Exclusive Interview With Richard Smith, Number Two of IBM Systems, in Charge of Storage

As VP WW storage solution sales, Richard C. (Dick) Smith, 53, is number two of IBM Systems behind GM Janie Thomas. He oversees the sales of storage software and hardware for enterprises around the world. He has held many senior leadership and international roles within IBM. Recently, he was GM of sales for IBM’s systems and technology group for Asia and growth markets. He also served as the VP of IBM’s software division for North America, overseeing thousands of practitioners and nearly a sixth of IBM’s global software revenue. Prior to this, he led WebSphere’s global sales organization and has held positions as VP for business partner and SMB sales, VP for the AsiaPac software group with responsibility across five regions, and as WW sales operations director. He is also a member of IBM’s growth and transformation team, a task force of the most senior IBMers who devise key growth strategies across all IBM brands. “I am 21 years at IBM, and before that I was with a long, long gone company Digital, Equipment Corporation (DEC), for ten years. I was an engineer at DEC team and VAXes,” he said. “I am an Australian citizen and an US resident,” he added. Along with Jerry Blakeslee, he published a book titled Strategic Six Sigma: Best Practices from the Executive Suite. His hobbies are technology and skiing. He is graduated from Victoria University of Wellington, Australia.

StorageNewsletter.com: What happens for IBM in storage products with disappointing results since more than three years and losing globally market shares against its bigger competitors (EMC, NetApp, HDS)?
Smith: I need to give you some context. So first and foremost, and this is public information, actually our storage business since the first quarter of 2014, has being sequentially improving quarter on quarter, relative to our revenue.

Growing quarterly, not yearly?
Quarter on quarter we have posted sequential improvement and in the same time we’ve also being transforming our storage business as you would know with a number of significant advances. The most significant was in third quarter last year with the divestiture of our Lenovo, of our ex-business, to Lenovo, as you may be aware. As part of that agreement, we also gave Lenovo the exclusive rights to allowing storage business, our V3000 offerings. So when you look at our results and performance overall, reality is we have being improving our position over the last year and a half, and moreover improving and against the backdrop where we actually removed some significant components of the underlined business. The other thing that would just share with you in strategic areas of growth for us, notably flash. Among others, we are actually the global market leader in flash shipments and petabytes shipped.

In petabytes only, not in revenue?
I’m not going to comment on the term the competition counts its revenue, but i would leave the comment at petabytes shipped.

IBM was an historical leader in HDD drives, tapes, and high-end disk arrays. In which storage category IBM is now number one?
So, we continue to be number one in tape, and in fact our tape business continues to grow quarter on quarter.

Are you in front of Oracle/STK?
Dramatically, yes. We are, as I said the global market leader currently in term of flash shipments.

That’s all?
So, from high-end disk, we believe, and I have an external market perspective on this not an internal one, but high-end disk, we believe that the market share position has remained within 2 to 3 points of the competition for the last 2 to 3 years, so for DS8000 for mainframes.

To try to reverse the situation, IBM spends a lot in acquisitions (Tivoli in 1996 for $743 million, Mylex in 1999 for $240 million, Softek in 2007 for 150 million, Diligent in 2008 for $200 million, Storewize in 2010 for 140 million, and more recently StoreIQ in 2010 and Texas Memory Systems in 2012. That’s more than $1.5 billion. For finally poor results.
We don’t think we’ve had poor results at all. So for example if you look at several of our strategic acquisitions, and I’ll focus on a couple here. I’ll put Tivoli aside here, because Tivoli is an important one, but Tivoli is a software acquisition that we actually made many many years ago. Tivoli has been organic for us for a very very long time.

Yes but it’s a part of the storage activity.
Actually it’s only just being integrated into the storage business. It was previously ran as a separate brand within the software organization, so the products that came across as you would know are Tivoli Storage Manager, VSC (Virtual Storage Center) and the Logs. But let me talk about some of the most specific acquisitions that we are delighted with. So, we talked about our flash acquisition, Texas Memory, and it has being phenomenally successful. We have posted a gain in external figure, we have posted high double digit growth quarter on quarter with that technology, every quarter since its acquisition. We’ve just announced two new products, the 900 and the V9000, which are the next iteration of the flash offerings for us. Now the capability in and around that are offering 2PB at industry leading prices. Now the marketing position, it’s just a statement in fact, our XIV business continues to grow double digit. XIV was a very successful acquisition for us, and in 2010 our XIV business continues to grow double digit every quarter. If you look it as Storwize business, so the V3, the V5, the V7, double digit growth quarter on quarter. Where we have seen contraction, to be very candid with you, is if you take a look at the revenue model, where we have seen contraction and price compression, where I think the industry has seen us, is in the high-end disk space. So if you take a look at the degree of units shipped in average, unit revenue around the high-end disk space, that is a compression that everybody in the industry experienced.

