Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'

Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'

Summary: Oracle didn't meet expectations for its hardware revenue stream and Larry Ellison wasn't happy about fielding questions.

TOPICS: Oracle, Banking

When Oracle reported its first quarter earnings Tuesday, there was only one major sore spot: hardware. Naturally, investors honed Oracle's declining hardware revenue and started asking questions.

See also: Oracle's Q1 strong, but hardware sales lag; New SPARC chip on tap.

The company line is that Oracle is transitioning from low-margin---more like no-margin server sales---to higher end systems like Exadata.

Oracle president Mark Hurd said the company is not as focused on the hardware growth rate so much as "growing the right things in the product line." Specifically, Hurd said Oracle wants to focus on "the things that are attributable to long term profitability, long term positioning with the customer (and) long term positioning in the datacenter." Oracle is more concerned with pushing its own Exadata, Exalogic, and the core SPARC product lines.

"I think you are seeing some of the momentum, and if you see the market share charts, they send tend to show you the facts that we are gaining share in that core part of the product line," said Hurd, trying to reassure investors.

However, CEO Larry Ellison didn't mince any words. Ellison sounded annoyed with analysts who questioned Oracle's hardware strategy.

Let me add a little bit to that just so we're really clear. I don't care if our commodity X86 business goes to zero. We don't make any money selling those things. We have no is interest in selling other people's IP. X86 includes Intel IB, Microsoft IB. We don't make money selling that. Sun sold that stuff and we are phasing out that business. We have no interest in it whatsoever. We have interest in selling systems that include our IP. That's how we're going to drive the profitability of our overall hardware business. Eventually I think actually in fairly short order our engineered systems will be -- are growing at such a high rate that the overall hardware business top line will grow also. But what's really important is to continuously grow our margins in the hardware business and our profitability in the hardware business so we can meet [Oracle CFO Safra Catz's] goal, which is getting back to our pre-Sun acquisition overall margins.

Perhaps Ellison was bitter over having to show up in court to talk Java settlements with Google CEO Larry Page all day. If you had to sit in court for a few days, you might be a bit grumpy too.

Nevertheless, Oracle is most likely making the right move with its hardware business. It's just common sense: If something isn't making you money you're better off cutting it sooner than later.


Topics: Oracle, Banking

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Sometimes, sometimes not.

    If something isnt making you money youre better off cutting it sooner than later.<br><br>There can be (and often are) complementarities between unprofitable and profitable businesses. Where that's the case, it can make sense to keep even loss-making businesses that have little prospect of ever becoming profitable. This is why good managers don't listen to myopic investors without industry expertise who demand that they sell off all of their unprofitable businesses. It boosts profit in the short run, in exchange for a loss of strategic advantage, which leads to lower profit in the longer run.<br><br>In Oracle's case, firms interested in buying both x86 and Risc servers might prefer to do business with a vendor offering both -- or even with a vendor who can offer everything from large servers to desktop/notebook PCs. HP were the primary firm able to do that, which is why shedding their PC business, which takes away one of their few advantages over competitors like IBM, might not be a good move (even if the business doesn't generate any profit at all).
    • RE: Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'

      First... HP does not sell their own hardware... they sell Intel hardware
      Second... Companies don't want to do business with a vendor offering both... they prefer that companies bid against each other... for the various pieces.
      Third, I don't agree that synergies will generate longer term profits... IBM's profits have grown, big time, since they shed their PC division many years ago. Companies have discovered that non-profitting units were draining the company of needed capital. Capital that can be used in R&D to keep the companies profits long term.

      HP is in a mess on many fronts...
      Their hardware flagship, Itanium, was faltering even before MS, Redhat and Oracle abandoned development on it. Because they didn't develop an x86 offering of HP/UX, it is now in trouble. Their service division has been decimated by layoffs and is demoralized. Their tablets have been put on the shelf... WebOS was an also ran, when they bought Palm... HP's pride maintained their involvement in the PC market, too long, but THEIR beancounters sacrificed the rest of the company for it. HP didn't abandon beancounting, Mark Hurd was known for it when he ran HP.
      • IBM continue to sell Microsoft/Intel IP

