back to article It's Big, it's Blue... it's simply FABLESS! IBM's chip-free future

IBM has coughed up $1.5bn to get Global Foundries to take its chip manufacturing arm off its hands as reported here at El Reg. There are a number of economic implications to all of this, perhaps the most important of which is why the eco-lunatic fantasy that we could reverse globalisation is simply nonsense (of course, whether …

  1. anatak

    scale and power

    The main problem with for example agriculture on a global scale has nothing to do with economics but with the power those global companies acquire over the food supply and how they will abuse that power. A monopoly might be economically speaking the most efficient model but economics never takes into account about the human psychology. Giving someone power will be a real test of character and it is obvious that our big businesses don't have a nice character.

    They can not have a nice character because if they would be nice they would be out competed by some other company without any morals / scruples.

    I am not against globalization but I am against giving big business too much power.

    1. MadMike

      large POWER8 servers are out:

      IBM has announced larger POWER8 servers. The largest, the P880 has 16-sockets and 16TB RAM. If you consider that Intels largest 8-sockets boxes has 12TB RAM, the POWER8 servers have little more to offer over x86. And Dell(?) will sell cheap POWER8 boxes with 2-sockets for somewhere like 2000 quid. That is not much. POWER8 is not high margin anymore, and that is the reason is killing it off.

      Accidentally, have you heard that IBM is going to kill off AIX too?

      http://news.cnet.com/IBM-Linux-is-the-logical-successor/2100-1016_3-982512.html

      It all makes sense. POWER8 is the last POWER generation. IBM Cell is killed off because of disappointments. AIX is going to be killed off, IBM has declared. IBM sold x86 hardware, chips, storage, etc. IBM is exiting all hardware, and transforming into a consulting company competing with Accenture. Mainframes are stagnating, there are no new customers, only old customers are upgrading.

      Bye, bye POWER and AIX. HP has also left the high end server segment. Left is Oracle who is betting more money than Sun ever did on large Unix servers. Everybody else are dead. And Fujitsu too, are selling large 64-socket Unix servers. Oracle and Fujitsu are the only one left in the high end Unix segment.

      1. Rob Isrob

        Re: large POWER8 servers are out:

        You like to trot that article out don't you? It's from 2003. I think I read an article in 1980 that said IBM was killing off mainframes.

        "POWER8 is the last POWER generation."

        http://www.theregister.co.uk/2013/08/27/ibm_power8_server_chip/

        "IBM was showing off a part, has systems of all sizes up and running in its labs using the Power8 chips, and has been designing the Power9 processor for quite a while already, according to Starke."

        Say... you aren't perchance a lib are you? Libs tend to make up facts, don't let facts get in the way of a good story, etc.

        1. asdf

          Re: large POWER8 servers are out:

          >Libs tend to make up facts

          "The announcement that Rep. Michele Bachmann (R-Minn.) is not seeking reelection will leave the Capitol a much less interesting place to fact check. As one of our colleagues put it, “The entire fact checking industry may have to hold a national day of mourning.”

          Bachmann is not just fast and loose with the facts; she is consistently and unapologetically so. No other lawmaker earned as high a percentage of Four-Pinocchio ratings as Bachmann"

          http://www.washingtonpost.com/blogs/fact-checker/post/michele-bachmann-a-fact-checkers-dream/2013/05/29/1481a6b8-c85e-11e2-9245-773c0123c027_blog.html

      2. Roo
        Windows

        Re: large POWER8 servers are out:

        "Bye, bye POWER and AIX. HP has also left the high end server segment. Left is Oracle who is betting

        more money than Sun ever did on large Unix servers. "

        Hmm, article about IBM going fabless (like SUN, Oracle, MIPS and ARM) and you post the same old FUD about the death of POWER... Are you trying to tell us that fabless = death of an architecture ?

        A cynic could be forgiven for concluding that Larry has forgotten to give you new instructions over the past 12 months. I wonder if you will get a new shilling to go with the "Oracle loves Clouds" message ?

      3. PowerMan@thinksis

        Re: large POWER8 servers are out:

        Ah, there is the Oracle FUD Machine! "MadMike" who has hijacked the moniker of "Mad Mike". A Oracle marketeer not interested in contributing or even stating competitive facts rather simply stating LIES. Yes, LIES.

        Does little good to refute him as he/she isn't interested in the facts as he/she has an agenda which is to spread more FUD about IBM, AIX and Power while trying to give the impression the only and best choice is Oracle, Solaris and SPARC. The ONLY thing he/she said that is true is that the E880 can have up to 16 sockets with 16 TB Ram. MadMike is pathetic.

        1. MadMike

          Re: large POWER8 servers are out:

          PowerMan@thinksis

          The ONLY thing he/she said that is true is that the E880 can have up to 16 sockets with 16 TB Ram. MadMike is pathetic.

          Well, I posted a link directly from IBM. Did I made up that link? It was not true? In that link, IBM explicitly says that AIX is going to be killed off. It IS true and nothing I made up. So AIX is going to die says IBM - was not a lie, as you IBM supporters claim.

          Regarding the 16-socket P880 server, I did not post a link to that one, but you acknowledge that piece of information as true.

