no servers, no hardware, no chips ...
... then what does IBM do, exactly?
IBM is rumored to have retained Goldman Sachs to help it put a price on its semiconductor division. The news was reported by the Financial Times on Thursday afternoon, California time, and follows IBM's long-expected announcement of the sale of its server division to Asian electronics giant Lenovo. At the time of writing …
IBM is going to exit Unix first, and later Mainframes. Where is the POWER roadmap ? Why dont IBM talk about their POWER cpus? IBM has slowed down on AIX development pace by shifting resources to Linux. IBM is betting more on Linux than AIX these days. And IBM has confirmed that AIX is going to be killed, and replaced by Linux:
POWER is loosing its edge, and transforming into low margin. POWER6 costed 10x more than x86 and were much faster than x86. POWER7 cost 3x more than x86 and is 10% faster. POWER8 will be the last fanfare from IBM, trying to make POWER high margin business again. But will fail as x86 is quickly becoming more and more powerful, rivaling POWER cpus. IBM does exit low margin business, so POWER will be killed off, probably after POWER8.
When AIX and POWER has been killed, IBM will turn to Mainframes. But that will take some time because Mainframes are still high margin. But no new customers buy Mainframes, only old locked in customers upgrade to new Mainframes. New customers bet on x86 and Linux/Unix. As old customers are slowly deploying new business on x86, the Mainframes will only be used for legacy systems. And when the legacy systems have been replaced by x86, they will not buy new Mainframes any more. The number of customers are diminishing, there are less every year. The customer base is not growing, it is shrinking. If we extrapolate the shrinking market, after some time, when very few customers are left, there will be no more Mainframes, IBM will only support the last dinosaurs, but not will release new models. And then Mainframes will die. Just like Itanium has been effectively killed off. And POWER soon.
You better start learning Linux and abandon AIX, which is next in line to die. Mainframes are in line much later (you do agree that a shrinking market will ineviatable lead to products being killed? When they can no longer carry their cost?).
I've read elsewhere that IBM is doing all of this bloodletting simply to get their earnings per share to reach $20 by some target date. I guess my question is what will be left of the company after that's done.
The thing that's a little scary about all this "let's get rid of those pesky Machines our International Business division manufactures" stuff is that they're really betting that software and services will be more profitable. IBM has been known in the past to be right about some of its predictions, like the commoditization of the PC marketplace. But they've been known to be wrong too.
The other thing that might happen is that their Research division will be clobbered because there will be no physical products or manufacturing processes to improve. And if they farm out semiconductors, they lose control of their ability to fab mainframe and POWER chips themselves. IBM's one of the last companies with enough resources and (I thought) forethought to support basic research.
I don't know, core competencies this, brave new world that...but divesting yourself of any physical products just to chase software profits doesn't seem like a good long term strategy. I guess I'm an old fogey, but I do remember when IBM was completely unstoppable and they employed way more people than they do now. It's going to be very weird having the next generation not seeing anything with an IBM logo on it and associating them with crappy outsourcing and WebSphere.
All to please Wall Street, one of the most toxic forces in the economy. That was the entire impetus to sell the X server line. So they will make their numbers then what? Who is going to fab chips for Z/OS systems with their low volumes? Getting rid of drives and PCs was a good idea. The X86 servers not so much, there were opportunities for differentiation. IF storage goes next, then chips, IBM will fade away over time.
Other reports say they are planning to keep chip design inhouse so this sounds like they pulling an ARM and going fabless.
Since the Cell business died I can imagine their foundry business volumes aren't what they used to be which makes it expensive to run a fab primarily just for their own Mainframe and Midrange Power chips.
Given the huge costs involved in keeping the fab processes competitive with Intel, Samsung, TSMC, UMC et al moving to 22nm and lower I imagine it wasn't hard for Ginni to decide it was no longer core business and was ready for offloading. And yes it will no doubt be welcomed by Wall St.
Hopefully they will keep the pure Research going. Who knows, maybe they are offloading now and preparing to return with the next big technological breakthrough
Nope, no master plan. Just far too many incompetent idiots in positions of power who have neither the desire nor ability to see more than three months ahead, and who genuinely believe that companies can be managed with a few spreadsheets.
If I was making a medium to long-term investment of any size, I wouldn't touch an American company with a barge pole.
> Don't these idiots *ever* learn from history?
That's not how it works. Decisions are ALWAYS local. They become global (to the detriment of most) only if a specific lever is legislated into existence to satisfy these local interests.
More on this in The Great Deformation.
IBM are very good at getting out of markets when it's the right time to do so.
They don't make typewriters any more
They don't make photocopiers any more
No more printers
No more workstations or laptops
No more commodity servers
No more consumer grade stuff in general
All of these markets (and more) had the ability to drag IBM down and break them, they got out before they became a millsone round the company neck. There are many other companies which didn't get out of markets, where are Novell now, for instance? Doggedly sticking with your initial reason for being and not changing with the changing marketplace is not good for company longevity.
"IBM are very good at getting out of markets when it's the right time to do so."
Really? So how do other companies manage to make a profit selling printers, laptops, workstations etc but IBM can't?
More like they got out because they didn't have the in house talent to drag those sectors up.
@Boltar - IBM are not in the pile it high sell it cheap business, they're into high quality high margin. All the markets the left, they did so because they were becoming commodity.
