El Reg drills into IBM: The storage biz's got that sinking feeling

Chart

Am I the only one finding the chart a bit strange? Would it not have been better to show each quarter on the X axis? Is this a new accounting trend in charting?

Additional Detail

Hi Chris, I have mentioned this a few times, but it is worth repeating. In the IDC charts they show both revenue share and capacity share quarterly and annually. There is no question that IBM is losing revenue share, but in calendar year 2013 they actually gained capacity share according the IDC, so again the devil is in the details and it would be if both sides of the story were told.

The question really becomes what is more important, how much revenue share a vendor has or how much capacity share they have. If you are an investor, revenue share (and profit share) is most important, but if you are a customer, as long as the vendor is financially viable, do you care that they made less money on the capacity they sold you? Probably not. If I were a customer I would see capacity share increases along with revenue share decreases as a bonus because you are getting the same product at a lower price. An alternative question would be if a company is gaining capacity share even if they are losing revenue share are they still not selling more storage?

One of the most dramatic examples of this is EMC compared to NetApp. While EMC was #1 with 34% or $7.2B in sales and NetApp #2 with 15% or $3.2B in sales revenue share in 2013. EMC was #1 with 28% capacity with 6.2PB share compared to #2 NetApp's 20.3% with 4.5PB. Net is EMC customers paid an average of $1,161/TB and NetApp customers paid $710/TB. From the previous year although EMC gained 0.7% revenue share, they actually lost 0.3% capacity share while NetApp gained 0.6% revenue share and also gained 0.3% capacity share a net gain of 0.6% capacity share even though EMC charged their customers more.

In IBM’s case, their revenue went down, but capacity shipped went up showing that they are actually selling MORE storage, but their customers are just paying less for it or have shifted to some of IBM’s less expensive products, probably both. For 2013 IBM $2.3B for 11.1% of revenue share and 2.5PB for 11.5% of capacity share at an average of $920/TB almost exactly between EMC and NetApp.

Form a customer’s perspective again what tier one vendor do you want to buy from? The one that whose acquisition cost/TB are going up like EMC or down like NetApp and IBM?

As a piece of trivia, IBM invented the HDD in 1956, doesn't seem like hard drives have been around almost 60 years.

The problem when working with raw capacity shipped is that it only tells a bit of the story; you would need to know what the capacity breakdown is before working out what customers are paying per terabyte. If EMC and IBM are still focusing very much on the high-end as opposed to the mid-ground that NetApp currently tend to operate in; for example EMC, IBM and HDS can sell storage into mainframe outfits, NetApp can't...this is going to skew their price per terabyte downwards.

If the various vendors were to start splitting out revenues and shipped capacity by product line; you could draw more conclusions. For example, what has the impact of the Engenio aquisition had on NetApp's capacity shipped? It's given them a whole load of partners that they didn't have before and has increased their channels to market.

Perhaps vendors should start to detail capacity and IOPs shipped...but that'd only be of real interest to the storage geeks.

I actually think IBM have some very good storage products; the Storwize products are very good, extremely good value and the low-end V3700 is so much better than the old OEMed LSI/Engenio products. If Lenovo do take the IBM storage products; the Storwize products could/should do very well. I think XIV has driven into a cul-de-sac architecturally; spindle-sizes have not done it any favours and the current product's inability to increase spindle-counts which is imperative to keep rebuild times down is a real problem long-term.

The DS8K is going to keep trundling along for as long as IBM sell mainframes....which is unfortunate really, it's the zombie product which needs getting rid of.

IBM is exiting hardware biz

The trend has been clear for quite some time. IBM is slowly but surely exiting all hardware biz. Not much hardware is left as IBM is transitioning to a consulting and services company, competing with Accenture etc. The margins are higher than in hardware, which is why IBM sold off the x86 biz. That was huge, as it was IBM who started the whole x86 industry. And now the creator has exited, because of diminishing returns. IBM has no sentimental why on business which explains why POWER is next in line to die. POWER8 will not be faster than high end x86 but much more expensive, so POWER will turn to low margin. That is the reason IBM has no POWER roadmap which is a very bad sign for the POWER customers, they can not plan ahead. Surprisingly Mainframes are still alive but the trend is declining as well. No new Mainframe customers, only old customers are upgrading, or slowly migrating away from the relics. There is no new customer that will build their new business around expensive Mainframes, they prefer Linux and fast x86 servers instead. There will be a time when the last few customers can not bear the cost of Mainframe R&D any longer, just like Itanium. But that will take another decade at least.

RE:The problem when working with raw capacity shipped is that it only tells a bit of the story

Hi Martin, totally agree, since mainframe storage is more expensive than midrange storage, cost per TB and TB shipped doesn’t tell the whole story either. Unfortunately IDC no longer breaks storage down into high end and midrange and NAS or SAN, so the only info available is revenue generated and capacity shipped. My point though is you can at least understand more of the story if you have both the revenue generated and capacity shipped evaluated in an article such as this, not just the revenue generated. Even though IBM generated less revenue, they shipped more capacity and it would have been good to include that vs. just saying IBM is going down the drain because their revenue went down when in fact they may not have lost any customers and their customers just paid less or moved to a less expensive brand of IBM storage.

One additional comment on your question What has the impact of the Engenio acquisition had on NetApp's capacity shipped?” One thing that is important to understand is that the OEM, not NetApp gets credit for those shipments as IDC bases their numbers on Branded revenue and Branded capacity shipped. So even though roughly 35% of IBM storage shipped is OEMed from NetApp, NetApp doesn’t get credit for that in the IDC charts, IBM does.

IBM: "I Been Misled"

IBM is leaving the hardware biz

I have to agree with MadMike. The long range goal of IBM is getting out of the hardware manufacturing game.

"Business Services" pays better. Round robin email, endless meetings and arcane contracts and employee rules are far cheaper than actually having real engineers and knowledgeable techs.

This of course means they will have to change their name to IBS, or iBS. This will more accurately reflect their new offering, BS. (bullshit for those not familiar with the abbreviation)

Because after all, if you can't dazzle them with brilliance, baffle them with bullshit. It's cheaper. And it certainly plays to their current strength.