          I also talked about DELL(?) or someone similar I dont remember, selling a cheap POWER8 server for around 2000 quid. Because you accuse me of FUD and making up false things, I googled a bit and found the link again. It was not actually DELL (which is why I wrote a question mark), no, it was TYAN who is selling a POWER8 server for $2700 quid. Which was pretty close to my recollection. Here is the link to the cheap POWER8 server:

          http://www.tyan.com/campaign/openpower/index.html

          My point is, POWER8 has left the high end segment. They are now cheaper than a decent x86 server and the largest POWER8 server is tiny at 16-sockets. No more POWER in the high end segment. We all know IBM only does high margin business. POWER is going to be axed next, besides x86 servers, storage, and whatnot. POWER is now cheap low end-midend servers. IBM knows that POWER can never compete with cheaper and faster x86 servers - why keep developing POWER?

          Regarding the "only and best choice" being a lie: well if you need some serious power, way more than the weak P880, if you need 32/64 sockets and 32TB RAM or so - what can you buy on the market that fulfills these requirements? IBM has no such large server, the P880 is hardly nothing more than two 8-socket servers glued together. HP has no such large server. Who is left other than Oracle and Fujitsu SPARC servers? I invite you to post links to a large server with 32 sockets, other than IBM, Oracle and HP. If you cannot, then I spoke true. Again. And all the time it was you the IBM crowd, that tried to FUD about my post.

          Actually, if you so strongly believe that POWER is NOT going to be killed off - why are there no further roadmaps? The POWER9 is not mentioned. AIX latest release was pity and in effect development has stopped. Show us the roadmaps of POWER and AIX. There are none. Why? Because "IBM bets heavily on AIX and POWER"? Are you fooling yourselves?

          It seems that you believe I will rejoice and be happy when AIX is killed, and POWER is soon dead. But that is not true. Competition is good for us customers, and if that happens there is no reason for Oracle to keep up the innovation pace and double performance every generation. Oracle has released five cpus in four years. Oracle will slow down when AIX and POWER has left the party. That is not good.

      4. Anonymous Coward
        Anonymous Coward

        Re: large POWER8 servers are out:

        @MadMike, Funny you say that, because the guy I was chatting to a couple of days ago about Power8 is now working on Power9+. Funny also that IBM subbing CPU fab to a 3rd party (like HP/Dell/Sun/Oracle) means death of Power - seems that ARM is doing pretty well without ever having any CPU fab capability...

        - Don't let the facts get in the way of a good story....

        1. MadMike

          Re: large POWER8 servers are out:

          Funny you say that, because the guy I was chatting to a couple of days ago about Power8 is now working on Power9+.

          And you expect people to trust something that an Anonymous Coward had heard somewhere from someone's aunts friends dogs owner? It sounds as when the IBM crowd flooded the forums about "SPARC64 is dead, it is cancelled, I talked to an Oracle executive who said they are going to kill SPARC64, I work at a large bank and we are migrating away from SPARC (but I love SPARC) because it turns out POWER is much cheaper/faster/stable/etc". The last thing is funny as POWER always have been much more expensive than SPARC. Everything was pure FUD. They "overheard something from someone on the train".

          And now you are keeping up the IBM tradition of FUDing made up and negative rumours. It is very easy to not FUD - show us credible links. Then you are not FUDing. I always show links: from IBM executives saying AIX is going to be killed, IBM roadmaps showing no POWER9, etc. Hence, I do not FUD (make up false things). Everything I write is true, and comes straight from Amarok (IBM HQ).

          Show us links if you want us to take you seriously.

  2. Gordon 10

    Globalisation

    Nice article but could have done with being a bit more precise about what form of globalisation you were talking about. There aren't actually many yurt eating fundamentalist anti-globalists in the work.

    There are also fairly few anti-product globalists in the work - especially where their own consumption is concerned - cheaper socks - bring em on.

    The only vocal protests about globalisation are the pure cost of labour based ones. ie Globo mega corp #265 cant be bothered to fix its fundamentally broken processes - so it uses volume of labour to slap a plaster over the problem for the next few years - that gets everyones goat coz a) they have lost their job b) most genuinely believe they could be part of the solution - not part of the problem.

    So the fundamental question not asked by this article is what was the true motivation behind IBM flogging off its FAB's - pure cost of manufacture due not enough economies of scale - or something else like a CEO desperately selling off the family silver due to a complete lack of ideas?

    Its also worth noting that for the finances of subsidiaries that are essentially internal cost centers is nearly always screwed up - because most globomegacorps don't actually have efficient internal markets where costs and revenues are distributed correctly.

    1. Tim Worstal

      Re: Globalisation

      "Nice article but could have done with being a bit more precise about what form of globalisation you were talking about."

      The piece coming up at the weekend goes into that subject using a different example.

      1. Gordon 10
        Thumb Up

        Re: Globalisation

        Cool. What would be really nice is one on the race to the bottom on labour outsourcing and how long we have until a global equilibrium of sorts is reached by where the incremental savings from outsourcing are not enough to pay for the cost and disruption of that outsourcing.

        Most companies in my industry are already on their 3rd or 4th offshoring country because the previous ones have become too expensive.

        Wage inflation in the India tech cities is quite big considering the sheer size of their labour market. Employee cycle time is massively high.

        Would make for an interesting model.

        1. Tim Worstal

          Re: Globalisation

          If I knew all of that I'd be writing up for a journal and then preparing to meet the King of Sweden a few years later :-)

          But there's certainly some general answers and that's a good idea for a couple of weeks' time. I've got a couple of subjects lined up already.....

