Dell, EMC, HP, Cisco are the walking dead
wired.com
wired.com
This is how the rest of the world works -- I run an import/export business and have to manage logistics. It doesn't matter what industry, it could be crude oil or widgets from China at half a penny a piece. It's my job as CIO to choose an implementation.
I'm going to choose SAP or Dynamics (or one of the handful of other ERP's with name recognition) on DB2 or Oracle. I'm going to store my data on EMC dual array filers for HA and pay the tax. Why? Because if I choose node.js and MongoDB to run my 3 warehouses and the system goes down, I'm getting voted out next board meeting. If my SAP R/3 on IBM hardware goes down, they'll do the dance and at least fly some people out to retain their image keeping your job effectively safe.
Banks aren't going to move over from z/OS anytime soon. I don't want my dialysis machine to be running on Riak.
1. Cloud providers are making it easy for legacy businesses to take baby steps. You don't have to use a NoSQL solution with AWS Dynamo, you can use Postgres with RDS. Or if you're hyper-conservative/stuck with legacy stuff, you can run Oracle with RDS. Or get an EC2 instance and install your own personal copy of Oracle on it. You don't have to jump to full modernity right away, you can take the easy wins of hosted infrastructure first, and only later move up to the bigger wins of hosted platforms.
2. Yes, legacy stuff takes forever to disappear. There are still businesses running VMS inside of VAX emulators with COBOL apps on them. But once you're in that "super-conservative businesses won't move away from you anytime soon" area, you're in a declining market. New businesses unencumbered by legacy won't go down that road -- small businesses starting today will never buy an on-site server to start with (thanks to Office 365/Google Apps/etc.), and as they grow there's no reason for them to change that policy anymore. EMC might never get a new customer, and even if they keep their old ones for some time, it's just a question of how fast their decline is.
Or get an EC2 instance and install your own personal copy
of Oracle on it.
1) another one of those fun vendor software needs Oracle things
https://aws.amazon.com/blogs/aws/amazon-ec2-dedicated-instan...
:)
Btw you'd cry if I were allowed call out some companies that still run lotus notes.
I worked for one (Raytheon). :/
How many Amazon, Google or Microsoft engineers (or even SREs) have ever set foot in one of their datacenters.
Hint: the answer is basically none.
The other thing is the writer seems overwhelmingly fond of Pure Storage...until they are bought by one of his Walking Dead. Instead of calling these companies "Walking Dead", he should be highlighting more the approach Microsoft has done. What Dell/EMC, IBM, Cisco, HP, and more will start to see is that they can operate in both worlds. Amazon will never turn over there stack to a company so that it can operate inside that companies physical control, but the listed companies (like Microsoft) can make there stacks available "in the cloud" and palpable to the public furthering there market.
To me, whereas Amazon and the like have plenty of growth ahead of them, these "Walking Dead" have higher ceilings although it will take quite a bit of inertial change. I can see IBM already working and heading in that direction. If they (like Microsoft) become another success story, it will be hard for the others to ignore that strategy.
But the author got his clicks and his eyes...so I'm sure we'll see of this.
The old joke was "no one ever got fired for buying IBM".
It's not even about what the best technology is for a solution but rather the intersection of what's the best I can buy that isn't gonna get me fired if it goes wrong.
Ps. Having migrated an SAP off of AIX, I'd say you'd be better served with something like a VBlock or even VirtuStream or Amazon cloud to run it. There also are price and performance incentives to run R/4 on HANA than on Oracle.
I've been a contractor for a firm that still runs their own mainframe for data transformation, and had Cisco experts on the payroll. They ran their own vCenter instances integrated with their Active Directory, so moving with them to Vblock was painful mostly in giving up control.
I personally couldn't look myself in the mirror , speaking as a former VP of IT ops at a large company, if I made a decision where the primary concern was for my job stability. In fact, the fact that I could get fired for making the difficult choice usually was a good sign. I'm not talking about making reckless decisions, I'm talking things like firing IBM outsourcing because they're price gouging and risking the inevitable call to our board of directors that is attempting to get me and my boss (the CIO) fired.
That said I do agree many IT people make the decision for their own stability first, and the company second. That usually is due to a poor business leader that promotes a risk averse culture - unlike a Reed Hastings or Elon Musk.