More than that, IBM seems now enable to invent any successful storage technologies and acquire them. Why R&D is now so poor?
I strongly disagree. From an innovation standpoint, we have been developing technologies notably around bringing the software intellectual property and the software intellectual value to our storage demand. So, we have completely read about and announced the software spectrum portfolio, which is all in-house developed and grown, it is a billion dollar investment.

Software-defined storage?
Software-defined storage.

Is that completely your own development?
It is. So, and this is important, because if IBM declared an external number into the marketplace, it is measured, you know, from the analyst standpoint. We are investing at billion dollars in software-defined storage in innovation and development. So with all due respect, you cannot say to me that we are not in innovation and investment, quite the contrary.

I could also say that about EMC. If they need the technology, they buy it. Like you, buying start-ups.
Well, I am not going to comment on EMC. What I am going to say to is that the IBM storage business is a critical business in IBM, and quite frankly you don’t invest a billion dollar in R&D which is higher than any of our competition, higher than any of our competition in R&D and software development in and around the portfolio. No one is spending that amount of money.

Generally when a business is decreasing and not profitable, IBM finally sells it. Is storage business for sale?
No.

On April 1, for the fun, we suggested in StorageNewsletter.com that Lenovo will acquire IBM’s entire storage business for $3.4 billion. Can you guarantee that IBM will keep its storage business in the five years to come?
Well, first and foremost, I’m an officer of the company, so I would never comment on anything that would be explicit to …

So you’re not sure?
No, not at all, you are putting words in my mouth. No, I’m just, I’m not gonna take it… the storage business is not for sale. We are investing billions of dollars, billions of dollars in the model, we would not be doing that if weren’t completely based in growth going forward. And by the way, the article on April 1 was foolish.

It was on April 1, it’s foolish.
It was ungrounded, unbased in fact and…

We will see …
Storage business is not for sale.

What are precisely your relations with Lenovo in storage?
So, Lenovo has the right, the exclusive right, for the sale of our V3000 product and Storwize V3 entry products. We have some ability in IBM to continue to support customers and upgrade, but Lenovo is our primary route to market now for their channel. In addition to that, we have an OEM relationship with Lenovo, where they can resell V5 and V7 storage technology. The V3 is OEM, V5 and V7 is resell.

But Lenovo also worked with EMC.
(…)

Tape is a declining market worldwide.
Not for us.

Is it growing?
Absolutely.

Thanks to the acquisition of Texas Memory, IBM is now one of the leaders in all-flash arrays. Who are your main competitors?
EMC XtremeIO, Pure. That’s probably the top two.

Storwize and XIV are also one of the few products doing relatively well at IBM. But now users prefer open scale-out solutions based on software-defined storage with commodity hardware rather than big proprietary silos. How are you going to react to this trend?
That’s a great question!

Thank you, I hope for a great answer.
Customers can now buy XIV as software and by deploying it against industry standard hardware. So, we’ve taken all of the logic out of the controllers, out of the silos and you can now buy the XIV offering as part of our spectrum offer as software. You can acquire it in the cloud, you can acquire it as a software package and manage your own environment, or, as many of our customers already do, you can utilize it in existing XIV storage arrays.

For how long there will be a market for mainframe high-end storage subsystems because it’s diminishing, for Hitachi, for you, for EMC…?
Actually, if you look at what we’ve just, with the z13, the latest mainframe announcement, we shipped more MIPS of capacity in that z13 launch in the first quarter than we have done in any of the previous z mainframe launches ever. So, I think the attached opportunity for high-end storage, the DS8000 storage, is commensurate with that continued need for z. What we have done with the z13 is completely redesigned it, quite frankly to deal with the intensive in storage demands of cloud and most notably mobile. Because, I mean as you know the transaction rates and the data rates that are being created by mobile are just phenomenal. So, I think the storage opportunity for high-end mainframe continues to be very significant. Having said that, average unit revenue around that segment has contracted.