Suggesting that IBM don't have the required in house talent suggests that you have no experience of IBM in house.
"All the markets the left, they did so because they were becoming commodity."
So IBM trying to pitch itself as the Rolls Royce of ... err , what exactly? Software? GIve me a break, I wouldn't touch AIX, DB2 or anything from Lotus with a Bargepole. It was the hardware that made IBM, not their software and CERTAINLY not their appalling "services" (I use that term loosely). If their dump all their hardware divisions, even if they are "commodity" , then thats the end of IBM as a serious player.
"Suggesting that IBM don't have the required in house talent suggests that you have no experience of IBM in house."
Admittedly I don't, but from the outside, it looks like IBM is trying to become another Accenture, BCG, etc. I find it hard to believe that the market can absorb yet another snake-oil peddling "consulting firm" as well as losing another hardware manufacturer. That's one thing I worry about in the PC and server markets -- fewer companies fighting over lower margins means they'll just stop doing anything cool and figure out the best way to squeeze a $299 price point out of a 4-socket Xeon box with 2 TB of RAM.
Watson is really cool technology, but how is IBM going to make money selling "decision support" and other stuff? More importantly, where will the next hardware innovations come from? Startups are cool and sexy, but sometimes you need a company as massive as an IBM to invest the kind of money needed to make huge changes. Oh, and keep half a million people employed...
Sure, IBM will get out of all markets, that would be right thing to do. If you dont take any risks, there will be no rewards. Sell everything.
It reminds me of when IBM sold the hard disk division many years ago. Why? Because IBM management had found out that pretty soon, a single future huge enormous 1TB disk would suffice for a whole company. IBM did not realize that the more disk you have, the more data you store. Just like cpu power, or RAM ("640KB ought to be enough for anyone"). The more resources you have, the more you spend. Storage is exploding and last year mankind stored more data than ever. Storage is growing, not dying. Wrong conclusion by IBM. Again.
HP tried to sell off it's PC division recently by new CEO Meg Whitman, because PC division were low margin. But Whitman changed her mind because, even though PC is low margin, it brought in $4 billion annually. And that is not chicken shit. If you have a low margin, but sell very much of it - it will give a good revenue, no matter how low margin it is.
IBM is exiting all hardware business, and has been doing it for decades. They have a long term strategy to slowly transform into software and services company. Is it why IBM has not release POWER road maps? POWER is next in line to be killed?
I call Bullshit. This is probably IBM's PR Department.
Dicthing the chips division - and the R&D division most likely after that - has nothing to do with consumer grade stuff. Selling the server biz to Lenovo has nothing to do with consumer grade stuff. Consumers don't buy servers or Power chips - in case you didn't get the memo.
Speaking of your "not changing with the changing marketplace" (you learned that phrase in Business School): what's left of IBM then? AS/400? Mainframes? Tape? (there's future in tape as we all know). Price Waterhouse? Talk about "not changing with the changing marketplace".
The New IBM - The Accounting Company. Next year I'll do my taxes at IBM. HR Block is a tough competitor, I hear.
Obviously IBM do not read the British Press and so have not heard about the value for money achieved for the British taxpayer by taking 'advice' from GS and UBS when Royal Mail was sold for 40% less than market value, despite the fact that many banks who failed to be awarded the contract quoted perfectly realistic valuations. I myself tried to obtain some shares, but failed. It is not every day that a sound business offering a 7% yield at the offer price comes to market.
I think it is easy if you look some raw numbers and isolate them from the rest of the universe, to make financial decisions. Also it is fairly easy to make financial decisions if you do not have proper insight into technology, engineering and research.
So looking at how much money the semiconductors division is making, it seems to be a good sense to sell the division. I could not argue that. However, looking closer - and understanding better the technology and how each division of IBM makes money, suddenly such a decision is not as sexy...
I remember how IBM supporters treated Sun too. You could often see small variations of the same post in Sun related articles here:
"I work at a big investment bank / telco company / etc and I just love SPARC and Solaris, but sadly our benchmarks show that POWER is much faster and much cheaper, so we are migrating to POWER now because SPARC has no future."
One time he even worked at a nuclear plant I think? There were small variations, I can't remember. I remember one post saying that they migrated off SPARC to POWER7 - at the same time the POWER7 cpu were announced, and no servers existed yet. IBM supporters took every chance to FUD and lie about how bad SPARC was. They did not mind making up numbers and use cases.
They also attacked HP, I remember people writing they loved HP-UX but it was too unstable so they migrated to POWER. HP-UX is arguably the most stable Unix there is. What crazy posts the IBM supporters wrote. I remember one prolific poster here, I showed him benchmarks where SPARC T2+ had more throughput than IBM POWER6 to which he rejected the benchmarks on the ground of "throughput doesnt matter, latency is more important". Later I showed him a benchmark where SPARC had lower latency but POWER6 had higher throughput to which he rejected the benchmark on guess what ground? "Latency does not matter, throughput is more important" !!! Some would call him a bit biased. Or, when someone called him ugly names to which he got upset "you should not call someone names, etc" - forgetting that he used to call me ugly names quite frequently. Yes, some would call him a bit biased. You can not discuss with such persons, if you say A, he will say B. If you say B, he will say A.
Where are Jesper Frimann, Matt Bryant and that other IBM supporter... forgot his name.... was it something with Alice?
Biting the hand that feeds IT © 1998–2019