back to article Cloud computing aka 'The future is trying to KILL YOU'

What do all ailing enterprise IT companies have in common? Trouble in their core businesses due to the rise of cloud computing. The repercussions that the technology is having on the IT business are all around us, and its effects on the industry are as inevitable as gravity on a dropped bowling ball. Cloud computing's rise …

COMMENTS

This topic is closed for new posts.
  1. T. F. M. Reader

    The cycle of change spins ever faster... Really?

    Gliding over the dubious validity of overall comparisons between VMware, Hadoop, and NOSQL, let's take the statements in the article at face value.

    So, VMware didn't have a serious effect on the industry for a decade? [I'd beg to differ[1], but I'll accept the statement for commentarding purposes]. But Hadoop "started to cause change" after 7 years (2005 to 2012, according to the article), and NOSQL "already having an effect" also after 7 years (2007 to, presumably, 2014)?

    This does not show any significant acceleration. On the contrary, the timescales look very similar to me: 7 years - with "started" and "already" qualifiers - against 10 (or, arguably, quite a bit less[1])? Meh...

    NB: The above does not, by itself, invalidate other main points of the article. But this particular argument does not hold water, IMHO.

    [1] VMware had a very significant impact several years before 2008. From personal recollections, not only was it widely used for workstation virtualization by 2000-2001 (x86 *servers* were not as dominant then as they are now, btw), but starting from about 2004-2006 VMware was a really major platform for server and networking companies on the supply side, and (at least) big banks on the demand side (see also below). EMC bought it for $625M in 2004 - its impact had to be pretty obvious at the time (that's just 5 years after the first product release).

    To emphasize the dubiousness of the article's comparison, VMware got a real boost after Intel and AMD built virtualization support into x86 (starting from 2006). This helped VMware win over paravirtualization (e.g., Xen, which is still kicking - think AWS and Citrix - but no longer has the performance advantages of the olden days).

    Neither Hadoop nor NOSQL needed this kind of CPU redesign to take off. And still their industry penetration timescale is no faster. I would also argue VMware's impact is a lot wider - Hadoop and NOSQL are very significant niches, but niches nonetheless in comparison. Arguably (yes, one can argue both ways, so don't start), big banks alone were such a niche for VMware before 2008, comparable in scale to big data today.

    A more direct comparison to VMware may be provided by KVM, which is already widely used in the Cloud even though its first *stable* release was just over 18 month ago. However, even KVM: a) was ready enough for Red Hat to buy Qumranet back in 2008 (and leveraged the pre-existing QEMU); b) didn't need to wait for CPU support, either, which helped; c) never had to fight for the basic virtualization business case as the pioneers - VMware and Xen - had won that battle several years earlier.

    1. Anonymous Coward
      Anonymous Coward

      Re: The cycle of change spins ever faster... Really?

      "This helped VMware win over paravirtualization (e.g., Xen, which is still kicking - think AWS and Citrix - but no longer has the performance advantages of the olden days)."

      I think that statement is a bit misleading. "VMware" which is a company, did not "win" over "paravirtualization", which is a technology that is in some cases has better performance than CPU level virtualisation.

      Oh, are we talking about Windows? I thought we were talking about servers. (snickers and leaves)

  2. Destroy All Monsters Silver badge
    Holmes

    Numbers, please!

    Show me the ROIs which are currently being rendered "irrelevant" by deformed markets and Keynesians policies of the worst sort.

    Show me that all of this won't come thumbling down with a fat case of Cloud Gout when:

    0) China implodes as its housing bubble bursts, taking "cheap asian manufacturers" iith it.

    1) The US implodes in an accelerated and decidely deeper version of 2007 with no way to kick the can further down the road via TARPery, Bailouts, Bailins or Quantitative Belching.

    2) Large numbers of feelers into "new technologies" are being pulled back as cash flows driven by cheap money wither. Crazy-horse companies collapse while those with ways to still pay the bills retrench to stable platforms and actually try to finish and stabilize projects for once.

    3) The cloud turns out to have dotcom levels of overcapacity and hype as customers pull out or die off, leaving large fixed assets that are rapidly deprecating against balance sheets that are already redlined.

    1. Don Jefe

      Re: Numbers, please!

      I'm always surprised when a dotcom v1 hype train gets scads of attention, but ignores the fact that businesses who actually use the technologies are cruising right along. This is, effectively a Snap-On vs Mac Tools issue: It doesn't matter who makes the tools you're using if you're getting the job done and making money.

      The tool companies will adapt, or they won't (considering 'adaptors' is what every single 'dying' company on this list used to call entire categories of their offerings, it's not a positive sign if adapting is beyond them). If tool companies, which are all high profit operations, choose to give all those profits to shareholders instead of into R&D and financial firewalls then I can't have a lot of sympathy for them.

      Others will step in and fill the gaps. The world always needs toolmakers.

      1. Destroy All Monsters Silver badge

        Re: Numbers, please!

        Some numbers, actually!