        2. Anonymous Coward
          Anonymous Coward

          Re: Globalisation

          > What would be really nice is one on the race to the bottom on labour outsourcing and how long we have until a global equilibrium of sorts is reached by where the incremental savings from outsourcing are not enough to pay for the cost and disruption of that outsourcing.

          This for me is a very interesting question indeed. Presumably, a lot of those areas would have lifted themselves out of poverty and doing quite nicely but I suspect that a lot of those working in sweatshops to feed our desire for "things" would think otherwise.

          It's all rather complex but the "great global leveling exercise" is going apace rather nicely. We just need to give some of the foreign governments bent on the destruction of their countries a good kick in the nuts.

  3. FutureShock999

    This saddens me greatly. I remember growing up across the river from the Fishkill, NY fabs, and going on a tour of them as a kid, and getting an IBM keychain with a bit of IBM chippery embedded in Lucite. I thought it was SO COOL back then.

    This effectively spells the end of the POWER architecture too. Knowing that the support contract to produce new chips is only for 10 years, any business that is not already on POWER systems would have to have a huge second thought before committing to a future purchase of POWER-based systems, knowing that they are now effectively a dead end. And business on POWER-based systems have little reason to not look at alternatives very quickly, as they cycle their hardware, and thus consider other vendors.

    What that leads to is the following conclusion: this is really a HUGE abrogation of leadership (and perhaps fiduciary duties!) from the current IBM management team - they have, as usual in America - optimized their business for next year's quarterly reporting, and F*CK what happens to the business in the long-run. Can you imagine a Japanese company being so short-sighted as to effectively kill off their product line in the marketplace in a single gesture? Or, more to the point, the Chinese?

    No, Ginni & Co. have decided that all that matters is next year's executive bonus comp, and fattening their own pay checks, and to hell with what happens to an American business icon in 10 years. They ought to be lined up against the wall and shot**...

    (** - that is a _metaphor_ , for any plod or NSA or FBI reading this)

    1. frank ly

      It's not a metaphor

      It's hyperbole.

      1. FutureShock999

        Re: It's not a metaphor

        No, it's a metaphor. It would be hyperbole if the real solution was to do something painful that did not involve shooting them - say a good spanking.

        But since the real solution does not involve corporal punishment or violence, but more likely shareholder lawsuits, it IS a metaphor.

        1. Michael Wojcik Silver badge

          Re: It's not a metaphor

          No, it's a metaphor

          You're both wrong. The statement in question is exaggerated for rhetorical effect, which makes it hyperbolic; and it involves an (implied) association with no direct connection between vehicle and (implied) tenor, which makes it metaphoric. Tropes aren't mutually exclusive.

          While we're at it, it's also a cliché, argumentum ad baculum and ad populum, amphidiorthosis (in the footnote), epiphonema, exuperatio, deinosis, praemunitio (again in the footnote), and no doubt several others.

          And this whole thread, of course, is an exercise in correctio. Or diorismus, epanorthosis, epidiorthosis, or epitimesis; though the last sometimes is used to mean epiplexis.

          HTH. HAND.

    2. Anonymous Coward
      Anonymous Coward

      You are completely right on all counts except one

      This means the end of POWER outside embedded one-offs. Even console makers will stay away from them.

      Don't agree with your viewpoint on Ginni & Co., While I don't admire them for what they are doing (IBM is slowly self destructing itself) they are after all just trying to maximize shareholder value. The "in the long term" bit is missing from their mission statement, and that is very convenient for them since they are themselves shareholders and want to secure a nice retirement. Remember their "Roadmap 2015" initiative.

      But it is the same with HP and others: the original founder's objective had nothing to do with shareholder value but for creating some valuable contributions to society and get rich as part of doing it. This no longer holds true, publicly held corporations are about maximizing profit now, not tomorrow. The invisible hand will take care of the rest. Or not.

      1. FutureShock999

        Re: You are completely right on all counts except one

        They are NOT maximizing shareholder value in anything more than a year or two.

        Once the market digests this, they will find they cannot sell a POWER-based system to any of their major corporate accounts, due to lack of future support for the architecture. And having sold their x86 business to Lenovo, they have little to take up the slack. So overall, sales will plummet on hardware. And, systems integration work based on hardware sales will decline.

        That will take 2-3 years to really kick in, but it IS totally predictable, and shareholder value at the 5 and 10 year marks will be greatly negatively impacted. Of course, Ginni isn't planning on being their then....

        1. PowerMan@thinksis

          Re: You are completely right on all counts except one

          Can't tell if you are an instigator, competitor, maybe a disgruntled (ex)IBMer but because your boldness to claim what will happen with Power in the future I'll go with a competitor trying to inject FUD into the conversation. As a IBM BP who is invested in the product and who has a reason to be concerned - know that I am not concerned. I saw Sun use TI, UMC, TSMC. Guessing Oracle uses TSMC. I could look it up but don't care. Point is they use somebody and if you couldn't tell where Dell, HP and Lenovo get their Intel chips from, yet Lenovo and not from their own fabs. Good try IBM competitor.

