EMC’s XtremIO is crapping on the badge; it’s an immature ball of destruction that shows how much architecture matters. Or so my favorite storage bloggers say. But customers and resellers seem to have a different take on the destructive XtremIO 3.0 update: They don’t care. Not at all.
Although the numbers haven’t been released officially, I’m hearing very, very strong sales for XtremIO both before and after the hubbub about the XIOS 3.0 upgrade. How is it that such a flawed product can still be selling so well?
XtremIO Outsells All Flash
First, let’s get one thing straight: There is no all-flash array market, despite what Gartner’s Magic Quadrant ™ tells you. XtremIO plays in the high-performance primary SAN array market alongside notables like Nimble Storage, Pure Storage, HDS USP, HP 3PAR, and EMC’s own Symmetrix VMAX. It’s not an “extreme-performance” array, nor is it particularly inexpensive or scalable. Until this update it was significantly lacking in features, but customers didn’t seem to care.
XtremIO is a pretty good product backed by the best sales force in the storage industry. It’s been aggressively pushed to customers and they’re eating it up. No doubt they’ve bought into the promise of predictable performance and believe that EMC will support and enhance the product. So they’re willing to buy what is in truth (if not numerically) a 1.0 product.
This is why it’s not shocking that XtremIO has apparently racked up twice as many sales this year than every other “all-flash” array combined. You read that right: XtremIO has outsold the entire alleged all-flash array market by a wide margin. This just proves that “all-flash” isn’t a market. It’s just another storage array, people!
But what about all the Sturm und Drang about data destruction? That news apparently reached the ears of resellers and support staff way back in May at EMC World. Although many of the techies are said to have been concerned by it, sales continued apace. Customers haven’t deployed XtremIO in production “raw”: They’re either still testing it or deploying a few behind a VPLEX or other virtualization product and performing a rolling upgrade to get all the XIOS 3.0 goodies.
Far from tarnishing the badge, XtremIO has given EMC an up-to-date product that competes well with the upstarts. And in a conservative world like enterprise storage, that matters a lot more than a disappointingly destructive upgrade process. I expect to hear EMC trumpeting this sales success like crazy once the numbers are in.
Stephen’s Stance
It’s somewhat disappointing that customers don’t care more about smooth upgrades. But who can blame them? Storage isn’t sold on technical merits alone, and EMC has always been the king of the sell. The real story isn’t this destructive upgrade; it’s the fact that XtremIO is rapidly eating into Symmetrix VMAX sales! Far from being a failed product, XtremIO is EMC’s future hope in the high-end SAN array market!
Disclaimer: Let’s get this out of the way from the start: I’m no EMC sycophant. Not by a long shot. But I call ’em how I see ’em. Every company mentioned here (apart from HDS) has sponsored my Tech Field Day event so I don’t think I’m biased financially. EMC gave me their EMC Elect thing this year and flew me to London for a big launch event, so take that for what it’s worth. One more thing: I have no NDA and no official inside information about any of these products or sales numbers.
Stephen
A few additional thoughts from me;
1. EMC sales people will sell their own grandmother if the margin was better than on VMAX. Same goes for XtremIO. It’s not about the technology or doing the right thing to them, it’s about the margin they make on a deal.
2. Q3 & Q4 figures this year will give us an idea of what to expect. Chad has already “spilled the beans” so we know numbers are expected to be good. Q3 is typically similar to Q1/Q2 for the Information Storage division. Figures for 3Q2013 were $3806 million at 55.3% margin (product + services); anything less will be a failure. Tracking back over the last few years this division’s figures have been completely flat. Will XtremIO alone change this headline number?
3. You are right, there’s no such market as all-flash arrays. We may be able to categorise by application or workload (e.g. capacity or high performance) but focusing purely on all-flash as already discussed it a red herring from Gartner and others. It’s all about requirements.
4. Clearly the days of the storage admin/architect are definitely over. It’s quite astounding that companies don’t care about disruptive upgrades. Let’s see what their opinion is when they are managing 20,30 or 40 separate non-upgradeable arrays. Remember, NetApp technology was seen as revolutionary and wonderful, until you had so many boxes the management overhead became a nightmare.
5. If EMC see XtremIO as the high end replacement for VMAX, then they have a LOT of work to do. Features, functionality, resiliency, are nowhere near – today. However maybe EMC is betting that in 5 years time, XtremIO will be capable and the slide from VMAX to XtremIO will be a gradual one of attrition.