        The Brutal, Beneath-the-Surface, Slo-Mo Crash of Stocks

        One place to find some of them grouped together is the Cloud Index maintained by VC firm Bessemer Venture Partners. There are 37 publicly traded “cloud” companies in the index. The list below shows the top 23 by market cap, ranging from Salesforce with $31 billion in market cap to LogMeIn with just under $1 billion. The average decline from their individual 52-week highs is 41.1%. A serious crash.

        Image

  3. (AMPC) Anonymous and mostly paranoid coward
    Mushroom

    Of course the cloud will eat your lunch, time for a new diet

    If you work from the premise that industrial evolution inevitably tends towards a more efficient system of production (a bit Darwinian, yes), many of the author's statements make complete sense.

    Firstly, any startup business which needs IT or data center services will naturally gravitate towards cloud based offerings. Why? If investors must choose between building a million dollar computer site and a no capex, pay as you go model, which one do you think will win?

    Secondly, if you are building and operating cloud centers and must choose between expensive proprietary OEM hardware OR cheap bulk-buy commodity hardware, what do you think will happen? Razor thin margins cut both ways and will continue to do so in future.

    Thirdly, if you are a business with conventional data centers and about to renovate...... (unless you and your share-holders stuck your heads in buckets for the last few years).... you will have noticed the above trends and act accordingly.

    All this spells a very disruptive future-present for companies that depend on proprietary hardware sales and outdated service models.

    There are some clear winners already:

    Google (sheer volume),

    Amazon (sheer staying power),

    and probably Microsoft (sheer presence in the enterprise and sheer marketing skill).

    The losers will be the companies that refuse to adapt to the new reality.

    The winners will be the ones who make the new reality work to their advantage.

    And so it goes....

    I see a future where the giants simply consolidate further and beat everyone else on price. New arrivals are going to have trouble competing with players lhat increasingly resemble the power utilities or telco monopolies of old.

    Incumbents selling whips and buggies will need to up their game or find a new line of business. Has anyone seen a new PC or server manufacturer startup lately?

    Of course, business owners and individuals will continue to shop around for cheaper electricity, gas, phone service etc. And the same will happen with computing resources.

    But replacing the cloud utility with your own solution (unless it is a backup generator in an area with spotty power supply) will soon become foolhardy, because it just doesn't scale.

    1. Anonymous Coward
      Anonymous Coward

      Re: Of course the cloud will eat your lunch, time for a new diet

      But replacing the cloud utility with your own solution (unless it is a backup generator in an area with spotty power supply) will soon become foolhardy, because it just doesn't scale.

      I keep hearing that 'scale' argument being put forward, like the ability to 'scale' is of some value to an SME business. I'm afraid it's wrong, scale matters when you're a big company, it matters when you are going through transformation, but in everyday month to month operations of most companies scale is of very little relevance at all, requiring only enough scale to take them through the next couple of quarters.

      I understand why cloud is taking off, and it's nothing to do with scale. It's about cost, it's only about cost. I see businesses with one (often patchy) line into their business being 'sold' cloud because it reduces their cost. They no longer need in house expertise (cost saving), they no longer need all that equipment (cost saving), they no longer need the electricity (cost saving), they no longer need to dedicate floor space (cost saving). etc. etc. All those companies big enough to require large 'scale' capabilities, built their own scaleable infrastructure.

      The article misses one very important point, each of those big tech companies played their part in creating the current technology market place. They knew it was coming long before Amazon ever thought of running any kind of computing business.

      I know for a fact IBM was moving towards this model long before Amazon had even heard of the concept of cloud. Whilst the likes of Amazon may shake up the marketplace, and require those big tech companies to adjust their plans, they are still running their plans and moving the market where they want it to go.

      In case you haven't noticed, the small startups who come up with something interesting, get brought up by a big tech company, and the providers like Amazon (and many others) are all dependent upon the technology and developments that come out of those big tech companies.

      Don't go deluding yourself into thinking big tech isn't still running the game. They might be changing the game they're in to something you don't think is 'their bag', but they're definitely still running the game.

      /Mainframe_is_dead off

      1. P. Lee

        Re: Of course the cloud will eat your lunch, time for a new diet

        The Cloud isn't changing the game - that's just PR.

        The Cloud is a reaction to a changed game. The problem is that we've pretty much finished IT development. We have networked pretty much everything that needs to be networked. We have storage to store the enterprise data we need. Our accounting systems, desktop productivity hosts, PLCs, email servers have already been bought and implemented.

        IT has little left to improve. There are improvements to be made, but nothing like the gains of the last 20 years. Therefore, the prices of new kit and consultancy can no longer be justified, because the ROI isn't there. Companies stopped buying desktop PC's. Vendors have reacted by removing upgrade options. New laptops now have un-upgradable RAM, lower RAM limits and more stuff glued together to make them unrepairable.

        The Cloud is just moving workloads around, its unreliable and inefficient since the infrastructure isn't tuned for the jobs required. That provides a lovely new market for more hardware and more consultancy. It requires massive amounts of engineering to solve problems which smaller units simply don't have. How many times have I seen adverts about using the cloud to scale your email. Really? I've never met a company that didn't know how to scale its email. Before it was shut down, myrealbox hosted email was serving 250,000 people with a two node cluster for mail pickup and another two nodes for SMTP on *Netware*.

