IBM's bailed out of the server market - will they dump Storwize next?

January's not even ended yet and already we have an interesting technology market happening; IBM’s withdrawal from the x86 server market does lead to a number of questions. Not only the future of IBM is under the spotlight, but also what IBM execs feel the future of the market is. Could this be yet another sector they withdraw …

[A Non e-mouse]

I suspect IBM will stay in the storage space while they think they can differentiate and still make a reasonable profit.

Desktops and x86 servers have became commodity: Everyone selling identical boxes at wafer-thin margins. A tough environment where you need large volumes to make a tiny profit.

[The Godfather]

Good on yer...

Nothing wrong in dumping big chunks of your business if it's those parts are not making money and no real future point of profit is visible.

[SirWired 1]

Seriously?

Firstly, you are getting Storwise and the SVC overly conflated. The Storwise is a RAID array that happens to run SVC under the hood; it has nary an xSeries part in it at all. (Nor, for that matter, does XIV.)

Even the SVC, which does indeed run on xSeries hardware, is not coupled that tightly into xSeries. There are some parts of the hardware that are designed to run on the particular xSeries model for that generation SVC node, and of course the mgmt software talks to the mgmt software within the xSeries, but they could just as easily use an x86 box from anybody. Most of the price of an SVC install is in the capacity licenses; the hardware node is practically a rounding error in comparison.

On a practical basis, the decision to keep or dump xSeries doesn't have to be made for years; there's no reason for them to stop using the xSeries boxes any time soon; at worst, they'll see a minor H/W cost bump if Lenovo decided to increase the price over what xSeries was charging to supply the boxes.

[Anonymous Coward]

Re: Seriously?

And yet without either a scalable NAS

IBM Storwize V7000 and Storwize V7000 Unified Disk Systems

IBM SONAS

[Anonymous Coward]

Re: Seriously?

In IBM, we call SONAS as "So-Nasty" or "SonASS". It has not worked. A prototype developed from IBM Germany but was rushed to become a project so executives can say something is being done on the world of scaled out NAS. Typical IBM failure. Talent has left the building and the only way for IBM to have new products is to buy a company like Storwise and TMS.

[Anonymous Coward]

Re: Seriously?

In IBM, we call SONAS as "So-Nasty" or "SonASS". It has not worked

It worked when I was an IBM'er.

[Cast]

Re: Seriously?

how much has it sold since they released it? hehehehe. it worked when you had 10 engineers watching one box to make it work.

[Anonymous Coward]

IBM's storage "strategy" was to turn SW products into HW

What is v7000 if not SVC's SW manifested as a HW appliance?

What is SONAS if not GPFS SW manifested as a HW appliance?

Not only did IBM not recognise the Software-Defined-Storage trend, it went so far as to take SDS front-runners like SVC and GPFS and bury them in the ground with v7000 and SONAS.

This strategy failed miserably with IBM not being a player in the SDS world nor being a very successful HW player either.

[Anonymous Coward]

Re: IBM's storage "strategy" was to turn SW products into HW

What is v7000 if not SVC's SW manifested as a HW appliance?

What is SONAS if not GPFS SW manifested as a HW appliance?

Ummm that's true of every manufacturers storage arrays. You might as well say what is DS8000 if not software defined storage (SeaScape).

The entire point of a hardware storage array is being in control of the performance the storage platform delivers. You can run the SeaScape code base on any old RS6000 system, but the performance of it will vary depending upon the pyhsical hardware it contains, and is attached to. Which is why IBM only sell it as a complete and integrated storage platform.

As some have pointed out the same is true for the SVC code base, but if you run it on a couple of dell pizzabox servers don't expect IBM to care what performance you get out of it, or what you're going to do about recovering your data if it all goes tits up.

[Solmyr ibn Wali Barad]

Re: IBM's storage "strategy" was to turn SW products into HW

"Ummm that's true of every manufacturers storage arrays. You might as well say what is DS8000 if not software defined storage (SeaScape)."