        @ scotth_z<br><br>First, IBM stopped selling desktop systems, not x86 or Windows systems (i.e. Intel/AMD and Microsoft IP). To this day, IBM are a leading vendor of x86 servers and sell a huge number of them, including machines running Windows Server. They clearly understand that many buyers do in fact prefer to do business with a vendor offering a broad range of server solutions, so your argument there goes out the window.<br><br>Second, HP are indeed falling apart, and with their many weaknesses, one of the few advantages they have over IBM and Dell is the ability to offer everything from notebooks to 'big iron'. If they lose that, they'll be even less able to compete with IBM (who offer a broad range from entry-level x86 servers to mainframes) from above, and Dell (who offer a broad range from notebooks to large x86 servers) from below.<br><br>Third, a low- or zero-margin business does not consume capital. A loss-making business can, unless there are synergies leading to sufficiently higher profits in other businesses, in which case it doesn't either. That's why decisions of which businesses to keep and which to drop are best made by managers who understand the synergies involved, and not by myopic investors who just want a short-term capital gain.
    • Don't confuse sales with support ...


      I *think* that what Ellison is referring to is the packaging and sale of commodity products. I *think* that he is not intending to stop offering *support* for commodity products. And that is an important differentiation. Why should Oracle repackage and resell Intel hardware and MS software? Is the customer going to care whether Intel and MS products get superficially rebranded with an Oracle label? The customer can buy their own Wintel products and Oracle, I am sure, will still be happy to load and configure an Oracle product on top of that hardware, for a price, of course, and also offer support for that Oracle product. It is simply that going forward, IF the customer wants an *integrated* product, their only choice will be a *fully* integrated product with Oracle hardware and Oracle software through and through. Otherwise they will have to deal with Oracle as one vendor among many. What's wrong with that?
      George Mitchell
      • Ellison

        @ George Mitchell<br><br>What Ellison is saying is that he doesn't care how many commodity Intel servers Oracle sell. Neither would I in his place. That is absolutely not the same thing as 'cutting' the Intel server business.

        Edit: I had missed that Ellison actually said Oracle are 'phasing out' the x86 server business, so I was wrong there. We'll see how that works out, but the leader in the business (IBM) remain heavily involved in the x86 server business.
    • RE: Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'

      I quite clearly stated PC... and you quite clearly stated..
      "large servers to desktop/notebook PCs"
      WHAT part of 'desktop/notebook PCs' did you not understand in your own diatribe and/or what part of PC did you misinterpret as x86?

      You also acknowledge that IBM is doing quite well and that HP is quite clearly struggling... I think you made my point that shedding a PC business is clearly not a handicap in services or in being in the IT business.

      As for the capital expediture... HP spend $25B to buy Compaq...I think that's a lot of capital....don't you? That's a lot of money that they could have spent to 1... make their high end servers more viable... 2... have spend to make their tablets leading edge.. rather than watching apple take the market...

      and FYI... all these decisions ARE being made by managers...not by the beancounters...
      Ellison and Palmisano are company men..
      33 and 38 years with their respective companies...
      Fiorina had only been with HP for months prior to her tenure as CEO... Hurd and Apotheker were outsiders... and rumors are that Whitman(outsider, now on the board) may be taking over.
      So which companies are you referring to as 'good managers who don't listen to myopic investors'?
      • Selective quotation doesn't work when the original comment is right above.

        @ scotth_z

        My comment on PCs was clearly parenthetic, as anyone reading the comment can see. The focus of my comment was x86 servers, which is also the topic of this article. Selective and dishonest quotation won't help you when the original comment is right above yours.

        The issue here is x86 servers, i.e. Intel and Microsoft IP. That's what Ellison was talking about and that's what I was writing about. Your comment on IBM is therefore at best irrelevant (referring to desktop PCs when the topic here is selling Intel and Microsoft IP), and at worst wrong (confusing IBM's PC division with selling Intel and Microsoft IP in servers)
      • Myopic investors

        @ scotth_z

        My comment on myopic investors was a general response to the author's suggestion that: "If something isn???t making you money you???re better off cutting it sooner than later." I even included the quote at the top of my post!

        You really ought to try reading posts as they're written, rather than trying to read between the lines and imagining they say all sorts of things they don't.
      • RE: Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'


        You're being parenthetical is BS, since you mentioned IBM in your comparison. So either
        you really don't know anything and only later discovered that IBM sold x86 and you were incorrect in your assessment... or you just confused yourself... or you are just compounding your incompetence by CLAIMING you were being parenthetical. So which is it.. you meant desktops and brought in IBM and are wrong in your assessment since IBM is doing ALOT better than HP in services or you made a mistake by including IBM and are now trying to lie your way out by claiming you are parenthetical.
      • RE: Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'


        Compaq was more than deskside PCs. HP got Proliant (and threw away their netserver line which was vastly inferior) and both DEC and Tandem assets. Sun x86-based servers cannot hold a candle to Proliant--most particularly the x4800 versus the DL980.