        Once everyone had bought all the software they needed, they didn't need to go back for more. So now we have OS's which time-out and to which applications are tied. Software is moving to rental and will be upgraded whether you need it or not, simply because you don't need it. Supply has outstripped production which normally means vendors should go out of business, but with software legalese, you now can't have an asset because the vendors need it to be a liability for you.

        You used to buy licenses to run software on a single computer. Now the software is hobbled to core counts, because a single CPU can serve a major enterprise and that is too cheap for the vendors. When that isn't enough revenue, they provide "appliances" running dual-core desktop CPUs to make sure you aren't using cheap hardware to get more use out of their software than they think you are paying for. The whole reason for faster hardware is being nullified with licensing and the hardware market has not many places left to go except consolidation. Once people have finished consolidation of major sprawl, server hardware vendors will be back in the same place that desktop makers are now.

        It will be interesting to see how things develop with FLOSS. Just as Intel is supporting MongoDB, I'll be interested to see if other hw vendors begin pushing open source, just as a way to allow better utilisation of their hardware. Current licensing schemes are driving poor design, because you can't just spin up more instances / new hardware easily. Why not replicate lots of databases (with high read-to-write ratios) across lots of nodes and load-balance so that new applications have only a small overall impact to any one node. If I were HP, I'd be pushing postgres rather than oracle in order to sell more HPUX nodes. By "pushing" I mean putting in resources, not just encouragement.

        This is the opening for new companies: software licenses are far too expensive. The downside is the big players wrapping up the industry in patents to discourage competition. As Google v Apple shows, you need deep pockets regardless of merit.

        1. Anonymous Coward
          Anonymous Coward

          Re: Of course the cloud will eat your lunch, time for a new diet

          Interesting heads up there. Many things do level off, and the markets respond accordingly. :)

      2. amanfromMars 1 Silver badge

        Time for a New Macro AIdDiet in a Hostile Systems Takeover Space

        In case you haven't noticed, the small startups who come up with something interesting, get brought up by a big tech company, and the providers like Amazon (and many others) are all dependent upon the technology and developments that come out of those big tech companies.

        Don't go deluding yourself into thinking big tech isn't still running the game. They might be changing the game they're in to something you don't think is 'their bag', but they're definitely still running the game. …. Obnoxious Git

        It is delusional to imagine that big tech is still running the game because of the small startups and smart upstarts who come up with something interesting to get bought up by a big tech company. And do not discount the one man/indie band outfit which be of interest to countries into bigger and higher tech advantage for massive titanic leverage, which can be equally well suited to be either creative or destructive and also both at the same time to deliver something new and interesting covering all points of influence and importance in between.

        You know, taking full advantage of ye olde quantum bit communications thing for out of this world endeavours, which can be either heavenly alien treats of devilish cunning or vice versa and much worse and/or better.

        Micro rules Macro just as surely as the Few who rule the Many but without the Foresight of Advanced Intelligence is Expensive Hindsight the Price that Banks Pay Human IT for Loss of Future Definite Vision.

        And do imagine that Systems Admins don’t realise that and fear for their future lives if they don’t pay the pipers who calls the correct tunes?

        There are no prisoners in cyberspace*, only hostages to great fortune for memory wipes from the map as certainly as corrupt and evil players are similarly summarily dispatched and permanently removed from dodgy teams on earthed planets ….. and/but you were all forewarned and forearmed ….. Be Good, Do No Evil ….. so you cannot complain about your choice of final dessert, can you, although most every ignorant and arrogant fool and inevitably useless tool will and does to no merciful avail.

        * That be where layers of cloud host all of the future and bury memories of the past which be blights on currents of presents.

        The Cloud isn't changing the game - that's just PR. … P.Lee

        The game is changed and Cloud is a major miner part of IT, P.Lee, and that’s just PR surely.

  4. Matt Bryant Silver badge
    Stop

    Re: AMPC

    ".....if you are building and operating cloud centers and must choose between expensive proprietary OEM hardware OR cheap bulk-buy commodity hardware, what do you think will happen?...." Gee, now where have we heard the 'cheap hardware is all I need to think about' fallacy before? Cheap, white box servers have been around for decades but it hasn't stopped hp, IBM and Dell building massive x64 businesses, and the reason is not anything to do with the hardware, it's the ability to ensure service on that x64-based platform. Quite simply, hp, IBM and Dell put the added management software and support services in to ensure your datacentre can run 24x7 to the agreed SLAs. Saying 'I can just move a VM and swap out a cheap x64 or ARM tray if required' is only actually going to work if you have the tools to even know there is a problem, let alone where it is and whether a tray needs to be swapped out. Cloud companies like Google and Amazon have to build up massive back end structures by hand to provide that custom service. Companies like hp, IBM and Dell spend millions in their labs to make sure their products not only bolt together into solutions where you can find the problem, but make it easy for you to do so without having to handcraft code to link it all together. That is why they are pushing OpenStack solutions, they know the real issue is not how cheap you can make the hardware (Capex), but how much confidence the end customer has that they can actually use it easily (which means cheaper Opex). The costs of running a solution over its life cycle are usually a factor of three to five times the original purchase price, so saving pennies on Capex with cheap hardware can actually cost you more Opex in the long run. So, no, I don't think just cheaper hardware is going to kill the big vendors.