          1. MadMike

            Re: You are completely right on all counts except one

            PowerMan@thinksis

            Why are you so strongly claiming that POWER is not going to be axed next? Why, are you working with POWER? Ooops, forgot, yes you are. But I dont believe IBM pays you to write this FUD, because IBM has stopped investing in POWER. Why spill money on astroturfers? It is clear that POWER is dead end.

            What is left of IBM hardware? Not much, everything has been sold off. Only left is POWER, Mainframes and Storage servers. Everything else is sold off. IBM knows that you earn the big money from consulting and services. That is why IBM has slowly been exiting all hardware bizz.

            POWER is going to die. A word of advice from a friend; abandon AIX and learn Linux instead. IBM says Linux is going to replace AIX.

            If you want to convince us that POWER is NOT going to die; you can do that simply by showing us a roadmap of POWER9 and POWER10. But there are none. Which means it is you that FUDs. :)

          2. Anonymous Coward
            Anonymous Coward

            Re: You are completely right on all counts except one

            " I saw Sun use TI, UMC, TSMC. Guessing Oracle uses TSMC"

            Not exactly examples of runaway successes selling machines.

    3. A Non e-mouse Silver badge

      @FutureShock999

      This effectively spells the end of the POWER architecture too. Knowing that the support contract to produce new chips is only for 10 years

      This doesn't mean that IBM will stop producing POWER systems in 10 years time. All it means is that in 10 years time IBM can go to other suppliers to make their silicon. (Or sign up again with GF.)

      The 10 year sign-up was probably a requirement from GF to ensure some revenue for the fabs to make it worth while buying them.

    4. theblackhand

      Re: FutureShock999

      "This effectively spells the end of the POWER architecture too. Knowing that the support contract to produce new chips is only for 10 years"

      I think you maybe reading a little too much into the move by IBM. I don't think IBM is trying to stop POWER (even unintentionally) - this is about securing GlobalFoundries medium term future as a manufacturer by allowing GF to present rosey forecasts to banks and investors to ensure they stay in business. And I would expect future agreements.

      With AMD struggling against Intel and a lot of TSMC/UMC/Samsung capacity tied up for the foreseeable future, keeping GlobalFoundaries in the game is important for IBM (and AMD).

    5. Ken Hagan Gold badge

      "to hell [...] in 10 years"

      There are two problems with planning 10 years ahead. Firstly, your company has to get through the next 9 before it can reap the rewards of your cleverness. Secondly, you might guess wrong about what the world looks like in 2024.

      1. Anonymous Coward
        Anonymous Coward

        Re: "to hell [...] in 10 years"

        "There are two problems with planning 10 years ahead."

        Guess there are quite more than two, but...

        "Firstly, your company has to get through the next 9 before it can reap the rewards of your cleverness"

        That's assuming your strategy is "big bang" success. What usually happens is gradual evolution. Nobody is surprised for example with outsourcing contracts losing money the first couple of years then recouping it in the next three.

        "Secondly, you might guess wrong about what the world looks like in 2024"

        I have to agree. Risk is exactly the reason the activity know as "investing" is something whose outcome is not 100% certain. To be clear, the longer the timescale, the riskier the predictions are, and that is always true.

        But in this concrete case, this move certainly does not seem to reach beyond the next 12 months. And the IBM board know very well where they want to be by then: vesting the stock bonus of their "Roadmap 2015" targets. What happens after that may or may not be difficult to predict, but essentially they don't care. Much less what happens to IBM in the following 9 years.

    6. Anonymous Coward
      Pint

      @FutureShock999

      Quite a few companies but let us keep it simple. Sony.

    7. PowerMan@thinksis

      Your comments started out as a sentimental note reflecting on memories of your youth only to have it turn into an attack piece full of unfounded statements. There is no reason to think that Power is dead because of this. It makes complete sense for IBM to continue using the technology, processes and personnel that built the chips today as they work on future products. They will begin working with Fabs - probably GF but maybe others like TSMC to see who can deliver the quality products needed at the price point.

  4. Cipher

    Krugman...

    ...has gone from being a Keynesian fanboi to Nobel Laureate to comedian. Or is that laughing stock?

    Krugman is a funnyman now

  5. Anonymous Coward
    Anonymous Coward

    It was all they could have done*

    IBM made a small number of chips in an expensive Fab.

    They were making a loss.

    Solution 1: Pay someone else to take it off their hands

    Solution 2: Sell the chips for peanuts and flood the market, it would cause a race to the bottom in the chip industry. Making the margin on bulk sales, would also have a negative effect on their brand and more importantly to their bonuses the share price.

    Solution 3: Retool, make (and sell) more chips, not necessarily IBM branded chips. Any chips with a high margin, e.g. FLIR sensors. Or any chips with a high volume. The problem with this is that you are making a profit from a product line to cover a loss making line. Any economist would say scrap the loss making line and lets make some real money.

    Solution 4: Shut it down.

    *So the only option to keep their stock stable was the first. And the long term effects for the U.S. of decisions like this will not be good. But it is all about the share price, and the next quarter and the quarter after that.

    1. Voland's right hand Silver badge

      Re: It was all they could have done*

      They already tried Solution 3. If memory served me right, FishKill was one of the sources for AMD and video card manufacturers. That proved insufficient.

    2. FutureShock999

      Re: It was all they could have done*

      You miss the point - the entire IBM mid-range product line is BASED upon these chips. It is also their main market differentiator, their USP.