No, not always true. Big monoliths use a lot (as in hundreds of pounds) of custom-made hardware. This hardware plays an important part.

Just a small example - should that fabulous software stack crumble, then dedicated RAID controllers are unaffected and will keep the data intact. But RAID logic built into the main software, as most midrange boxes have, will crash with the software. Then there are bugs - memory boundary violations, cache sync issues, and other niceties. If the important bits are isolated by the iron curtains, it certainly does help against the data-munching bugs.

[pPPPP]

Re: IBM's storage "strategy" was to turn SW products into HW

That custom-made hardware has firmware running on it to provide the actual functionality. It can fail just like any other software. Nobody actually does hard-coded hardware as you wouldn't be able to patch or upgrade.

The software running on black-box hardware products isn't a single process you know. Yes, there is usually a Linux or BSD kernel under there and these aren't entirely bulletproof as has been demonstrated a number of times. But the code itself is a number or separate pieces of code which are designed to interact with one another.

Any product designer knows, like any architect, that both software and hardware will fail. With hardware the only real way to deal with this is to provide redundancy. In software, you can do this too, but the code itself can provide routines to deal with exceptions. Some are sophisticated, others simply turn it off and on again.

[Solmyr ibn Wali Barad]

Re: IBM's storage "strategy" was to turn SW products into HW

"That custom-made hardware has firmware /.../ It can fail just like any other software."

Yes, it can fail, and does. In its own little tincan. Where it probably cannot bring the whole house down.

Separation of functions, or shall we say 'sandboxing', is important. Packet storms against the external I/O ports should not be able to crash the main code, especially not all instances at once. Invalid pointer in the main code should not be able to overwrite the data cache. Yet both of these examples are real. And hard to avoid completely in the flat memory model.

So it is quite wrong to claim that *any* storage system is just two servers + software. Like Storagebod did in a cocksure way ("Shark is, and was, simply a pair of RS/6000 or pSeries boxes"), and obnoxiousGit in a more cautious manner ("that's true of every manufacturers storage arrays"). No, it is not true for Shark, nor is it true for Symmetrix. Do not belittle those ancient behemoths, please.

[Javapapa]

The clue to understanding IBM's behavior

They long for the days of 500 fortunate customers and clueless governments who will pay 40% profit margins. You need that to justify the fleet of bizjets at White Plains, NY.

Holmes in honor of Cumberbatch.

[zb]

Or perhaps if they concentrate on software and services...

The rest of the article was padding.

[Spookybuz]

Storwize (V3,V5,V7) does NOT run on xServer- they just use x86 CPU (What kind of CPU is in a VNX?)!

XIV does NOT run on xServer- they just use x86 CPU!

DS8000 does not run on xServer- here, IBM uses POWER.

Tape does NOT run on xServer either.

Their new Flash840 ...guess what...

The only product that uses a mod. xServer is SVC- but how cares? It is HW and if Lenovo will not build it (after IBM design guides) maybe foxconn or some other will do...

[Anonymous Coward]

There's actually no real reason why it needs to run on x86, just like there's no real reason why it needs to run on System X hardware. The only reason it ever did is because it was built elsewhere within IBM and it seemed like an obvious option. I suspect that many of the features of that hardware were overkill for SVC purposes and being able to pick and choose from elsewhere in the industry will free SVC from its shackles to some extent.

[Anonymous Coward]

BS move

IBM dumped PC because they said it was not profitable and they were last in the US then in 3 years Lenovo becomes #1 in US market. So is it that it is not profitable or that the current IBM executives need to be terminated. I believe it is that latter they are substandard management that don't want to do any work and are looking for the easiest why to show a temporary uptick in the stock to get their bonuses and slink off. Terminate all senior management

[Anonymous Coward]

In China

They can do it all so much cheaper in China. Lenovo will make a killing in their domestic market. Whether they can then sell into the West (especially in Govt) is questionable. Look at how Huawei has been banned by many governments.