1. As to your first point I would like to clear up your misconception. You realize that sales is not some monolithic group I would presume and are engaging in hyperbole for comedic effect. I would let it go, but I have seen so many good people give so much of their time and give themselves ulcers, practically killing themselves trying to do the right thing for the customer that I wanted to comment.
I am an engineer not in sales. However, EMC reps have not been paid on margin since the 90s. So your statement is factually incorrect. Like almost all reps in the industry they are paid primarily on revenue. So ironically selling a smaller AFA is probably not in their financial interest and yet they have done a ton of it to give their customers the best solution. The vast majority of EMC and competing companies do their best to do the right thing and give their customers the best technology for a given situation (from their perspective). The fact that EMC has a portfolio of products for different use cases actually ensures the customers have choice. Many of the players believe a single product can cover all situations. On a fundamental level EMC is structured to allow a choice of the best technology. (Scale-out NAS, Scale-out SAN, Unified, AFA, Hybrid, EVO:RAIL, Software-Defined etc. on the micro-level and the Federation: VMware, Pivotal, EMC II on the macro)
Are there bad actors at every company who do things for the wrong reasons, probably. I am sure you also realize that most of the reps at the small companies and storage divisions of other companies have worked at EMC and vice versa.
The bottom line is everybody wants to like who they see in the mirror and do their job well. So to make some blanket characterization about a group of people is probably not good. Also it seems a little arrogant to think that customers are so stupid and naive that they are bamboozled by some EMC sales magic. There is no magic. Every single deal is competitive and all the vendors are in there. Customers are not idiots they weigh business, financial, and operational benefits and risks and then decide.
2. Waiting for the numbers.
3. Agreed, It is all about the requirements (thus EMC portfolio of products)
4. Agreed. Obviously they see the rise of converged infrastructure. Perhaps you should check out EMC ViPR and other efforts toward simplifying and automating (VCE, EVO:RAIL, EVO:RACK, ViPR SRM etc). EMC is not only aware but leading the charge.
5. It’s all about choice and requirements. Depending on those it could be any of the platforms today, not 5 years out. Five years is an eternity in storage. We might all be talking about phase change memory, memristors, quantum arrays etc.
This is all just my personal opinion and I don’t speak for the company.
How is it that your comment is more eloquent than my whole damn post?
JDG, you’re right and I apologise for being slightly facetious, but you get the point. Even if it’s not a margin incentive, there will be additional incentives applied to sales of XtremIO to boost the numbers.
Now, that’s not to say that’s a bad thing. EMC are in business to make money, like us all. Customers just need to be aware of everyone’s drivers. Oh and agreed, there will be many sales people who do the right thing.
Chris
Not quite, Stephen! Some times writing stuff is easier than other times. Always best when its from the heart and not forced. 🙂
There isn’t an EMC rep alive that would prefer to sell an XtremIO install over a good VMAX install. This is customer demand, plain and simple. And I’m sorry, but there is absolutely an “All Flash Array” market. While there is overlap for sure against both mid-market and high-end arrays the use cases, capabilities, and features vary greatly across them. The erosion of the higher end arrays is expected. People that bought them just for high performance have no need to do that now. This supposedly non-existent AFA market is filling that gap. Now you buy a VMAX for 6-nines and advanced data services.
Jason
Our XIO just shit the bed for the second time, immature is an understatement. The folks at tech support are slow and our fix now for them not being able to present half the VMWare luns that were running just fine for many days is to upgrade the firmware and hope for the best.
Where is this XtremIo revenue number coming from? I’m calling BullShxx on it. I’ve seen in the Forbes report that they hit $500m…ah but if you read the fine print it says “on an annualized run rate”. Know what that means? It means they extrapolate (make the number up using your favorite parabolic function) from now until the next 3 quarters.
I also see in their latest report that “EMC Emerging Storage” hit $580m for the Q3, but that includes Isilon, Atmos, and XtremIo.
The SolidFire storage cluster is beating XtremIo on almost every POC because XtremIo is behind in technology and integration. XtremIo is winning when they have the incumbent deals and don’t have to go through POC. Still the financials from EMC don’t seem to support what you claim here.
EMC can and will give these things away though, and that’s what they’re doing with their most loyal accounts to keep intruders out. Yeah customers don’t seem to care about upgrades yet because they’re not taking the platform seriously yet, so they’re not putting any applications on it that they really care about…and if they are, EMC is spotting them a swing box and free services to make them happy. There, now I’ve made my own outlandish conjectures. 😉
Please correct me if you’ve got more accurate figures for your claims and please source them so they can be verified.