    1. ToddR

      Re: AMPC

      IBM just gave up on x86 Matt. Oh and HP don't have a clue what they are doing. Dell has been taken private, Mr Andreesen was bang on!

      1. Matt Bryant Silver badge
        FAIL

        Re: ToddleR Re: AMPC

        "IBM just gave up on x86 Matt...." No, IBM decided to concentrate on more margin-rich areas, especially software. You obviously failed to notice they have not stopped making plenty of software for x64. Lenovo has bought the old IBM x64 biz because they think they can still make plenty of money against the white box boys, and they seem to just be better at it than IBM.

        "..... Oh and HP don't have a clue what they are doing....." Well, apart from being the leading x64 vendor, you mean? Please do show where one of the white box boys has passed even Fujitsu in market share, let alone hp.

        ".....Dell has been taken private....." Did you stop to think that maybe Mr Dell and his backers thought it was a good investment because they think they can still make oodles of cash from x64? Oh, I see the problem for you, that whole thinking bit.

        1. ToddR

          Re: ToddleR AMPC

          I suggest IBM have moved out because of the white box guys, which you derided in your original post.

          HP IS still king of x86, but make little money, buy sw companies them don't understand, but sell lots of printers at good margins. Long term decline, I think so.

          Dell has been taken private in the hope they can still be competitive in x86, (with Intel's support), against the companies they buy from. Difficult to see that working long term.

          Supermicro make the best quality x86 servers OEMed by many including: Bull and oh Fujitsu!! They may not be anywhere near the size of Fujitsu, but selling to them as a positive sign they are on the right track.

          Tier 1 x86 vendors are still the main players of global corps, but whereas 10 years ago there was zero white box suppliers, today its single digit %ages, but its growing, which, HP, IBM (no longer in it, that's x86 hardware Matt, as the article was) and Dell are NOT!

          1. Matt Bryant Silver badge
            FAIL

            Re: ToddleR AMPC

            "I suggest IBM have moved out because of the white box guys, which you derided in your original post....." No, IBM wanted to concentrate on higher-margin areas, including HPC x64. Lenovo have shown there is still plenty of profit to be made if you are willing to play the low-margin-but-high-volume parts of the x86/64 game. If you want to believe otherwise that's because you can't comprehend market share reports, which show the white box vendors all together have been struggling to get even 10% of the server market share for years.

            "....HP IS still king of x86, but make little money...." Yeah, backtrack some more. Now all you have to do is admit the major vendors are stills owning the market by a massive margin and you can go be wrong elsewhere. You also forgot that hp, by offering the support services the white box vendors can't, makes extra cash for every server sold that the white box vendors can't. Concentrate real hard and go read the bit of my last post where I pointed out that such services make three to five times the price of the hardware, then you might just understand why hp, Dell, even IBM, are not giving up on x64 server market.

            ".....Difficult to see that working long term....." Really? Why? Oh, did you forget to provide any actual argument or reasoning to support your premise? This is my surprised face, honest. Don't worry, you're probably too young to remember people saying exactly the same about data warehousing and hosting killing storage and servers in the Nineties.

            "....Supermicro make the best quality x86 servers....." That is your opinion, but it would seem strange then that they have such a tiny share of market compared to hp or Dell. After all, they've been trying since 1993, yet they just don't seem to be able to steal that market share. Now, I wonder if it's because they're too,niche, or maybe they just can't match the big vendors' services, or maybe they just can't make the money without OEMing their kit to companies like Fujitsu with actual services capability? No, it can't be any of those, right? So you tell us why, if being cheap is all it takes, why didn't Supermicro kill hp and Dell years ago? Should be good for a laugh! Before you waste too much time, you might want to consider the Warcraft case, where Blizzard insisted on the highest reliability, resilience and ease of management for their game server farms, and chose hp blades over white box vendors, including Supermicro.

            If you had actually read some background material before posting your whimsy, you might have read that both hp and Dell (and IBM) have been investing in tray designs for cloud companies. They are going for exactly the type of cloud companies that might consider buying tens of thousands of cheap trays from Taiwan and offering to tailor-make them similar trays, only with the added services of proper integration testing in their labs, full software stacks, and proper after-sales support. If you seriously don't think the big vendors have seen this coming and have been preparing for years (did you even notice their very vocal support for Openstack this year?), then all I can say is you would seem just as stupid as your posts.

            1. ToddR

              Re: ToddleR AMPC

              Oh Dear Matt, Neanderthal as always.

