IBM's first quarter 2013 results were disappointing, and if you're in storage and servers at Big Blue, it appears the numbers guys are focusing on some of your product lines. In the earnings call, CFO Mark Loughridge said: "There are parts of our business that are in transition or have been under-performing like elements of our …
Of that list the "At Risk" products are:
DS3500 (V3700/V7000 is the replacement)
DS5000 (V7000 is a natural replacement here)
N Series (V7000U and SONAS will replace)
Safe products should be:
SVC (Although I would expect a V10000 ish system with 8/10 Core CPU's and 240/480 12gbps SAS Disks + Clustering to replace it eventually)
Tape (TS3500 stays for Mainframe, TS-1140 and LTO-6 is not a problem if tape stays and IBM have a reasonable OEM business in selling bare tape drives)
ProtectTIER (Someone needs to take on EMC/Data Domain and HP is not getting it done)
SONAS (Needed to replace big NetApp boxen from a CIFS/NFS perspective, maybe a "Gateway" that can attach to XIV/SVC/DS8000 as a future SONAS offering ala VNX VG2/VG8 Gateways)
My $0.10 (Inflation is a bitch)
"....DS8000 (Mainframe)...... Tape (TS3500 stays for Mainframe...." Isn't the overall trtend in mainframe downwards?
I think he was referring to storage as the product line "in transition" (i.e. moving from DS3,4,5 to V7000, XIV). The LSI rebrand DS systems, all except for DS8, are obviously tanking because they are going away, the new systems, XIV, V7000, are doing well.
"Isn't the overall trtend in mainframe downwards?"
Not really, they were up 7% on the worst quarter IBM has had since 2005.
Re: AC Re: Mainframe?
"......Not really, they were up 7% on the worst quarter IBM has had since 2005." Gosh, would that be the quarter of new production introduction? One quarter a recovery does not make. Overall trend is down, as pointed out here in Register articles before.
Re: AC Mainframe?
"Gosh, would that be the quarter of new production introduction?"
No, the zEC12 was introduced last year.... There were only two years between z11, i.e. z196 and z114, and EC12, so many people have not finished their depreciation on the z11 generation. The BC version of z12 has not even been released. z never fluctuates up or down very wildly. The hardware, physical hardware, might go up and down with releases, but the MIPS is pretty steady for decades.
Re: AC Re: AC Mainframe?
Oh boy, the denial is strong with this one!
"....but the MIPS is pretty steady for decades." So, stop for a moment and think carefully - if IBM are introducing more powerful machines in each generation that can do more MIPS each, but the number of MIPS is staying steady, that must mean IBM are selling less and less machines, which means less income from mainframes. Either that or you really want to contend that IBM are selling they customers new machines that actually have no perfromance advantages over the old versions. So, which do you want to go with? LOL!
Isn't the DS3500 just an OEM'd model from NetApp/LSI Logic anyway? The GUIs to manage them are slightly different, but the DS3512 and 3524 report their controllers as LSI Logic, and you use LSI's multipathing software on Linux to talk to them...
That said, the storage line does need to be simplified a little bit. Especially in the middle tier, they have a lot of slightly different models that could be combined to save costs and improve focus. Hopefully storage doesn't get sold off like the servers and PCs...
Matt, The Mainframe business is low revenue but high margin business and you will pry it out of IBM's cold dead hands. It is a no brainer for them to keep pushing as they have 80+% market share and a captive audience as the path to move off a mainframe takes YEARS.
Erik, Yes the whole DS, DCS and N Series range are OEM'ed from NetApp/LSI and that is why they will go, the StorWize range is resigned to replace them, the new V3700 is the box to replace the DS3500 as the V7000 replaced the DS3950/DS5020 and the DS5100/5300 ranges. The only product from that line that might survive the end of the year is the dense DCS3700 for Video etc.
IBM will get rid of as many OEM'ed systems as they can, they want to own their storage destiny and they have either acquired the products to do it (XIV, TMS) or built them in house (StorWize, SONAS).
HP are doing the same thing of out with the old (EVA, P2000/MSA2K, P9500/XP) and in with the new (StoreServ 7000/3Par, StoreVirtual, StorEasy).
wasn't storewize an acquisition ?
though maybe was another co..
SONAS does that still primarily use DDN hardware with GPFS on top?
Storwize was an acquisition, but it is not all that it is in the V7000.... IBM marketing apparently just liked the name. The V7000 is essentially SVC, IBM's storage virtualization engine with other software features, including Storwize compression, with internal disk/SSD... although you can also virtual third party arrays.
SONAS has different underlying hardware options. Basically XIV for large scale, V7000 for lower scale or more nodes with less capacity. You can run GPFS on anything, e.g. x86 servers if you want a Hadoop type architecture. I don't believe IBM uses DDN for any of it.
Re: Michael Duke
".....HP are doing the same thing of out with the old (EVA, P2000/MSA2K, P9500/XP) and in with the new (StoreServ 7000/3Par, StoreVirtual, StorEasy)." Ugh, don't remind of that awful new hp naming convention! That is what happens when you have too many marketing bods running riot. Besides, StoreVirtual is not new, it's just the latest version of the LeftHand software; StoreEasy is just a re-jig of the old D2D range on new Proliant servers; and hp have kept the XP arrays as they seem to do quite well as (drumroll) storage for IBM mainframes, something hp can't seem to be bothered to make available on 3PAR (which probably hints at how small a market mainframe storage is now).