        HP should have cut over their BCS (Business Critical System group) offerings over to Xeon in time to catch the Nehalem EX (Xeon 7500) train but stuck with Itanium. HP established a really tight relationship with Oracle which culminated in being the first platform to earn a port of Oracle Exadata (the HP Oracle Database Machine). HP put *a lot* of company resources into making that Exadata go-to-market effort. That leads me to remind, or inform, people that Exadata X2 is just commodity x86 server hardware. Exadata Storage Server software is actually fully functional running in a Virtual Machine. That's how non-hardwired it is. It only needs x86. That's why the cut from HP Proliant to Sun was done so fast. It's just x86 software with a port to the BMC (Baseboard Management Controller).

        Most people don't have more belief than fact on this particular topic.
    • RE: Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'

      When you mention reacting to myopic investor... you were clearly referring to Ellison... the guy who founded Oracle... Also when you were referring to 'good manager'... again who were you referring to', surely not Ellison... You cited HP as your example...the company that has 4 different permanent CEOs in the last 12 years... all of them outsiders...(aka.. the ones who would MOST react to myopic investors).

      Read your own comments and ask what conclusion anyone can come to.. Your arguement that Ellison is reacting to myopic investors is not credible.

      My comments were clearly aimed at you... not at the article... YOU were the one who included PC and desktops as part of the conversation.
      You were the one that brought HP into the conversation as a company that should be emulated...your example is flawed... and you later admit it. If you can't cite a company that offers a full offering.. that is doing well... please cite it.. It clearly isn't HP. So even your parenthetical arguement is incorrect.
      The real question to ask is who IS reacting to myopic investors and who is doing it for a pure business sense. IBM spun off their PC division when people were questioning WHY. They sold it off not because the margins were bad, but that they ANTICIPATED that they would become bad. It clearly hasn't hurt IBM.
      Companies in an ever changing market such as IT, constantly need to re-invest themselves... Just remember that IBM was at one point the dominant typewriter company. They spun off their typewriter division in 1991... when Lexmark was created. Do you think THAT was a bad move? A move like that doesn't dovetail with your philosophy of maintaining a non-profit division, does it?
      To be fair, I agree with you that Oracle shouldn't leave the x86 market, but for different reasons... Clearly the x64 market is in serious decline and would leave Oracle with no hardware stronghold. Of course, it wouldn't surprised me ,if Oracle leaves the server hardware market completely in 3-5 years and that their dropping of x86 may be a precursor.
      • It's like talking to a brick wall.

        @ scotth_z

        I absolutely was not referring to Ellison. The first paragraph of my comment was a general response to the author's suggestion that any business that doesn't make money should be 'cut'. That's why I started my second paragraph with, 'In Oracle's case' -- moving from the general to the specific. I apologise if discussing two different issues in the same post confused you, but I should have thought it was obvious to anyone with normal mental abilities.

        I suppose I shouldn't be surprised that you're confused about IBM's exit from the PC industry too. IBM's PC division was loss-making when it was sold to Lenovo, and having ceased to be either a technology leader or a cost leader, IBM had fallen to third place in PC market share. In contrast, HP's Personal Systems Group generates a steady profit (but with low-margin -- about 5 per cent), and accounted for nearly 18 per cent of HP's overall profit in 2010. Moreover, HP have become the leading seller of PCs. It's a completely different situation.

        Mind you, since you're very easily confused, I'll point out that I am not making the argument that HP should keep their PC division, and nor was I ever. I'm simply saying that the fact that it's a low margin business does not imply that disposing of it is the optimal strategy. Maybe it is and maybe it isn't. The same applies to Oracle's x86 server business. My point is that low margins or even losses are not sufficient to make a disposal optimal. Full stop. If you're unable to grasp that simple, straightforward point, there's nothing I can to to make it simpler for you.

        Not surprisingly, you're confused about the x86/x84 server market as well. According to IDC, x86/x64 server revenue in 2011Q2 was up more than 15 per cent -- the seventh consecutive quarter of revenue growth. If you think 15 per cent growth is 'serious decline', you're using very, shall we say, 'creative' definitions. I assume it was just more confusion on your part, rather than any deliberate attempt at deception.