      Getting rid of the fab, and getting only a 10 year guarantee, doesn't just close down the fab - IT SHOOTS THEIR PRODUCT LINE IN THE HEAD!

      If that fab is running at a loss, then guess what? Consider it the Cost of Goods Sold, and ameliorate it with one of your other strategies. Solution 3 works for this. Because Option 1 and 4 basically tell your customer that you cannot guarantee a future state of POWER chips beyond one or two more IT lifecycles, and so they should begin to consider other options in the marketplace. There are customers out there with thousands and thousands of blades running POWER...and you have just told them they have an infrastructure that will likely go out of support, and the smart ones will cut their losses sooner rather than later, killing IBMs mid-range sales. (remember that they just sold off their x86 server business to Lenovo, so WTF they are planning to sell without POWER escapes me. And I suspect Ginni as well...).

      The point is you cannot just consider the economics of the fab. You have to look at it in light of the entire IBM value proposition to the marketplace - anything else is a broad breech of fiduciary duties. And just f-ing stupid. And right now they have just cut their mid-range architecture off at the knees in the marketplace, and sold their x86 business. As I said, line that "management" team up against the wall...

      1. Denarius
        Meh

        Re: It was all they could have done*

        not quite full story. Remember PowerPC uses Si on Cu, a dead ended technology it seems. Good at time, just superceded. Intel has a better chip making technology. Accordingly, sell plant and contract output back. IBM have been encouraging Linux for a long time. How long do you think AIX has to run ? 10 years maybe ? Did IBM see this coming ? I think so.

        Finally, it is an excellent illustration of the problem for _BIG_ industries as Tim writes. The cost of buying those chip writing machines, one supplier in Germany ITIRC, requires massive turnover at any unit cost to be worthwhile. Hence as CPUs and signal processing chips get more capable and complex one needs a bigger market. Even RISC like ARM is growing in size. Hence the steady erosion of architectures and profits outside the dominant one, if any. When did SPARC last make a decent buck ? MIPS ? PA_RISC, Itanic ?

        In aviation Boeing and Airbus have same issue. How many flying cattle trucks does the world need ? Fortunately for Airbus, Boeing have employed, IMHO, a thoroughly modern merkin CEO and board, so they wont be around many more decades. Probably be bought out by Sukhoi around 2050.

        Finally, I regard food miles, buy local food concepts as just another european or merkin greenie protectionist misrepresentation. Fortunately, chocolate and coffee have to be traded to the frozen wastelands of the North where these creatures have a heartland of sorts, so the slogans will never quite ring true.

        1. PowerMan@thinksis

          Re: It was all they could have done*

          Intel has a better chip making technology because of "why"? Both Power8 and Haswell use 22nm. Power8 is delivering 2X the performance in SPECint/fp/jbb and SAP off the top of my head. I recall Power uses something like 15 layers in their manufacturing process compared to something like 7 with Intel. When you stack the highway you get 3D efficiency so if anything I'd say IBM's approach is the better technology even if their fab appears dated by mass production standards.

          1. MadMike

            Re: It was all they could have done*

            PowerMan

            "Power8 is delivering 2X the performance in SPECint/fp/jbb and SAP off the top of my head"

            How about you post links instead of making things up? I always post links. You dont.

      2. Anonymous Coward
        Anonymous Coward

        Re: It was all they could have done*

        Sorry - don't understand this logic at all.....does this mean that by definition if you don't produce your own chips you are out of business? ARM doesn't produce any chips at all, but it seems to be doing okay.

        I think that the point here is that IBM is committed to GF for its chip needs for 10 years. After the end of this period, IBM has the option to go out to the market to see if anyone can do this better/cheaper than GF.

        The point that you have missed is that IBM has just INVESTED $3B in chip R&D, which includes the design of future POWER architectures. It just can't afford to make them itself, so the manufacture is being outsourced to someone who does have the economy of scale.

        The jury is still out on whether POWER will survive, but that's a market discussion rather than "because we no longer make them ourselves"

      3. PowerMan@thinksis

        Re: It was all they could have done*

        More FUD. 10 years is an eternity. Look over the last 10 years just to see what has happened so looking at the future 10 years doesn't impact the future of Power at all. Your "sky is falling" is unfounded, without facts and baseless.

    3. Rol

      Re: It was all they could have done*

      Have you considered option 5?

      Take the R&D for a completely new way of providing cpu grunt and blow the old fab process into an economic backwater. Perhaps in ten years time IBM will be pushing the positronic brain that it has been secretly scheming over for all these years. Or perhaps an optical processor? Either way the fab it had would have nothing of use to the new process and thus be more of an hindrance than an asset.

      I assume IBM is in business to stay in business and therefore will no doubt have thrashed through many of the arguments raised here and more before bringing down the axe. Thing is they are privy to more information than we are and so we can only assume they have come up with a cunning plan...and hopefully Baldrick was not involved.

      Personally I think the processor is on the verge of being completely revolutionised and that is why IBM has been carefully offloading its "antique" technology, to be nothing more than a millstone around the necks of its competitors.

    4. JHC_97

      Re: It was all they could have done*

      Or someone went oh the cloud/<buzzword of the day> is fashonable right now and chip fab doesn't sound sexy. Lets sell it! It frightens me the way we see our business leaders as all knowing and far seeing when in reality they don't have a clue what they are doing.