AWS is absolutely killing it. Every release cycle, they put out quality products that directly attack Oracle. And, Oracle is in the bad position of having to eat their existing profitable business lines to compete.
Now, of course, cloud isn't here yet. There are certainly lots of issues around security, auditing, privacy, data ownership, etc. But, the trend is accelerating. Why? Because the cloud is so much cheaper. Why maintain and manage your own data center and infrastructure for millions of dollars when you can reliably outsource for a fraction?
I used to do a lot of work in the utility industry. I remember years ago giving a presentation on the cloud. The audience was passionate that it'd never, ever happen to them. They were extremely protective of their data. But, then I slowly saw it happening anyway. It crept in slowly. The first to go were the lowest risk systems. Then, the next lowest risk systems. Then, the next.
The incumbents will hold position for a long time to come, propped up mainly by government spending, hellishly long sales cycles, a corporate aversion to micro products, and FUD. They have plenty of time to catch up. But will they? That's the question. Unless oracle is actively researching new cloud platforms (and I'm guessing they are), they're going to be in pain in 5 years. Corporations that don't move to the cloud will be in a cost disadvantage.
On the other hand, serious work needs to be done to overcome the severe limitations of the current cloud. Lots of certification and auditing and risk assessments and SLAs need to happen.
Oracle is ultimately a software company. Where their customers run that software is largely irrelevant to their core business. Is there margin to make by running an Oracle cloud? Sure, but the bulk of their profits are made just by licensing their LOB software packages and the associated annual support contracts. Oracle will be fine, and Larry Ellison will be able to buy an even bigger yacht in a decade.
That said, the Oracle database product, while extremely expensive and obnoxiously licensed, is not without technical merit. There is a genuine market for it, just not a $50b one. One could say the same for what these other so-called dinosaurs are selling.
Oracle makes the majority of their money off of industry-function-specific ERP software that runs on top of their database platform. That's a market that's very hard to disrupt or displace, because you have to develop deep domain expertise in dozens of different industries and functions. Their technology platform licenses often sneak in the back door -- the customer wants the solution, and if they have to pay for the DB as part of the solution, then it's just part of the price.
http://www.wired.com/2014/11/facebooks-new-data-center-bad-n...
That's basically the point of the article: These companies that actually had to get up to that kind of massive scale quickly discovered that the way hawked by these big expensive "enterprise" companies didn't work. It was expensive and insufficient. So they've developed their own custom stuff (often built on open source; sometimes open sourced back out) that works a lot better while also being massively cheaper.
AWS isn't just a bunch of Dell servers hooked up to EMC SANs and Cisco switches. The only people who still use that stuff are people who a) have a ton of money and a desire to spend it freely, plus b) don't have really serious demands, which would mandate something better.
That's not a good place to be.
Similarly with Cisco, you can run the internet exchange of a small country on a single stock Cisco switch, and many do. Cisco isn't going anywhere in the near future. Amazon, Google and Facebook might not have a need for them, but thousand of other companies do, even large datacenters.
There are millions of SMBs and tens of thousands of large corporations around the world that own server hardware. A tiny percent of that is the major internet companies.
The designs for hardware are changing but the providers are still going to be the usual companies that are good at this and have all the manufacturing in place. For example a lot of banks have started to switch to Open-Compute designs [1] but they will still buy these from Dell or HP.
Software is a different story.
1. http://www.datacenterknowledge.com/archives/2015/03/11/open-...
Very few companies have the kind of network requirements of Facebook or Google.
And if Cisco/Dell/EMC stuff was bargain-priced and the ultra-high-end stuff was expensive, well, there you go.
But the problem is, if you want your stuff hosted on Amazon-class hardware/network topologies, it's really cheap to do so with AWS (or Azure, or whatever). And the Cisco/Dell/EMC stuff is painfully, ruinously expensive.
If the high-end thing is also the cheap thing, what's the rationale for buying the low-end, expensive thing?