[Hans 1]

I would like to know the impact this will have on Power, AS/400, and Mainframe sales. I suspect they were giving away quite a few of the x86 servers to their customers in exchange for big datacenter deals with AS/400 & Mainframes etc thrown in ...

[Ken 16]

Yeah, totally chasing those AS/400 sales!

[Barry Whyte]

Software needs Hardware - but good software can run on any hardware

(Disclaimer - I work for IBM in the SVC and Storwize development team)

Quite a few comments here, the first I would contend with is the statement above that all IBM can do is buy in technology like TMS and Storwize. Well, if you actually knew what you were talking about, you would see that the 1% of IBM's revenue last year that came from SVC and Storwize was 100% organic IBM development from the last 14 years of work from the team I work in.

Yes, we re-used the Storwize name, but the product range is based off of SVC software.

SVC is a software product and always has been - its part of IBM's Tivoli group, the Storage Software Group - but all software needs some hardware to run on.

We've looked many times at running SVC software on anything, but the main reasons we don't do this are performance and reliability. You need to guarantee certain performance levels, certain response time aspects, and ensure you can report when something has gone wrong. For now, thats why SVC software runs on specific SystemX hardware, and the Storwize controllers.

At the end of the day, just like Printers, Disk Drives, Laptops, POS, and now x86 hardware, these all become commodity items, that a vendor who is geared up to produce millions of them, and has routes to marked for the masses are in a much better position than IBM to increase sale and make profit.

Storage is still a growth area, and IBM has some great technology in SVC and XIV that have proved they can scale and grow with customers right from small 12 disk systems to many tens or hundreds of petabytes.

All Storwize product hardware is based on x86_64 hardware, but its custom built planars to fit in the form factors needed, and had no SystemX components. XIV is a custom "storage rich server" that again has no SystemX components. Only SVC used a vanilla (with a little metal-bending) SystemX platform, and more often than not, some of the server components just got in our way - like IMM's service processors and the like. With the sale of SystemX, yes we could goto Lenovo in the future and get a standard server from them, or anyone else, or we could tender for a specific server planar that has just the bits we want, and none of the bits that get in the way, or cause additional development to work around.

SVC software has run on over 15 platforms in the last 10 years - including MIPs - so the actual base hardware is almost irrelevant to us.

All that said, the roadmap for SVC and Storwize is rosey, growth is meeting target and some of the exciting things we are working on at the moment will take the bar to the next level.

[Anonymous Coward]

Re: Software needs Hardware - but good software can run on any hardware

All that said, the roadmap for SVC and Storwize is rosey, growth is meeting target and some of the exciting things we are working on at the moment will take the bar to the next level.

And you can't manage a better hint(s) than that?

[pameron]

As an ex IBM employee and now reseller of IBM technology I strongly agree with the position Barry has taken. The crown jewels in IBM's Storage portfolio (SVC / Storwize / XIV) are successful and have a loyal customer following because of the way IBM has taken standardised hardware components (disks, CPUs, Networking etc) and integrated them with software in a design that creates a highly competitive set of products. Sourcing components from a company like Lenovo that has already risen to World No 1 in PC/laptop sales in the last year makes good economic sense. The OEM agreement to sell some of the Storwize product back to Lenovo also enables IBM to scale the product, keeping costs down. Let's also not forget the genius that is GPFS...another Jewell in the crown. If only IBM had got the packaging of this software asset for scale out NAS sorted, this would be a much bigger player in the market today. That said, GSS and bespoke offerings that Partners like us can put together are still credible and cost effective alternatives to the EMC offerings we see in the market today.

Finally...to address the question of whether IBM will sell of Storwize next....my view is that this is inconceivable. I can tell you with some knowledge that the gross margins of products like Storwize do not fit in the same category as xSeries servers. If IBM can also scale production through its alliance with Lenovo, then I think IBM can enjoy margins that support the business strategy. Let's also not forget that the really clever thing about IBM's Storage portfolio is the software....a key strategic direction for IBM.