      We don't know if it was a viable concern, whether the sale had another motive and pointing to it as evidence of the unstoppability of globalization is illogical.

  6. Rol

    Equitable distribution

    It's all well and good rationalising the whole worlds production to meet demand as efficiently as possible, but some countries are not going to get a piece of that pie and further to that, some groups of people will find themselves living in the wrong country for their skill set.

    In the UK we see ourselves as a tertiary economy, we provide services rather than dig the dirt or churn out widgets and that is a precarious existence.

    If we look at the UK's economic advantages, they all focus on one aspect, people, and a very particular kind of person, one who is educated and intelligent. These are the people who run our financial institutions, design new processors or delve deep into the fabric of our universe, without them our country would be uneconomically viable

    The problem with people is they can readily move to the other side of the planet, they are not a thousand acre fabrication plant costing billions, they are probably just the cost of a forty foot steel container, plus shipping.

    Look at the brain drain that started several decades ago and consider the worth these people have added to foreign economies or seen another way, denied our own economy. (Although it must be said they probably had little opportunity to achieve a similar success in this country)

    More importantly, how soon will it be before our only economic asset is equalled and perhaps surpassed by other countries? I'd suggest that time is upon us and the service industries this country has relied on so heavily will soon disappear to India and China, leaving what behind? What kind of product might the UK in a globalised world be excelling at to the point it would be uneconomic to move it elsewhere? Perhaps if quirky eccentricity could be "profitised" we just might have a future

  7. TheOtherHobbes

    >one who is educated and intelligent. These are the people who run our financial institutions

    I think you may have confused 'criminal' for 'educated and intelligent.'

    1. Rol

      Ah yes, I do believe they exist at the union of the Venn diagram (criminal (banker) educated and intelligent)

  8. Anonymous Coward
    Anonymous Coward

    IBM is in crisis, so is HP and Yahoo. There's a connection here I think.

    Oh yes, they all decided that they would hire female CEOs to try a different approach. Working out well huh?

    Next time, hire on merit and stop box ticking?

    1. Anonymous Coward
      Anonymous Coward

      And your pouint is?

      1. All 3 CEOs you are referring to have done a reasonably good job in other roles prior to that in their career.

      2. There are not that many people out there who are brave (or reckless enough) to take on a job like Yahoo or HP. None of those are "easy assignments". IBM is a more complex one, but not an easy one too.

    2. Denarius
      FAIL

      relevance ?

      gender has nothing to do with being blinded by dominant cultural views such as Harvard Business college, trickle down economics and managment myths. The companies are doomed, but I seriously doubt even Dave and Bill could help HP now. Thomas Watson of any mark would be helpless in modern IBM with the stranglehold process has on it. The world and market has changed. No longer near monopolies in IT.

    3. Anonymous Coward
      Anonymous Coward

      Female CEOs

      I'm a recent Big Blue escapee (my choice, not theirs) and while I am no longer a Ginni fan, you're way off base in asserting that gender or box-ticking has anything to do with it. IBM's executives' biggest problem is what we lowly field people used to call "breathing our own exhaust". They hire bright marketing people to craft attractive stories to highlight the positive aspects of their offerings, and they sincerely believe every word of it. And like the "telephone" game, the story gets farther and farther from reality at every retelling. "It's twice as fast as our previous generation" becomes "It's twice as fast as anything on the market today". And then they stare dumbfounded when customers don't buy in droves, go back into their opulent meeting rooms, and breathe some more exhaust. They stopped listening to their field employees years ago.

      Ginni's course was largely cast by her predecessor, and pressure to meet an unrealistic goal. She was chosen on merit, had a very strong history prior to becoming CEO, and was a heck of a lot better choice than the previous heir apparent, who one might recall ended up doing time for really stupid insider trading in which he wasn't trying to make money, he was just trying to impress his (ahem) "girlfriend".

      So fine, damn Ginni for her performance... but not for her gender.

  9. Stern Fenster

    Chip fab, the ultimate selective example?

    Globalisation: yet another one of those words that cover several things, so can be used by different folk to argue in completely different directions. OK, chip fab needs that scale of capital, but that's exceptional (it certainly doesn't apply to farming, nor to most ordinary things). The bulk of complaint against globalisation is focussed on companies who abruptly close down workplaces and leave everyone in the shit merely because there are still some places around the planet where people haven't heard about money yet, so fat profits can be made... for a short time.

    Of course, though it took us a couple of hundred years to organise the collective leverage to get workable wages, it's not going to take these other people anything like that long; after which we'll be looking around going "Oh. Where did all the skills and all the plant go? Why is our culture just flapping about selling insurance policies to itself?" But the guys responsible will have long since vanished with their bonuses by then.

    This is logic/common-sense. No quasi-hippy cliches (yurts, sandals, tree-hugging etc etc ad nauseam) required to subscribe to it.

  10. Joerg

    IBM losing money? Please..managers faking accounting!

    That is the only truth. These IBM managers are thieves and should be put in jail. They stole billions inside the group. The have destroyed the biggest and strongest IT Corporation ever. They kept faking accounting to show that they needed to sell this and that. Nothing was ever true. They were just stealing money and getting bribes from competitors to sell the assets and so on.. It's all dirty business, it's all dirty money..it's a mafia thing!