In the long run the "walking dead" from the article are doomed when due to technological growth an entire data center can be converted to a single rack in a closet without any expensive monthly upstream bandwidth charges and nickel and dimeing from a cloud. Those suppliers have expensive corporate infrastructure that depends on shipping $50M data centers in 2005 not a single $100K rack in 2015, or 2025 or 2035 or whatever, to do the same task. Maybe a great sales campaign can convince them to double their usage, that still leaves them $49.8M in the hole, non-inflation adjusted (ouch).
Computational use for non-computer businesses is kinda like electricity in the VERY long term. If you look at the early years of the growth of electricity vs GDP or whatever, you can rapidly get strange ideas like the average office in 2015 based on projections from 1915 should be radiating heat like the surface of the sun. However, here we are, there is a natural leveling off of how much electricity or municipal water or food a business can usefully consume per unit of GDP. Assuming GDP only goes up, inflation adjusted (LOL). When companies bounce off those limits, life will get very exciting for "eternal Moores Law growth forever" oriented suppliers for those businesses. Based on facebook use stat growth rates in 2007 we should all be spending 78626 hours per day on facebook... didn't happen... building a supplier company based on that assumed growth trend will lead to tears. Moores law says the transistor will shrink not that people will buy it.
All this also because these companies have never really tried to attract the best talent by paying above market compensation. It's fine.
[1] The kind of detachable GPU configuration we saw in Surface Book is harder to pull off outside of MS. Doable, but MS will have the advantage of getting inputs directly from the DX12 team.
Also, you should never underestimate corporate inertia. Switching to a new system is immensely expensive, for many companies it won't happen, ever. Just look how many companies are still using decade old hardware and scour ebay for replacement 286 processors.
I'm not being snarky - I've seen many a large business make many a horrible decision, not due to a dispassionate evaluation of the pros and cons of each option, but because the people in charge hate change. And it's not that said change would be necessarily painful or demonstrably bad, it's an irrational bias.
Change hate appears to scale linearly with organizational size, more specifically, number of managerial staff.
If nobody was using them anymore, they'd just be the dead-dead. The "walking" part comes from their existing support base propping them up.
"Oracle and EMC are fine, just like 286s!" is not an argument that would comfort an executive at those companies.
It is not only about cloud. It is that these companies did not evolve to match current needs: it is about sales process (just try to buy cloud from Oracle and you will see), it is about support, it is about pricing, it is about technology (SOAP?) ...
We're going to keep local servers for a looong time to come.
Sound familiar? This fate awaits all these companies:
- Dell: I don't even know what Dell makes anymore. I guess they sell PCs, servers and consulting services to small businesses that don't know any better. But this market is being eroded by an army of local MSPs and resellers, and it's not a model that benefits from scale.
- HP: HP has been a disaster for a while, but they will probably endure since their hardware is generally pretty good and reasonably priced. But servers and printers are not a very high-margin business, and that's caused their leadership to be constantly on the hunt for a newer, higher-margin revenue stream.
- EMC: Storage is increasingly distributed, and EMC's centralized storage solutions really are a dinosaur in today's market. They just make the wrong product and never found anything to replace it.
- Cisco: By all rights, Cisco shouldn't be on this list given that the Internet still largely runs on their hardware. But they convinced themselves they were a consulting company, and that ran off all their good engineers. Cisco today is a shadow of its former self.
Their biggest liability is their customer base that IS NOT changing and continues to pay for these products. I see guys everywhere trying to cajole customers down the cloud path because they'd rather be at the seat of that inevitable transition than pushed out. Many customers don't want to. So they're stuck doing the old thing that's dying and the new thing is being done by smaller customers that wouldn't even consider them.
The story of if/how they fail or rise again will be a lot more complicated than you are suggesting. Keep in mind Apple was near bankruptcy 18 years ago.
Example - EMC. I'm not sure what you're talking about, there are a lot of products. Isilon is distributed storage and selling like mad. ScaleIO is free for anyone to use unsupported and a distributed Elastic Block Store. VxRack is scale out compute and storage. XtremIO sells more and is growing at pace with Pure (which just went public). Dell in swallowing them up which seems like the least worst option given Elliott is at their back and HP would be a death sentence. (I don't work at EMC, but a subsidiary and have some knowledge of how they work and sell.)