              Supermicro, or Lenovo will take time to win market share and I have already agreed that services and having a global presence has been key, but this is changing. I am guessing you are ex Tier1, probably HP, and yes guess. My reasoning is speaking with IT personnel at large enterprise shops and them telling me this information. Also don't remember saying you had to be the cheapest, but it does help. The services part is being eroded away, long term, and HP irritating customers with silly bios warranty charges is bound to annoy, but it all takes time,

              IBM may have moved from x86 HW to sw, but the original piece, dumbo, was about the cloud eating the Tier1 HW vendors.

              As for age, don't see the relevance, but I started with PDP11s. As for WOW, oh dear you need to get out of the house chum. It takes time, but the line for the HPs and Dells in th ex86 space is slower going down.

              1. Matt Bryant Silver badge
                Happy

                Re: ToddleR AMPC

                ".....Supermicro, or Lenovo will take time to win market share....." So, one minute you're insisting the white box vendors are feasting on the corpses of the big x64 vendors, then you're backtracking again and insisting the biggest white box you can think of needs time, despite having had two decades of time pushing their 'cheapest is best' product. Oh, and Lenovo are flogging IBM's designs, so if you want to claim the white boxers killed IBM then surely Lenovo can't possibly grow with exactly the same offering? LOL, your logic simply doesn't follow, please do take a deep breath and try again, mmmkay?

                ".....I have already agreed that services and having a global presence has been key...." No you haven't, all you have insisted is that the big vendors are dying on their feet simply because cheaper white box products are taking over the market by selling straight into cloud providers. You can't have it both ways.

                ".....The services part is being eroded away, long term....." Ah, don't you just love the guarded qualification of 'long term', as in you have zero proof of it happening but you want to believe it is so, probably because the Supermicro sales rep told you so. LOL! Once again, provide some quantifiable proof or STFU.

                ".....IBM may have moved from x86 HW to sw, but the original piece, dumbo, was about the cloud eating the Tier1 HW vendors....." I nearly got trampled in the rush of IBM trolls insisting I point out IBM PureSystems, complete with x86 Flex nodes, all packaged up into ready-to-go cloud offerings. Oh, didn't the Supermicro sales rep mention those? Looks like you need to talk to a lot more than a few industry reps in future if you actually want to claim to know something about the market.

                ".....As for WOW, oh dear you need to get out of the house chum....." So I'll take that evasion as just your admission that you actually didn't know of that event, nor the interest it raised in the then young cloud market, as Blizzard were expected to go the Google DIY tray path. Then again, I can understand why the Supermicro rep probably decided not to tell you about that lost sale.

  5. Anonymous Coward
    Anonymous Coward

    And who is the winner..?

    Ironically, the winner of all this cloud mess is... Intel, a hardware company. Basically almost all cheap, commodity servers are equipped with Intel processors. Parisc is dead, sparc is dead, power is struggling to cope, AMD is a marginal and ARM is not there yet. Software defined everything bullshit has to run on some hardware and everything runs on Intel. So, yeah, good game, Intel :)

    1. Don Jefe

      Re: And who is the winner..?

      Intel has always been a serious business (where serious business means getting a piece of nearly every in a business sector). I like to blame engineers for the success of Intel. Up until fairly recently Intel has always been led by real engineers with real industry experience, not just a numpty with a degree.

      Regardless, comparing Intel to ARM isn't a valid exercise. Intel is a manufacturer and ARM is a product designer who sells licenses, not hardware. They're the tech equivalent of Intel has always been a serious business (where serious business means getting a piece of nearly every in a business sector). I like to blame engineers for the success of Intel. Up until fairly recently Intel has always been led by real engineers with real industry experience, not just a numpty with a degree.

      Regardless, comparing Intel to ARM isn't a valid exercise. Intel is a manufacturer and ARM is a product designer who sells licenses, not hardware. They're the tech equivalent of Pininfarina. Which is a pretty good spot to be in, but not one comparable to Intel or AMD or Intel.

      ARM is a pretty small company, wee tiny for a 'brand name' tech company to be honest. Even if, for some reason, they wanted to compete with Intel, the combined revenue of their entire history isn't enough to buy even the tooling necessary to punch out enough product to compete with Intel.

      I think ARM is a shining example of how the patent system was intended to be used. They're a smart, very neat company with a long life in front of them. They just aren't in remotely the same business as Intel and AMD.

    2. proud2bgrumpy

      Re: And who is the winner..?

      Well intel is the winner for the moment, but Cloud also makes the software hardware agnostic, so long as the service delivered meets the end users operational requirements - who cares if it is running on intel / AMD / POWER / ARM?

  6. John Smith 19 Gold badge
    Unhappy

    "It's the end of history I tell ye"

    Some how I get a feeling of Deja Vu all over again.

    Know what I mean?

    1. Phil O'Sophical Silver badge

      Re: "It's the end of history I tell ye"

      Know what I mean?

      Not really. The important thing about Deja Vu is that it's the feeling that you've seen something before when you know you haven't

      1. Don Jefe
        Happy

        Re: "It's the end of history I tell ye"

        I suffer from vu jade: the feeling I've never before.

        1. Destroy All Monsters Silver badge

          Re: "It's the end of history I tell ye"

          Deja Vu: That's when they change the number of zeros on a dollar bill.

