlvinTheNerd
      • 10 years ago

      Legal issues and competitive issues.

      Would you expect AMD to buy wafers from Intel even if Intel opened up its fab to do so?

      You can’t get ahead of your competition if you are paying your competition in the process.

      However, now GF doesn’t compete on the processor level, they don’t make their own processors. So other companies are willing to use them.

      • kuraegomon
      • 10 years ago

      The usual spin-off story – your direct competition isn’t very likely to want to give you access to their IP in your fab. Spinning off the fab business allows you to not have to make expenditures to maintain/grow it, gain revenue by selling it, and have it be self-supporting (or even profitable) because it can now draw business from even your direct competitiors.

      That may in fact help said competitiors by giving them access to higher quality processes, but as long as you can out-engineer them and out-market them, you’re fine

        • A_Pickle
        • 10 years ago

        You know, both you and AlivnTheNerd basically give the same reason as to why this is an advantage, but I see a general lack of coverage for the disadvantages.

        What if, someday, AMD is unable to contract a semiconductor manufacturer to synthesize it’s product? There aren’t that many semiconductor fabs in the world, in fact, the industry of producing small processors like that is relatively limited in terms of it’s manufacturing capacity. There’s Intel, there’s TSMC, and there’s GlobalFoundries.

        g[

      • NeelyCam
      • 10 years ago

      Absolutely. AMD will be losing money because of this.

      AMD hasn’t been profitable for ages when GF was attached to the company. Part of it was due to the process development cost, but part of it was because the chips were too big and costly considering how much they could fetch in the market place.

      Now that AMD and GF are completely separated, GF is going to want to make a profit off the silicon they are manufacturing for AMD, and this is coming out of AMD’s profit margins. If AMD couldn’t make a profit when they owned the fabs, why would they make a profit now that the fabs are aiming for profits themselves?

      People claim that having multiple customers allows GF to spread out the process R&D costs across them, but AMD is still the only customer for GF’s high-performance SOI process: AMD will have to support its development alone. And that will cost money; those SOI silicon prices won’t be cheap.

        • Anomymous Gerbil
        • 10 years ago

        Brilliant massively-over-simplistic analysis. Fortunately, you’re not a decision-maker at these firms.

          • NeelyCam
          • 10 years ago

          Overly simplistic, maybe. But we’ll see how it goes…

          I’ve made my guess: AMD posting a loss now, and will post a loss for every quarter until a small profit in Q4/2011. What do you think will happen?

        • not@home
        • 10 years ago

        l[

          • NeelyCam
          • 10 years ago

          Yes, more or less. IBM is offering SOI foundry services, but few customers are willing to pay the price. AMD has been the big guy using SOI, and Cell processors are made with SOI.

          But beyond CPUs, there are few applications that require 3+ GHz core clocks… Graphics chips mainly benefit from massive parallelization instead of insane clocks, so slower and cheaper bulk silicon makes more sense.

          And the last time I checked, the only companies selling $100+ CPUs are Intel and AMD.

            • poulpy
            • 10 years ago

            q[

            • Anonymous Coward
            • 10 years ago

            Also some of the SPARC stuff. Sun has always been fabless and I could see them going to GlobFound.

            • NeelyCam
            • 10 years ago

            Not sure about SPARC – I’m wondering if Oracle will continue process development, considering the competition. But if they did, SPARCs would probably be on SOI, you’re right.

            • NeelyCam
            • 10 years ago

            Yes – I missed Power. And even Cell might cost more than $100. My bad.

            My point was that only AMD and Intel CPUs need the performance and are selling a high enough volume to support process development costs. Since Intel has its own bulk process, it’s up to AMD to pay most of the SOI development.

    • Arag0n
    • 10 years ago

    Advantages for AMD in short term selling the company:

    -Reduce financial costs.
    -More money for R&D to improve designs and architecture in GPU and CPU.

    Disadvantages:

    -Maybe lose of income in long term.
    -The fab capability dependant of third players could make hard to get back the leading from intel.