EMC is generally just fucked because their potential customer base is shrinking. While I don't think everyone everywhere will move everything to the cloud, enough migration is heading in that direction that storage appliances seem like a risky business (especially considering the big cloud providers just roll their own rather than buy from EMC).
Don't get me wrong -- I fight the same battles with big companies about how their stupid new $50million data center can't hope to compete with AWS on cost, reliability, security or functionality. AWS runs datacenters at the billion dollar scale. CIOs/CTOs have the mindset of "it'll be better if we control it" -- but then don't realize how expensive it is to replicate all the niceness that you get from AWS like management interfaces, workflow control, etc. All those systems have to be grafted on later, and are naturally behind the curve.
I find it strange that the Author didn't include Oracle in the title although he talks about it at some length in the article body.
Oracle doesn't do this; the bulk of their value comes from software that is much more difficult to replicate.
Microsoft fucked because they have Windows on 95% of computers?
Agh, i'm too exasperated to continue my train of thought. Yes, everybody is fucked because of Facebook and Google and MongoDB.
You'd have to start by defining "computer". If we use a technically reasonable but broad in journalism's terms definition of "device containing a CPU with integrated MMU", then we're including every mainframe, server, desktop, laptop, smartphone, tablet, industrial control system, laser printer or multi-function document management system, router, network or storage switch, storage system controller, sound mixer, and countless other devices. A very small and diminishing fraction of those machines run Windows.
Even if you limited the definition to include only things we traditionally think of as computers, namely mainframes, servers, desktops, and laptops, Microsoft's market share is something like 30%, though it's very difficult to measure and once again subject to definition; see https://en.wikipedia.org/wiki/Usage_share_of_operating_syste.... And again, that's been declining for a long time. Even the most generous interpretations that include only desktops and laptops don't get you to 95%; you can narrow-definition your way up to about 90%, and that assumes "market share" includes copies that were made in violation of license and/or law (and thus for which Microsoft received no money).
The reality is that Microsoft is a niche player, not because they lost a market they were previously winning but because that market itself became a small part of something much bigger, in which Microsoft was never especially competitive. As noted by the author, they've been investing heavily in changing that, but the early returns aren't much to talk about. Windows Mobile/Phone is a complete dud, Surface is a weak seller despite recent improvements, and Azure is at best a very distant second in IAAS (see http://www.datacenterknowledge.com/archives/2015/05/28/gartn...). What's remarkable is not that Microsoft is doing so poorly but that they're the only enterprise computing company that even appears to be trying. I think that's what the author was expressing. The characterization as "fucked" comes from Ashlee Vance and should be given the credibility one would assume considering the source. There's little question that most of these companies' positions of relative importance are in decline and have been for a long time. But some, especially Microsoft, still have enormous sums of cash and businesses that are continuing to generate more of it. However irrelevant such a company is, it's still a long way from going out of business. Conversely, no amount of cash is a substitute for organic sales growth.
Microsoft are so far from niche it's unreal, and it shows a great ignorance on behalf of the author to look at one product from one arm of a diverse organisation, which has been at the forefront of CS research for decades, has competences that Google et al absolutely do not, and compare it to other companies which only have vague product-line overlap so blithely.
This is ignoring the manufacturing, supply chain management, distribution, retail competences that the other mentioned companies have, which Google and Facebook may never have. It's also ignoring fundamental aspects of marketing such as brand equity, positioning, etc. It's also ignoring the fact that some of these companies are already embedded in industry, via their product line and also via consultancy. Lastly, it's taking a single current snapshot of the industry, and is ignoring the fundamentals of organisational strategy and competition and capitalism, which would suggest that the companies mentioned will adapt and compete to survive (in b4 Innovator's Dilemma). Not all companies will do this successfully, but the companies mentioned will.
It seems like the author is judging these companies via their perception of a small subset of their products, and the market's perception of them (some of their share prices are up and down) without looking at the fundamentals of each business, their diversification strategies, product line strategies, their financial metrics (which the author could've looked up in 10 seconds), and so on.
I mean to suggest that these companies are in trouble because of OSS such as MongoDB, which they have partnered with, support, and sell (and more than likely funded at some point), is very silly, but that's precisely what the author has done.
It's not 1992 any more. At the very minimum, it's reasonable to assume that "computer" also includes mobile devices and servers, two additional markets that are well-known to both technical and lay audiences.