    If there was any justice bankers and these managers would have been put in jail long time ago. Unfortunately we live in a world with no justice and criminals are allowed to become managers, steal money, steal assets, destroy people lives along with top-notch products.

  11. Buzzword

    So why Starbucks?

    So why on earth do we have global coffee chains? Making a cup of coffee is just as local a service as a haircut; yet our high streets are dominated by national or international chains. Where's the economy of scale in coffee production?

    1. The Axe

      Re: So why Starbucks?

      Starbucks the brand is global, but the actual shops are usually local business, franchises.

      1. Buzzword

        Re: So why Starbucks?

        Ok, so what about Caffe Nero? They aren't a franchise: every store is owned & operated by the company.

    2. Anonymous Coward
      Anonymous Coward

      Re: So why Starbucks?

      Starbucks does not excel at making coffee, and they are downright bad at providing pastries.

      They do excel in providing a consistent, controlled, "nice" emotional experience across space and time; a customized beverage and lousy bakery item is just part of the ritual. (Crucially, the monetized part.) The ritual is very useful to many people who are dealing with stressful situations, such as meeting strangers in strange places, and also restorative for many people whose daily activities are not generally under their control.

      In this light, the large scale has a lot of value. For one, the local coffee shop can only provide this for locals, not even for those who move around town on a daily basis. For another, a great deal of care (based on design and research) goes into evaluating, improving, and refreshing this experience.

      1. LaeMing

        Re: So why Starbucks?

        The economies of scale in these types of business are at the marketing end.

  12. Anomalous Cowshed

    Big Blue

    Big Blue must look to Apple and dream...

    Dreams of shuffling papers, doing a bit of design, marketing...and selling stuff made by a Chinese company and rebranded IBM...Life would be much easier, without all those pesky blue-collar workers, clean rooms, engineers...Just a bit of paperwork, and cash, sweet cash...Mountains of cash. Why can't we also have that?

    1. PowerMan@thinksis

      Re: Big Blue

      What business isn't looking at Apple and their $100B+ cash pile?

  13. rjf

    I think they made the right call... it's call a "fabless" design model, you design and build your chips on other fabs. this gives them freedom. The cost of a current generation fab is about 5 billion dollars.. you need a huge volume to make that worth while and contract manufacture is not what IBM is about.

    as another poster said, the 10 year deal is about supply business to global foundry... there is a reason they had to PAY global foundries to take the fab, they refused to pay for it 6 months ago.

    A few years ago, I saw an article that said there would only be a dozen fabs worldwide in a few years time... split between 2-3 manufacturers. each fab tends to double in cost per generation but more than doubles it's output.. so eventually it's only possible to stay in the game if you have an ever decreasing number of fabs.

    1. Ken Hagan Gold badge

      "A few years ago, I saw an article that said there would only be a dozen fabs worldwide in a few years time... split between 2-3 manufacturers."

      So (with due respect to the suggestion that he probably never made the original quote) Thomas Watson *should* have said, "I think there is a world market for maybe five computers factories.".

  14. Javapapa

    Let's just call them IB

    Pretty soon there won't be any Machines to sell. Serves them right for treating the AS/400 brand with benign neglect.

  15. John Smith 19 Gold badge
    Unhappy

    The problem is not this generation but the *next* generation

    Bottom line, their internal needs were simply not going to cover their costs. Acting as an outsourced supplier to other companies should have covered their costs but it didn't.

    The gorilla in the room is that in a few generations time we will be down to a 1 atom wide transistor.

    And at that point everyone is f**ked.

    1. PowerMan@thinksis

      Re: The problem is not this generation but the *next* generation

      Which is why IBM is investing $3B to cross that canyon. They are already approaching it with a tremendous performance and feature lead over x86. SPARC isn't even in the game. They are going wide and big (ie M7 with 32 cores per socket & 1024 cores per server) that *if* it ever comes to market and *if* it even produces a SAP S&D 2 tier benchmark will be comparable to the 192 core Power8 server. At this point it is about innovation which the people I work with at IBM in Power engineering have it in spades.

      1. MadMike

        Re: The problem is not this generation but the *next* generation

        @PowerMan

        This is silly. You are speculating that a 32-socket SPARC M7 server will be comparable to a 16-socket POWER8 server. How do you know that? Isnt this just wishful thinking, or do you have access to 32-socket SPARC M7 servers and have benchmarked them before they reached the market?

        Besides, informed people say the SPARC M7 which is 4x faster than SPARC M6 - should be more than twice as fast as POWER8. If not, I will despise Oracle for bad engineering. Yes, truly. I mean, SPARC M7 has 10 billion transistors, and POWER8 has half of that - and if SPARC M7 is not much faster than POWER8, then Oracle has failed miserably with their transistor budget. And I despise bad engineering. So, I promise you, I will despise Oracle for SPARC if the M7 is comparable to a POWER8. Promise.

        But, everything points in the direction of SPARC M7 is more than twice as fast than POWER8. For instance, I would like to see one POWER8 cpu doing SQL queries at 120GB/sec - which SPARC M7 can do. How many POWER8 sockets does it take to do 120GB/sec queries? Four? Six? Anyway, hoping for SPARC M7 being as slow as a POWER8 is quite optimistic thinking, dont you think?