    So… the decision is not as easy to say wich option is the correct i guess…. My usually decision is looong term but if they have enought information to belive that they could do more with the money than with the invest, i guess it’s right.

      • NeelyCam
      • 10 years ago

      My guess is that they were so deep in debt that they had no choice.

      Overall, I think AMD is repositioning the company in the right direction. They will be in for a world of hurt in 2010 and 2011 (i.e., losses on top of losses), but will probably start recovering after that.

      I would say AMD should focus on servers, and similar high-price market segments (graphics?). They don’t have the cost structure to compete with Intel in the low-price market segments with decent profits.

        • SubSeven
        • 10 years ago

        Wipe yourself, you got crap dribbling out of your mouth… what you just said is the biggest load of shit I’ve ever heard (or read in this case). People have not freaking clue how the economy will perform this year yet you have enough knowledge and foresight to predict AMD losses for 2011. Take that crystal ball you have and return it to the 99 cent store you bought it from because you clearly got ripped off.

          • Welch
          • 10 years ago

          LMFAO…….. Great ahahaha.

          I don’t know how he can claim that AMD can’t compete with Intel in the Low-Cost market……. HA!!!!!! Thats what AMD does best, low cost parts. Sure argue that Intel’s Core I5 and I3 win out, but they were released very recently to AMD’s offerings, wait for AMD to put out their competition.. its always going to be a see-saw game of who has the bang for the buck, but what matters is who holds that title for the longest all the while selling the most of those while they hold that title.

            • MrDigi
            • 10 years ago

            Generally AMD doesn’t choose what their best CPUs sell for (currntly mid to low category), that is where their performance has ended up relative to Intel, since the last two or so years.

            I once heard said that over AMD’s existence they have never made money. I haven’t ever added up their earnings to check this, but it seems likely given all the bad years they have had.

            I would agree they are currently in survival mode.

            • anotherengineer
            • 10 years ago

            If they have never made money how did they buy ATI? I dont think a bank would give out a 5 billion loan to a company 5 billion in debt and never making a profit.

            I think they should have left ATI on its own, it would have been 5 billion in their pocket to help the weather the core 2 storm that followed.

            Aw well, I will keep on buying from them and hoping for the best, because if they disappear well will be paying 600 bucks for a 2.66ghz core2 and nehalems will be in the 1k + zone lol

            • NeelyCam
            • 10 years ago

            Loans, and giving away a bunch of stocks. You know, back then when AMD stock had some value… and back then they even had some profits. But after ATi acquisition: no profits.

            The last time AMD had a profitable quarter (GAAP) was in Q3/06. Over three years ago.

            • NeelyCam
            • 10 years ago

            AMD does low-cost parts because their products can’t compete with Intel at high-end.

            And don’t confuse performance/price (=what consumer sees) with price/cost (=profits). AMD CPUs cost too much to make compared to revenue they bring in. Intel’s manufacturing is just brutally efficient in cutting cost, from high yields to massive economies of scale to having a process node advantage.

            AMD simply can’t compete profitably in markets where low cost is the main driver for profit. They have to focus on high-price markets where performance is king, and cost inefficiencies don’t hurt as much. Unfortunately, they don’t have a CPU product capable of competing with i7’s and Nehalem-based Xeons on performance.

            Magny Cours might be an interesting twist, but due to its massive size, I don’t think it can compete with Intel’s offerings (Gulftown/Beckton) profitably. Bulldozer is supposed to bring AMD back to the performance game big time in 2011, but until then they’ll keep bleeding money.

          • MadManOriginal
          • 10 years ago

          Where did he say anything about the economy? 2011 was chosen for a different and more informed reason – that’s when AMDs next big new products come out. Bulldozer is the new CPU and I think Llano is the name of the CPU-GPU combo design. Or maybe Llano is the GPU part of the hybrid *shrug* In any case he has a reason for quoting the year.