Re: Michael Duke
Yes StoreVirtual is the new gen8 servers running V10 of the LeftHand software.
StoreEasy is the NAS range, running Windows Storage Server 2012 and or IBRIX depending on model.
StoreOnce is the D2D re-brand.
P9500 is HP's mainframe offering BUT they have single digits of the mainframe storage market, IBM and EMC have like 80+% between them and the Hitachi vendors (HDS, HP and formerly Sun) have the rest.
it is no surprise that HP do not want to engineer FICON into the P10000/V Class.
Re: Re: Michael Duke
".....it is no surprise that HP do not want to engineer FICON into the P10000/V Class." I think that's more of a case of hp not being bothered with mainframe seeing as they seem to be struggling to keep up with the demand for 3PAR!
don't compare XIV with SONAS
Yes, both store data, but other than that they are 2 totally different products.
From the headline, I thought maybe IBM was going to fire those of its engineers who aren't able to change the laws of physics. IBM invented the hard disk and set the standards for magnetic tape for a long time; and twice it's been responsible for significant innovations in PC hard disk density.
But this isn't about their cutting-edge activity; it's about routine disk servers, and for that, of course the cheapest adequate product will win for most customers, even if IBM's traditional reputation for and emphasis on reliability keeps a loyal customer base.
If I were IBM, I'd be content with that niche rather than damage the brand by trying to compete in the cut-price market, especially given IBM's exit from the PC field. The alternative will leave those seeking high-reliability storage with no place to go. But, on the other hand, neither IBM nor any other company should ever be content with selling product not much better than the next fellow's at inflated prices, so a critical examination of their server line-up is not a bad thing - as long as it has a sensible plan behind it.
IBM is about high margins...
Telling IBM to be content with everyday stuff that has low margins isn't listening to the company's goals. Higher margin products can be a good value ONLY IF they have special features or exist in special markets.
Of course, such product ARE poor values, otherwise
Re: Beachrider Re: IBM is about high margins...
Whilst I don't disagree that IBM is indeed all about high margins, the economic reality is that the market is trending towards more and more commodity hardware and software, which means the opportunities for higher margin offerings are reducing whilst those for commodity and lower margin ones are growing. Therefore, sooner or later, IBM will not be able to make enough money to carry on making only high margin offerings, at which point it will have to try re-entering the lower margin sectors or do a Sun and die.
A direct comparison would be that the UK and many European countries used to have hundreds of small companies that made expensive and specialised car bodies and interiors to order, called coachmakers. Often they did little more than add a fitting service to another vendor's car, just as IBM will be trying to do by offering services on other vendor's x86 and storage kit. The vast majority of those coachmakers have gone bust and the few survivors have shrunk to tiny companies. That is the future for IBM with their retreat into higher margin products.
Re: Beachrider IBM is about high margins...
You can disagree about IBM's mission being successful, but that is their mission.
They will not go the way of Sun. There are other ways. Remember that IBM got into the business machine business by making clocks. All the other clockmakers went commodity. IBM might fail, but they will fail with a novel approach (unless..)
Re: Beachrider Re: Beachrider IBM is about high margins...
Yeah, and Sun got into servers, software and storage by making workstations. All you're doing is reminding me of the Sunshiners that used to post here about how Sun was too big and innovative to fail, yet the fight against commodity took Sun from a $200bn marketcap to less than $4bn when Oracle finally put them out of their misery, all in less than ten years.
It seems all is not as rosy in IBMville as Ginny has been pretending. She can get away with selling the x84 server biz by claiming it's low margin and shrinking due to virtualisation, but storage is both a growth area and has better margins than servers, and IBM needs to be jostling for the lead in the sector. Or maybe Ginny thinks she can claim that selling the storage biz to Lenovo is another "success"....
IBM agrees that they want to be 'in storage' did you see something that say it's gone?
Re: Beachrider Re: Hmmmmmm.
"IBM agrees that they want to be 'in storage' did you see something that say it's gone?" Take off the Big Blue blinkers and read the article.
It's all about....
Licensing! There is a huge price tag attached to this. For example, let's look at their TPC product for managing storage. Want to buy it so that you can manage your XIV's, SVC, etc. from a single console? Good luck affording it. An XIV is licensed on a per drive basis (raw TB capacity) completely contrary to the way the XIV is sold.
Another product, SONAS. Their attitude is "don't bother us if you have less than 150TB" - go buy an N-Series, which has its own set of problems with the Cluster-mode junk NetApp is working on. No wonder SONAS, which is so promising, is not being called out; another waste of an excellent product.
They need to learn a lesson from their partner: NetApp. Pay a HUGE price upfront and don't worry about the TB licensing. Of course, they need to lose the NetApp behaviour: charge full price for their OnTap software every time you want purchase. Heck, even learn from the XIV guys: no TB licensing costs.
Oh and support....good luck with that. Is it sofware or is it hardware? I don't think they know what it is! And, if they screw up your records, its the customers "fault" to prove to them that you are a customer and that you are paid up for support.