    2. Peter Gathercole Silver badge
      Coat

      Re: "It's the end of history I tell ye"

      Wasn't DejaVu bought by Google when they acquired the Deja News Research Service?

      Just saying...

  7. ecofeco Silver badge

    Where have I heard all this before?

    "The intervals between invention of new technologies and significant changes in the industry appear to be shortening, which means that if you are either a slow-moving organization or a hardware specialist, things are tricky."

    Where have I heard this before? Hmm, let me think. 1999, maybe?

    "New tech" is not and never will be a panacea that replaces true innovation. All I see happening is more pointless complexity being added thus creating more confusion and points of failure.

    Much like the ornamentation of antique firearms, so much new software and hardware has become useless and costly greeble almost to the point of Rube Goldberg.

    Not to say there aren't real advances being made, just that most of it isn't.

    I don't think any even remembers what "bum the code" means any more.

    Holy shit! A quick google search shows that no, nobody remembers. Exactly one reference after phrasing the request just right. to bum programming code

    We're screwed.

    As for the whole cloud thing? Most of you know where I stand on that. Why would anybody give the keys to the kingdom to a complete stranger?

    1. Don Jefe

      Re: Where have I heard all this before?

      Ah, now you're into the philosophy of development, it's a subject near and dear to my heart. What I mean, is that innovation and invention can be viewed as two very different things.

      In my world, making physical things, we define invention the creation of an entirely new thing. Creating a new thing by combining a mental image and the invisible 'stuff' that fills our universe.

      We define innovation as the novel way of combining multiple, extant, inventions, or parts thereof, in such way that the assembly becomes something greater than the sum of its parts.

      1. TheOtherHobbes

        Re: Where have I heard all this before?

        >We define innovation as the novel way of combining multiple, extant, inventions, or parts thereof, in such way that the assembly becomes something greater than the sum of its parts.

        How do we define 'silly management fad'?

        Who's going to be surprised when someone invents the 'on-site cloud' - fast, commercially secure full-stack locally hosted hardware with a third-party SLA and built-in redundancy for scaling - about 30 seconds ago?

        1. Anonymous Coward
          Anonymous Coward

          Re: Where have I heard all this before?

          "Who's going to be surprised when someone invents the 'on-site cloud' - fast, commercially secure full-stack locally hosted hardware with a third-party SLA and built-in redundancy for scaling?"

          Pardon my ignorance, but what's the difference between that and "private cloud", which everyone who thinks they're anyone (Dell, HP, MS, Rackspace, to name a random few) has been talking about for a lot longer than 30s?

          Apologies if my irony detector has failed.

      2. itzman

        Re: Where have I heard all this before?

        And there is the philosophy of adoption 'when my horse dies I'll probably get a motor car'..

        As technology and the costs associated with it shift so too do usage patterns. When hardware was damned expensive, time sharing mainframes..then hardware was cheap, the PC, then networking..pushing the trends back to 'globalised' raher than 'localised' data. Add in security issues and that produces a bias towards 'localised' again..

        Software? The problem there is that actually all the big software stuff has already been done.

        WE have in *nix, the basic operating system that can run on anything from a supercomputer to a fondleslab.

        We have in the GUI a basic user interface paradigm that most people are pretty comfortable with, and needs but minor tweaks to be usable without special knowledge.

        WE know most of the basic tasks to which computing power can be put - storing manipulating and transmitting data in various useful ways.

        IN short its actually pretty mature as a marketplace. And when a market gets mature its a harder sell and a less profitable one. IN my youth Men would Talk Motor Cars and have Beliefs about Tyres and so on. Advertisements for brake pads, spark plugs and what oil you put in it abounded. Today no one cares. You buy a car, and apart from a few inessential details of style and user level features, its almost identical to any other car in its class.

        I dont own a fondleslab, or a smart phone, but what strikes me about the ones people bring to show me,. is how remarkably similar they all look.

        I have a strong suspicion that the IT boom is nearing and end. IN the sense of rapid penetration into areas that never before had computing. People are running ten year old plus software on ten year old plus hardware that still fulfils the purpose for which it was bought. I vividly remember a 21 year old IBM 360 covered in dust, still running the software it was bought to run..shortly to be replace buy an AIX PC running the SAME software.

        Cloud isn't new, it isn't even interesting. Its just another twist on the old game of data and communication.

        Bit by bit free software is catching up with the brand leaders of 20 years ago simply because the game is over. WE know what we want a spreadsheet, a database, a word processor or an email package to do.

        If Microsoft, Apple, HP, Dell, Adobe, Oracle didn't exist, what actual need would they fulfil to justify their creation?

        Ask yourself those questions and if the answer is 'yes. they still have something to offer' then that's a company that will still be there in ten years time.

        I think the reality is that the exciting innovative days of IT are pretty much over. Its now a tedious cost-benefit-risk analysis to combine the components we have onto a variety of standard solutions that will be increasingly margin squeezed.