        So, there is no way a SPARC M7 with 32-sockets and 64TB RAM will be as slow as a 16-socket POWER8 16TB RAM. The SPARC M7 cpu is at least 2x as fast as a POWER8 cpu. And the SPARC server has twice the amount of sockets. So in total, the SPARC M7 server should be at least 4x faster than the largest POWER8 server.

        Its a far stretch to hope for SPARC M7 server being as slow as a 16-socket POWER8 server. The SPARC M7 monster will crush the POWER8 server. Trample it in the dust.

  16. Squeezer

    The nearest gorilla in the room isn't the technical one (the "1 atom transistor"), it's the rising cost of building the new fabs and manufacturing chips in them, which means that the cost per gate has pretty much stopped falling beyond 28nm -- and the whole electronics industry is based on "more bang for the buck" with each new process node, not "more bang for more bucks". What happens when the entire planet can only afford one fab?

    Economics will kill Moore's law before physics does...

  17. Anonymous Coward
    Anonymous Coward

    long article completely misses the point

    Globalization has nothing to do with IBM's chip-making unit becoming a loss leader. It has everything to do with something much simpler: a competitor's chip - Intel - is cheaper and better than POWER, and nother competitor's operating system - Linux - is cheaper and better than AIX.

    IBM knew this for a long time. They've been trying to play the "Linux on POWER" game to save the POWER architecture. Turns out it's not so simple. The product - Linux on POWER doesn't really compete yet, it's not up-to-par with Linux on Intel. And even if it were, the hardware is much too expensive.

    What does this have to do with Globalization? Nothing. Economic competition and obsoletion - which is what happened to POWER and AIX - exists with or without Globalization.

    But, when the only tool Worstall has at his disposal is a Really Big Hammer, everything's a nail.

  18. Stephen McLaughlin

    IBM and Oracle

    I think IBM would have been better off had they embraced Oracle on Open Systems when the p4-690s were introduced. At the time our datacenter had loads of Sun E10Ks and we were looking to do a tech refresh. The performance numbers on the p4-690 in testing were amazing, they blew the doors off the Sun E25K servers. However, though Oracle did run on the IBMs, it was very, very buggy. IBM was staunchly behind DB2. So we upgraded everything to Sun E25Ks. Wonders what if IBM pushed Oracle instead of DB2 on Open Systems, how things might be different today.

    1. PowerMan@thinksis

      Re: IBM and Oracle

      IBM is Oracles largest customers in terms of the platforms their software runs on. Also, far more Oracle runs on Power than DB2. So, something changed and happened from the P4 days. As a side note. I think Oracle should actually abandon (by selling Solaris & SPARC) to focus on it's software stack optimizing it to run great on every platform. Let it exploit the best features of Power, Intel, AMD, ARM, Itanium, MIPS, whatever runs it. They are slowly trying to lock themselves out by tying everything together.

  19. Torque76

    My story as a contractor at IBM – I worked for IBM eighteen months ago and my contract was cut short by IBM two months early.

    I started another 6 month, hourly paid, PAYG contract with IBM through an agency recently but decided to leave early after working 68 hours (9 working days at 7.5hrs) with 2 days enforced leave, no email or intranet for the first week.

    I have been paid for 13 hours only (the hours worked on the first two days of the contract). When I queried I was referred to the:

    “No Compensation” clauses on Page 7, this indicates that “Should the Contractor for any reason, not complete the term of the assignment and have been engaged by agency for more than 10 working days, then agency shall not be obliged to pay any fee to the Contractor for the final 10 days worked by the Contractor”.

    I hope Global Foundries spend my salary wisely :(

    1. Anonymous Coward
      Anonymous Coward

      Surely it is the agency that has kept your money, IBM pays the agency for a contractor to turn up, first it was you and now it is A.N.Other.

      That you didn't read the contract between you and the agency is no concern of IBM or GF.

      Why didn't you read a page that has a heading 'No Compensation Clauses', that is just craziness on your part, the first bits I read are 'How much I will be paid' followed by their 'Get out of paying' pages followed by the description of how many pounds of flesh/numbers of unborn offspring they will require in return.

      1. Torque76

        "Surely it is the agency that has kept your money, IBM pays the agency for a contractor to turn up, first it was you and now it is A.N.Other."

        If I can suggest, given the reading skills you demonstrate later in your post you use your 'reading between the lines' skills at this point here.

        Also at my level, which is a temp employee, I don't get to enter into contract negotiations sadly and I have to accept the contract as it is if I want the role.

        I am glad you don't have that issue and given you are doing so well I am puzzled why you felt the need to post a comment that attempts to diminish me and absolve IBM? They are a $168Bn multinational organisation with plenty of resources as is.

  20. Paul J Turner

    Don't hold your breath

    "IBM's proposed sale of its semiconductor operations to GlobalFoundries will be scrutinized by the Committee on Foreign Investment in the U.S. because GlobalFoundries is owned by investors in the United Arab Emirates and IBM supplies chips to the Defense Department." - http://www.computerworld.com/article/2837426/ibms-chip-business-sale-gets-national-security-scrutiny.html

  21. Anonymous Coward
    Anonymous Coward

    Nuts!

    IBM bought borrowed entirely too much money to buy its own stock back. It should have ploughed all that money back into its research and fabs. Now it is going down the drain fast. Is there a word for "brillian dumbass?" Yes, it's IBM!

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like