          The only reason AMD can look like it’s doing well ‘all of a sudden’ is because they shuffled off the expensive part of their operations to another business entity which then turned around and effectively gave them ‘free’ silicon. It’s a lot easier to show a profit when your cost of goods sold goes down radically. This is the potentially huge problem with spinning off GF – once GF really starts to operate as its own company out to make a profit and other firms start competing for GFs capacity AMD could get squeezed. Things might turn out rosy and great with GF being AMD’s best bud (they should have named the company something that shortens to BFF instead 😀 hmm…Brainfart Foundries?) but once AMD loses a controlling interest in GF they lose leverage and there’s a lot more ways things could go wrong for AMD.

          People might not like Cam because of his brash Intel fanboyism but at least his posts have some rationale behind them and don’t read like a child throwing a temper tantrum.

            • SubSeven
            • 10 years ago

            Mad, Neely is rashly making foolish predictions on AMD’s profitability in 2010 and 2011. My point was that people don’t know how the economy would turn out this year let alone next year yet he somehow knows what will happen in AMD during this horizon? A very premature and erroneous notion in my opinion. Surely you agree with me that AMD success is heavily reliant on overall health and recovery of the current economy. Secondly people in here don’t understand that profitability is a highly subjective metric. Like MrDigi (#22) pointed out that he wasn’t sure if AMD ever made a profit. My point to that is who cares? Ideally it would be great if AMD posted losses every quarter for its whole lifetime as it wouldn’t have to pay taxes. As long as its free cash flows from operations are positive, that is all that really matters. Unfortunately, hiding profits via accounting transactions happens to be illegal but I give props to AMD for successfully “hiding” true profitability thus far (their real losses are nowhere near as dramatic as what one would read in the recent articles). People don’t comprehend that things like depreciation and write off of goodwill and “losses” on product impairment do not impact cash as they are only accounting entries. They are stuck on the stupid bottom line without understanding the whole picture. And with poor understanding, they make stupid and premature prediction on company profitability (a la Neely). I am sorry if I sound conceited but this kind of ignorance really gets to me.

            I agree on what you said about Neely, he does have some logic and makes valid points (unlike some other fools here; AMDISDEC and the what’s wrong with Intel guy come to mind) but his irrational fanboyism destroys his credibility and makes him look like the forum clown. With respect to the “a child throwing a temper tantrum” remark, were you implying me with that?

            • NeelyCam
            • 10 years ago

            Boy, you are confused – maybe you should go back to Economics 101.

            Do you think depreciation is just an “accounting entry”? A huge amount of real money was spent on the item (e.g., a fab). This cost is spread around over multiple years in accounting through depreciation, but make no mistake: cash flow did happen from AMD’s bank account to the contractors and suppliers.

            l[<"...he wasn't sure if AMD ever made a profit. My point to that is who cares? Ideally it would be great if AMD posted losses every quarter for its whole lifetime as it wouldn't have to pay taxes."<]l l[<"Unfortunately, hiding profits via accounting transactions happens to be illegal"<]l Unbelievable... Who are you - Jeffery Skilling?! And so let me get this straight: you are claiming AMD is truly profitable although they have posted losses every quarter three years in a row? Talk about destroying credibility...

            • SubSeven
            • 10 years ago

            Oh I’m confused? Interesting! Thank you for telling me where I need to go. Oh, and just as an FYI, I have been to econ 101, and 201 and 301 and 401 and quite a few others. Yes depreciation is an accounting entry you accounting imbecile. Do you even know the purpose of depreciation? Clearly you don’t so I’ll go ahead and answer that for you. Depreciation is a tax shelter designed by the IRS to provide incentive for companies to invest in equipment. Unfortunately the IRS did quite a poor job at specifying methods of depreciation originally and because of this, people rushed to depreciate virtually the whole cost of the equipment to offset income gains. To put it as an example, you buy 1M in equipment, on the first year you make 500K income. If you depreciate all of the equipment in the first year (which is what companies did in the beginning) your actual net income is a loss of 500K. So now explain to me how is depreciation relevant to true income and cash? What you fail to understand is that companies (especially asset intensive firms like AMD) love their depreciation because it’s a great way to offset income. With respect to initial cash outflow to purchase the equipment, don’t forget to account for cash inflows generated from the equipment. This is why I said as long as cash flows generated from operations are positive, that’s all that really matters.