        Its not quite all over, but the rush of the 70s, 80s and 90s is gone.The Turing machine has arrived, expanded to fill and empty market niche, and is now a permanent fixture. We haven an IT infrastructure, and we have pretty much all the software we need for the basics. The rest is niche stuff.

        A final message from 'Big Blue'

        http://www.youtube.com/watch?v=iUaevnP1LLg

    2. Kiwi

      Re: Where have I heard all this before?

      "I don't think any even remembers what "bum the code" means any more."

      Phrased like that? No, I don't remember it. But in practice I do. I remember a time when an update to a program either added significant new features, significantly improved speed and reliability (not one or the other, but BOTH), or significantly reduced the program size while also removing some bugs.

      These days "upgrades" seem to do all the opposites - add bloat, remove functionality, slow the program and maybe your machine (bye bye Firefox, you've usually been good to me for all these years but latest update is last straw). I can't think of when I saw any real improvements in an update (aside from security fixes)

      Lost art sadly, code optimization.

      Oh, and I agree - a lot of the "new" isn't that new and less of it is useful. We did a lot of this "cloud" stuff back in the "grid" days. Name has changed but the guilt is the same...

  8. Anonymous Coward
    Anonymous Coward

    Google breakup creates more cloud capacity

    If EU manages to break up Google into smaller entities, each one will have about today's entire cloud capacity, only competing against each other, accelerating the effect outlined in the article.

  9. Anonymous Coward
    Anonymous Coward

    Yeah, I think this is overstated. It has more to do with the commoditization of hardware than "cloud" computing.

    SAP - They are not struggling at all. The sale of on-prem licenses have been growing. They are laying people off, as corporations do frequently, but the business is not struggling in the least.

    Oracle - Hardware problems are no big deal. No one ever thought Sun was going to be a real growth engine. Anyone who thought that would be a growth business was deluded. It was in free fall when they acquired it (primarily because they wanted Java and MySQL... hardware was thrown into the deal). The software businesses are doing fine.

    IBM - They are struggling in hardware, but that has less to do with "cloud" and more to do with companies buying less costly hardware for on-prem (e.g. x86-Linux instead of pSeries).

    Rackspace - Many would consider them a cloud company, so they run counter to your "cloud everything is good, on-prem everything is bad" thesis.

    Dell - Bad business model, but they were struggling long before the term "cloud" was invented.

    HP - Same problems as Dell. Selling commodity hardware, the price of which goes down every year.

    Cisco - Their 10g upgrade cycle is over. Not many people are upgrading their cores from Nexus to Nexus just for fun. They also have more competition from HP, Juniper at the edge. Another proprietary hardware getting, much more slowly than servers/storage, replaced by commodity story.

    1. K

      Common sense at last :) very nice breakdown and couldn't agree more..

  10. Roo
    Windows

    Cloud really shouldn't be a threat to IBM, Oracle et al, they have enough IP lawyers and cash to sue and buy their way into the market if they really want to.

    The real problem they face is that OpenSource software provides an alternative with a low-zero price, which is slowly but surely eroding their margins.

    The vendors only have themselves to blame, if they punted good products at a reasonable price that the majority of folks could afford and took better care of their customers there wouldn't have been as much motivation for folks to roll their own and open source it in the first place, RMS notwithstanding.

  11. h3

    Funnily enough Oracle and Sun were right with the "The Network is the Computer thing" just too early.

    1. Roo
      Windows

      "Funnily enough Oracle and Sun were right with the "The Network is the Computer thing" just too early."

      I give Sun a lot more credit for that because they had enough conviction and vision to actually deliver (NFS, YP/NIS etc...) and folks actually used it - and some liked it. I recall Larry Ellison banging on about it, but I didn't see anything that actually backed it up. I put that down to Larry just doing the Visionary Salesman thing in between America's cup races, if anyone knows better I'd be interested to hear. :)

  12. Mitoo Bobsworth
    Mushroom

    CLOUD

    Corporate Lackeys Of Unashamed Dissimulation.

    Mushroom cloud award of the month goes to Adobe for a timely demonstration of this acronym.

  13. amanfromMars 1 Silver badge

    Irony Streams are Pots of Gold for the Purchasing of Untold Riches ...

    .... for the Tapping and Trapping of XSSXXXXual Pleasures' Treasures.

    OpenStack has been enthusiastically adopted by incumbents such as IBM, HP, Oracle, and others, to help them make tech to defeat their younger rivals. [In a sign that irony is dead, they have even referred to themselves as "The Rebel Alliance".—Ed).

    The forces which be struggling with law and order, and a prime example would be Britain's National Crime Agency (NCA) and most probably most other agencies and orders with needs in the provision of leading intelligence sectors and vectors, to combat and/or assimilate and driver and not radicalise fundamentally a growing metadatabase , are similarly into such cloakings of themselves, are they not ........with a particular and peculiar known unknown example being The Flag Day Associates?

    IT aint rocket science and it is a long ago well known fact that delivers a real live virtual fiction for home base command and control, that to create a new enemy has one issuing eventful controls for battling against crazy ghost hosts which are merely simple aids used to justify and maintain and retain proxy remote negative power with mad destructive capital spending ........ which be the product of intellectual property deficit and challenges which result in supplies of increasingly burdensome and inevitably unsustainable and unaccepted and unacceptable debt.

