IBM To World: Hey, We're Still In The Game!

Today, IBM did one of those mega-announcements around information, storage, security and related topics.

Now, before you think this is just another vendor-bashing competitive blog post, let me assure you it's not.

There's a lot to write about here, and -- for the more-than-casual-observer -- it provides deep insight into how IBM thinks about things, and how they do a pretty good job marketing the cards they've been dealt.

A Bit Of History

If you're a denizen of the storage blogosphere, you probably realize there's been a fair amount of criticism regarding IBM's unusual behavior surrounding some of its newer storage products. Beth over at SearchStorage has a useful perspective, if you're interested.

We were pretty sure at some point that IBM would use their brand and their credibility to do some sort of from-on-high omnibus announcement.

And now, with your permission, allow me to dissect this announcement from an industry insider perspective.

It's All About Information Infrastructure For IBM

I suppose that imitation is the most sincere form of flattery, and as a marketing guy for EMC, I'm feeling very flattered indeed by the IBM storyline.

They've chosen the term "information infrastructure" to describe what they do around growing information management challenges.

EMC started using that term to describe what we do for a living at the end of 2006, so I guess it's catching on a bit :-)

They also set up the problem using IDC-like stats without using IDC's material: growth of personal "information footprints" over time -- all familar stuff.

So, we all agree that information is pretty important stuff, it's growing like topsy, and we're going to use the term "information infrastructure" to describe an overall approach to managing and leveraging the stuff.

So far, so good.

Look To The Labs!

IBM spends a boatload on R+D in all sorts of areas, and they're well known for it. But, after watching them for more than a decade, I've noticed that they tend to point to all the money they spend on R+D when they're a bit short in the actual shipping product department.

Actually, you should go read the IBM Research press releases here and here. While you're reading it, please keep in mind some of the great technologies that came out of that group that really didn't go anywhere in the marketplace (StorageTank and storage "bricks" come to mind as quick examples -- remember those?).

Me? I don't give anyone credit for any of this stuff until a customer can actually buy it. The rest is just posturing, IMHO.

Andy Monshaw's quote seems to mention what might be possible -- rather than what's here today:

"The world is re-tooling its underlying IT infrastructure in a dramatic shift away from a decades-old client/server model to a radically more efficient Internet-style architecture. This requires different thinking and new capabilities, which we are addressing in this information infrastructure launch, with our investments going forward, and how IBM will do business with our clients," said Andy Monshaw, General Manager, IBM System Storage. "There is no bigger opportunity for our clients than to unlock the value they have in their data centers and help them create smart, innovative offerings for their end users – the consumer. IBM is the only company in the world – not HP, not EMC, not Sun -- with decades of research, industry knowledge, market leadership and the end-to-end capabilities to make this a reality for our clients.”

And, in a related approach, when you've got a key product gap, announce that you will ship something in the future, rather than actually shipping something:

"In the emerging area of information security and encryption key management for storage – where information is “locked” and can only be accessed by users who have “keys”, IBM will release new Tivoli Key Lifecycle Management software, which helps automate the management of keys where disk and tape storage devices cannot be compromised if lost or stolen."

Never mind that EMC RKM (RSA Key Manager) has been shipping for a while, and is already reasonably integrated with multiple encryption technologies from multiple vendors.

OK, we all can understand gap-covering in a competitive portfolio by a bit of skillful choice of words. From my perspective, no harm and no foul -- vendors do this all the time.

But there's even more interesting perspectives that can be gathered from some of the details.

Finally, We Hear About XIV and the DS5000.

IBM's new XIV-based platform has previously been positioned separately as "web 2.0 storage" as well as "high-end storage". If you read the text carefully, it seems that IBM is positioning it as a bit of both. I'm sure that's not the end of the story.

Further confusing matters, the new DS5000 is positioned in the same breath -- although it uses a more traditional architecture. Bottom line? It's still not clear where these bits really will fit in -- and we'll probably hear more from IBM on this in the future.

The DS8000 is not quite dead yet, it seems -- it gets RAID 6 for (presumably) support of larger disk drives at some point in the future. No terabyte drives yet, and -- of course -- no enterprise flash.

And, no, please don't ask me to position the RAID 6 in the DS8000 against the RAID 5 in the DS5000 and the "RAID X" (not really RAID) in the XIV. For that, you'll have to ask IBM.

IBM Loves Tape

It's amazing how much of IBM's "information infrastructure" is based on tape: new(er) tape drives, new(er) tape libraries, compliant storage "requiring" tape (???).

IBM is relatively strong in tape offerings, hence it's got to be part of the agenda -- regardless of broader industry trends. Now, I'm not saying that tape isn't part of the broader picture (it is), but it seems to be heavily overweighted in the IBM storyline from my perspective.

My primary beef with tape? It's really hard to make tape-based archival information part of active business processes, like search, customer services, etc. It's just so dang slow to get to the bits you might want.

And with dedupe and spin-down, it's hard to still make the same old pro-tape arguments of yesteryear.

It's All About The Investment?

Finally, IBM makes much of their "$2B in R+D and acquisitions over the last three years".

Simply put, I don't think that's enough guys -- really.

As a comparison, I think the comparables for EMC are something like $4B in R+D, and $8B in acquisitions -- roughly 6x what IBM is spending in this space.

And, despite how much money EMC spends on all of this, we've still got a wish list a mile long of stuff we'd love to do -- if only we could spend a bit more :-).

At this rate, I don't think IBM is spending enough to reach critical mass, let alone close the gap between themselves and EMC, for example.

And, Yes, There's Financing Available!!

Taking a page from US automakers, there's new attractive financing and incentives available for customers who want to go down this path with IBM. My interpretation? Sales are sluggish, so a little extra juice might be required.

Given IDC's recent announcement of just how fast this marketplace is growing, you'd think IBM would be finding plenty of willing customers. But I guess some extra incentives are needed ...

The Bottom Line

IBM does great storytelling -- this press release is a great example. And I'm somewhat flattered that they've decided to incorporate a majority of EMC's storyline as a basis of their own.

But telling a great story and delivering great solutions are two different things.

It's clear from this announcement that IBM desperately wants to be seen as "still in the game". And, at least from my perspective, it looks like EMC is framing much of the discussion.

But, frankly, I think IBM is going to have to do much more -- and do it much more quickly -- to still be seen as any sort of lead participant in this market.

[Update: you think I'm brutal? Go take a peek at how Chris Mellor savages IBM ... sheesh!]

Comments

Hi Chris,
I follow your blog very closely and felt the urge to write to you today. I do agree with your article about IBM but can i ask why most of your comparisions and articles speak primarily about IBM,Netapp and HP while nothing is mentioned about HDS and their announcements? I am sure you see them as a serious competitor!By the way, I am not associated in anyway to HDS

Posted by: Premkumar Subramanian | September 09, 2008 at 08:02 AM

Hi Premkumar ...

The reason that I don't talk much about HDS is that there's not much to talk about. Their strength is a single product (USP-V) which is primarily about a single feature (storage virtualization using the array).

As such, there's not much that they're doing that I find interesting, unusual or really that worthy of discussion. That being said, they seem to just keep chugging along, selling big arrays to their customers.

Plenty of other stuff worth talking about before I get to them. Now, if they ever do something interesting or controversial, you can be sure that I and a few other EMC bloggers would comment on it.

Thanks for writing!

Posted by: Chuck Hollis | September 09, 2008 at 10:12 AM