> I presume that everybody here would understand that I didn't mean "device which can compute", i.e. all servers, cash registers, calculators, difference engines and Turing machines.
You haven't shown that you know what an MMU is. There's a reason I included that; the definition excludes microcontrollers, devices that lack external memory, and other things that clearly cannot run Windows or any other competing operating system. Most of the devices you mention either don't have CPUs or the CPUs they have don't have MMUs, and they're not part of the Windows addressable market. Since you brought up Windows in the first place, doesn't it make sense to limit ourselves to the general classes of computer on which Windows could conceivably run?
> This is ignoring the manufacturing, supply chain management, distribution, retail competences that the other mentioned companies have, which Google and Facebook may never have.
I didn't mention either company. Both are advertising companies; Facebook doesn't make any kind of product that competes with Windows, and Google's competitor (Android) isn't really its own anyway, as the actual OS is Linux.
> Not all companies will do this successfully, but the companies mentioned will.
I'm not sure why you're going off on me. I explicitly said that Microsoft is likely to remain in business and making money for a very long time. In case it isn't obvious, making money and successful are the same thing. A company can make money serving niche markets, and many do.
> the fundamentals of each business
Microsoft's top product by both revenue and profit is Windows. Windows has a leading market share on only a single class of device (desktops and laptops), and that market itself has been consistently shrinking in each of the last several years (2013: http://b-i.forbesimg.com/anthonykosner/files/2013/04/ChartOf... 2014: http://www.engadget.com/2014/10/08/pc-market-share-q3-2014/ 2015: http://www.gartner.com/newsroom/id/3090817). Forecasts are for continued shrinkage of that market. I don't think it's fair or prudent to ignore those fundamentals, as they clearly are central to Microsoft's ability to make money. Azure may be doing well and may even grow into a strong #2 in its space, but it's still less than 10% the size of the market leader and it generated well under a billion dollars of revenue last year (http://www.forbes.com/sites/louiscolumbus/2015/04/15/sizing-...), accounting for less than 1% (http://www.google.com/finance?q=NASDAQ%3AMSFT&fstype=ii&ei=R...) of the company's total turnover. It's difficult to know if it was profitable, but it hardly matters; the operation is barely even material to the company as a whole. Azure would need to double in size 4 more times to account for 10% of Microsoft's revenue.
Again: I didn't say Microsoft is going to go out of business tomorrow. Or next year or even in 2025. But you should acknowledge that a backward-looking view of the company is risky, and that an objective forward-looking assessment of its business is less rosy than you would make it out to be. At best, the company will be much less dominant than it was in the 90s.
> I mean to suggest that these companies are in trouble because of OSS such as MongoDB
I didn't. As I said, there's a lot of stupidity here. There probably are some small e-commerce and ad companies using Mongo that should be using Postgres or MySQL or Oracle or Microsoft SQL Server, but I don't know that those companies would have gone into business at all if they had to pay an enterprise software vendor for database licenses. I agree that the impact of MongoDB on the named companies' bottom lines has been negligible, and that no evidence of such impact has been offered.
The author and many of the comments here have strayed well into hyperbole, probably fueled in part by ignorance and in part by schadenfreude and glee at seeing some poorly-managed companies in mature markets struggling to remain successful (HP is, after all, the poster child for mismanagement in this space). I don't believe I've made the same error. If you think I have, please provide supporting evidence, as I have provided for my position.
Obviously this is not economically efficient, and thus eventually the scheme has to unravel.
If that's overpriced I really need to talk to your hardware supplier.
(Of course, they all have high-end products, but the breadth of the offerings is usually high enough that they're always an option. They're not stupid.)
EDIT: why did this get downvoted - am I factually incorrect or you simply disagree and want an echo-chamber?
While reading this I was thinking, where does Apple fit? Does it have any big plan for enterprise at all?
Does Amazon buy servers from Dell, HP or IBM?
Thomas Watson, president of IBM, 1943
"Nuclear-powered vacuum cleaners will probably be a reality within ten years."
Alex Lewyt, president of Lewyt vacuum company, 1955
"Two years from now, spam will be solved."
Bill Gates, founder of Microsoft, 2004