    Welcome to the Flagship ZerodDay Explosive Exploitive Great Games Suite ..... where the Rebel be Renegade and Presidential and Prime Ministerial King and Queen and Ace Player of Master Piloted Suits/Advanced Persistent Threat Teams for the Active Cyber Terrain of Live Operational Virtual Environments ........ Post Modern Information Rich and Intelligence Led and Fed Media Reality.

    Or are you following something else, which has no base in and relationship with intelligence and which be always delivering chaos and mayhem/doom and gloom/fear and conflict/division and inequity?

    Speak up now, don't be shy and don't be fooling yourself with the sharing here of nonsense for such as is revealed here is future phorming all of your lives, and beta testing for all lies too.

    1. Alan Bourke
      Pint

      Re: Irony Streams are Pots of Gold for the Purchasing of Untold Riches ...

      Authentic frontier gibberish.

  14. Legal IT Law Technology

    RTW Hosting & Backbone UK - also thriving

    Some pure play cloud providers in the UK are thriving with the rise and rise of cloud computing, such as RTW Hosting, while those who provide Hybrid It that integrates on premise/ often legacy hardware with Cloud systems, such as Backbone UK are also seeing significant uptake

  15. Jim 59

    "Cloud computing"

    You can ascribe any industry shift to "cloud computing", because it is a flexible layman's term that can mean almost anything you want it to. In this article, the author has chosen to group all positive market movements and call them "cloud", and has categorized all market falls as non "cloud".

    The change cycle is unpredictable, unlike the hype-cycle.

  16. kmac499

    It's Ecological succession not Evolution..

    Digging out my old Biology education, the changes in computing devices and platforms smack more of an ecological succession rather than evolution.

    Broadly, evolution brings foward new forms of things which may or may not suit the niche\world they are born into. Evolution is undirected and random leading to failure and sucess. Winnowing of the weakest as well survival of the fittest over vast periods of time.

    Ecological succession is where a niche pre-exists, and organisms move in colonising it. Once there the pioneers change the environment by exploiting whatever resources they find. Even if this means eating themselves out of house and home. It's quite common for one organisms waste to overwhelm it, but provide the next niche for the next coloniser. This is the succession part of the process. Usually the ecosystem becomes more diverse and eventually plateaus into a final or apex state, often some form of forest in a much shorter timescale.

    I think computing is best described by the second model because there is very little evolution in the core 'organisms' processors, memory, storage etc..Yes individual components improve in performance but that's a change in word size, speed, instruction set etc.. Organic systems follow the flow of energy, computing systems follow the money, allowing old now expensive configurations to die off and new configurations to become economically viable as component prices and service prices fall,

    Computers Expensive one mainframe to rule them all

    Computers Cheap one personal computer per desk

    Computers dirt cheap one pc phone tablet per person etc..

    Computer management expensive inter-Netwoking cheap the Cloud

    Next development <insert £million guess\bet here>

    A final analogy is that software is mildly eqivalent to DNA in that it directs the hardware and being self replicating can overwhelm, repurpose and completely change the hardware that is directed by it. It is not for nothing that the word virus covers both live and software entities. and ecological disruption for good or bad in both spheres can arise overnight.

    Well that's my IMHO anyway..

    1. amanfromMars 1 Silver badge

      Re: It's Ecological succession not Evolution..

      Next development <insert £million guess\bet here> ... kmac499

      The AI and SMARTR IntelAIgent Machines revolution is the next and current stage in the ecological succession over evolution model, kmac499, and IT delivers the next priceless money pit development for wannabe lead chargers and wiser move makers [and coincidentally, also prime movie makers] and is nothing less than Live Serial Applications of Virtual Reality, which be way beyond the limited ken and puny proxy command and control of existing secretive orders and traditionally empowered and hierarchically established status quo systems admin.

      And IMHO, you’re on the right track, kmac499, with .... A final analogy is that software is mildly eqivalent to DNA in that it directs the hardware and being self replicating can overwhelm, repurpose and completely change the hardware that is directed by it. It is not for nothing that the word virus covers both live and software entities. and ecological disruption for good or bad in both spheres can arise overnight.

      And if the Wild Wacky West doesn’t pay and want to come out to play nicely, will it logically be offered to the Erotic Exotic East, although if you can believe what the Yanks have made up to pass off as global news today, a direct targeted approach to Unit 61398 and/or APT1 might be the right sort of top gun persons of interest to have a pow wow with. They certainly appear to mighty gifted and suitably impressively inquisitive.

  17. Anonymous Coward
    Anonymous Coward

    lose money on each sale but make it up in volume

    That's been the business model of a number of these "winners" like Amazon - and let investors named Ponzi make up the difference.

    It's a model that can succeed for a surprisingly long time - unless/until it changes there isn't much point in challenging Amazon for hosting. IBM and HP will have to find something else to do.

This topic is closed for new posts.

Other stories you might like