        If it makes you feel better to pretend I wrote things that I didn't, so that you can set up Aunt Sallies and then knock them over again, whilst at the same time making things up to support your incoherent arguments, go right ahead. I've no interest in playing along, however, so I'll leave you to it.
      • RE: Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'


        "Clearly the x64 market is in serious decline"

        Your nomenclature is off base. You probably meant IA64 (Itanium). The industry has settled on "x64" as shorthand for AMD64 (x86_64) and EM64T (Intel).

    • RE: Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'

      So when you mention desktops YOU really meant x86 in the article.... but when you incoherently rant about managers... even though the article is about ELLISON, not management decisions, you were really talking about OTHER management decisions... yeah clear as mud. Who do you think the comment was about... the article is about Ellison, not some other management decision/team? Also, you are showing your intellectual dishonesty by lumping x86 and x64 sales together to try and claim that x64 is doing great. You offer the same mix and match logic, by saying desktop, but trying to mean x86. Sorry... x64 is in trouble especially in HP land. You can whistle all you want in the cemetery... but HP is in deep trouble...
      Their flagship server is now poison in the industry... MS, RedHat and Oracle don't want anything to do with Itanium.

      Perhaps you can explain these numbers, HP fanboy..;siu-container

      FYI... in the article... Itanium is that section labeled BCS.

      FYI.. Losing market share is not the same as not being profitable and IBM wasn't the only manufacturer to leave the PC market... IBM was the #1 typewriter manufacturer when they bolted that industry so being #1 is not a deterrent to leaving an industry behind. Let's understand that HP bought their way into #1... Compaq was #1 until 2001 when HP bought them... and was #2 when they completed the sale. So HP could most certainly sell their PC division... especially if they don't have an answer for the up and coming Acer. The question to ask is will they get anywhere close to the money they spent to get Compaq($25B) and WHO.

      As for whether to keep a low margin division or not... It depends on the business... HP has been trying to buy their way into every IT niche... When a company takes their eye off their higher margin divisions... such as services... such as high end servers, such as software... YEAH it makes absolute sense to spin off the low margin units. An IBM VP said it best, about a year ago... "Volume without margin is just a distraction" Synergy loses it's value when your business is suffering and HP is clearly in trouble.
  • RE: Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'

    You do not make money with Hardware (PC's, Laptops 3% profit) any more. You make money to sell solutions, consulting, service.
    The best example is IBM , the hardware is just a part of the solution the will sell you. I thinks and the numbers are showing it, Oracle made a mistake to by the money loosing SUN company.
    old mainframer
    • RE: Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'

      @old mainframer
      IBM 2010 annual report
      % of revenue..
      software 44%
      services 39%
      financing 9%
      hardware 8%

      gross margin
      software 86.9%
      global financial 51.3%
      technology services 34.7%
      systems and technology 38.5%
      global business services 28.3%
  • Dell?

    IBM, HP, now Oracle. Nobody wants to make and sell these things; there's no money in them. It just ties up capital in inventory and receivables with no return. Who needs that? At least money market funds make 1%.
    Robert Hahn
    • No.

      @ Robert Hahn

      You're confusing two different markets. Ellison is talking about x86 servers (what he calls Intel and Microsoft IP), not desktop PCs -- Oracle aren't even in the desktop PC business, and never have been.

      IBM continue to sell enormous numbers of x86 servers, including Windows servers, and HP never suggested they were considering disposing of their x86 server business. If exiting the x86 server market were a good idea, you can be certain that IBM would have done it.

      Looking at the overall market, Intel servers continue to gain market share, year after year, against non-Intel servers. On the software side Windows Server also continues to steadily gain market share, even when measured by server hardware revenue (which gives an advantage to operating systems that run on expensive non-Intel hardware).
  • So Basically Sun Is All Down The Toilet Now

    So every single thing Oracle acquired as part of its Sun purchase has now turned to excrement. All the open-source software projects were destroyed and their communities driven to forks elsewhere. It looked like the hardware business might have been worth something, but now it's clear that's rubbish as well.

    So everything Sun achieved is now history, thanks to Larry Ellison.
    • RE: Oracle's Ellison: 'I don't care if commodity x86 business goes to zero'


      Agreed. Oracle has tried hardware several times and they've failed each time they just don't get that hardware has a lower profit margin than software