            No I am not Jeff Skilling, I am simply someone that understands accounting methodologies and tricks as well as he does. But he chose to use that knowledge to cook books, which for your information, all companies do to one degree or another.

            I never claimed AMD was truly profitable for all the quarters in the last three years. I was merely stating that they were nowhere nearly as unprofitable as their bottom line “net income” figure, which every moron in the world only looks at, would suggest. There were several quarters that AMD was truly profitable although their net income was negative.

            All I want to say is that you are pretty smart guy with good insights but you should use your head instead of buying what others say and shooting crap from the hip without ever giving a second’s thought to what you write.

          • mutarasector
          • 10 years ago

          I agree. That was a load of shite. AMD’s going fabless was an absolute necessary move. It was more important to become a complete platform vendor than it was to continue being a processor maker only. Yes, AMD’s fab operation caused a lot of arterial bleeding for AMD for a long time, but not because it was necessarilly a loser per se, but because the startup costs of newer fabs couldn’t be amortized quickly enough during a long soft economic period. It really was more of a poor cash flow situation.

          There really is no significant downside in AMD’s going fabless, quite the opposite in fact. The funny thing is, some of the industry news pundits who criticized AMD’s ATI aquisition a few years back are also saying the same thing currently about AMD going fabless. Roll the clock ahead a few years, and it turns out that AMD’s ATI division is what has ended up keeping ATI alive, and minimizing ATI’s cash bleed. At the time of the aquisition, NVidia looked to be ready to pound AMD into the ground on the graphics processor front, and today the situation is completely reversed. I guess AMD isn’t looking so stupid now, are they?

          Given the divergence of the CPU market into primarily two disparate markets (low power and high power, and virtually nothing in between), I think AMD has positioned itself for a recovery and a return to competitiveness with Intel rather *nicely*. They’re not quite out of the woods yet, but looking back in hindsight at the path they’ve navigated since the beginning of their financial woes, frankly I’m amazed at just how they’ve successfully dodged bankruptcy or an aquisition this long. The odds against them surviving were definitely stacked against them.

          AMD’s current structuring and developmental roadmap seems to be working quite well and offers much promise in keeping the CPU/GPGPU markets very competitive for some time to come.

    • TheEmrys
    • 10 years ago

    Not sure if this is a good move for AMD. With the signing of Qualcomm, GF could turn a tidy profit year over year.

      • Kurotetsu
      • 10 years ago

      I’m pretty sure this is something ATIC and AMD already agreed to a long while back. I seriously doubt AMD could just suddenly turn around and decide to keep their shares.

        • khands
        • 10 years ago

        Even if they could, I’m not sure they’re in a position to do so, they need to get the core of their business back up and trading blows with Intel.

      • Prototyped
      • 10 years ago

      This may have been the tradeoff agreed to in order to receive the cash injection when AMD was financially at its nadir, essentially having ATIC buy the fabs off AMD’s hands so AMD could keep on designing chips.

      Also if you keep in mind just how expensive it is to research, develop and deploy upgrades to fabrication technology, it becomes more apparent why AMD would want to spin off the fabs and not have them vertically integrated within AMD — becoming a commercial fabrication company would allow for greater income to directly finance fabrication technology R&D. (If anything, that should make production of AMD processors cheaper as the R&D cost is spread out between AMD-as-client and GlobalFoundries’ other clients.)

    • jdaven
    • 10 years ago

    Not to mention get a huge chunk of change for selling off their share.

Pin It on Pinterest

Share This