Why IBM missed online cloud storage against the Microsoft, Google and Amazon?
I don’t think  we’ve missed it. I think we have our own set of offerings must notably, SoftLayer, and the integration of SoftLayer, which was an acquisition as you know we made some years ago…

Yes but when you compare…
So again, I don’t think we’ve missed that at all. In fact, we have some very successful implementations where we’re integrating SoftLayer in the cloud, as well as on premise offerings for our customers. We have a SoftLayer part here in Paris. And it’s proving to be very successful. If there is one differentiation in characteristic, I would tell you that I think that for us SoftLayer started at the enterprise, as opposed to necessarily maybe some of the lower level business consumer environment, so it was definitely a high-end enterprise offering that we are extending.

Do you have a special frame for archiving data at low price?
Yes.

And what kind of hardware do you use for SoftLayer? It’s a mix of tapes, disks, optical?
Actually all three. We’re in the process of making all three. We have recently had some very significant advancements in tape, tape density and archive capacity.

How are you going to react to the success of VCE and Flexpod in integrated infrastructure and platforms, which is a good business?
The marketplace for converged and hyper-converged systems is still being defined and there is an opportunity for both. From a storage standpoint, IBM obviously has the ability to participate in both domains, and we have a very strong relationship with Cisco, where we have an offering now VersaStack, which is available in the converged environment. In a hyper-converged environment, again from a storage standpoint, we have the ability to provide flash technologies as an integrated component. This is my personal view, I still think this market is playing out.

Yes, but these new software-defined storage probably will decrease the global revenue of the storage industry.
I think the revenue of traditional storage will decline.

It has already declined.
I think it will. But, I think it is being replaced.

Yes but will it compensate?
I think it’s being replaced in no small part. But, with software-defined storage, the intelligence and the logic is really starting to be built out at a software layer. Now, quite frankly, cost of spinning disk and the infrastructure associated with spinning disk, I mean, it has long commoditized. The value to clients, not to vendors, come through the delivery of the functionality within software-defined storage. So, I think the market is changing, no question.

For you, which IBM’s storage product is without competition?
Well, the answer for me is easy. All of my products.

That’s just marketing.
Here are two areas I truly differentiate. Our flash technology and the technology that we are building out, with real-time compression and capabilities around flash. And the second area where we are unbeatable is software-defined storage.

What’s the acquisition strategy for IBM in storage?
I’m not going to comment.

So there is no strategy?
Well, look, as an organization, IBM is always looking at the marketplace. We have made strategic acquisitions in the past and they’ve been very good business decisions for us.

IBM stated it will invest $1 billion in software-defined storage over next 5 years. Does it include acquisitions?
No.

Is IBM is looking at another technology?
No.

Roadmap in storage?
Three areas. Continue to build out our, so let me step back for a second. We think that flash-enabled technology will become the de facto standard in storage. So, across our entire portfolio, DS8000, all the way through to the low end, we are flash enabling those products. You can buy with one level or another of flash variant, so we will continue to build out flash capabilities across our products. Point number two, in the roadmap, is software-defined storage across the entire portfolio, ultimately to a single pane of glass, single management infrastructure and complete elasticity across the model. And point number three, not explicitly on roadmap statement, but an important one for us, is to dramatically increase the size of our partner ecosystems. So, partnership with the likes of Cisco, very important, very strategic to us.

It’s strange to remark that IBM was the inventor of the first HDD drive and now defend flash chips.
Again this is a personal reflection, if you think about a 15,000rpm drives, 7,000rpm drives, the reality is that at the end of the day it is a commoditized product. The density and the capability and the speed, in and around their product, has a fine out window of continued expansion. Likewise, even in the flash arena, semiconductor and the degree of density and storage continue to build out around semiconductors, it will continue to grow. It’s an exponential challenge.

Flash have all the advantage against disk but one, today, the price.
So, I have a saying when it comes to this discussion. Today, you are right, flash is twice the price, but half the cost. The pure semiconductor arrays twice the price or more, but if you look into the financial model relative to flash, space, power, application, all of those factors you can very…

There is just a problem: you can lose some chips the time being…
So, and by the way with our flash arrays now, we guarantee a compression level, we guarantee 7 years. Maintenance cost free, we guarantee that any chip or array will be replaced. I mean, it’s still a destructive rock. And by the way I think the semiconductor price will come down.

Do I miss any questions?
No, I think you’ve been very comprehensive.