Breaking Analysis: Spotlight on IBM’s Systems Business


>>From the SiliconANGLE media office in Boston, Massachusetts, it’s theCUBE. Now here’s your host, Dave Vellante.>>Hi, everybody, welcome to this edition of theCUBE Insights powered by ETR. In this breaking analysis,
we’re going to look at IBM’s Systems business
and, specifically, the IBM system Z, and
talk about the impact that it’s going to have on IBM financials. Now, Alex, if you would kindly
bring up the first slide. So this is data from ETR’s spending intentions survey
for the second half of 2019. They asked customers, compared
to the first half of 2019, what are your spending intentions on the second half of
2019, specifically for IBM? So you can see the N here is 448 customers out of their panel of 4500,
of which around 11 or 1200 answered this question specifically cited that they were IBM customers. What this data shows
is 21% of the customers said we’re going to increase
spend in the second half relative to the first half with IBM. 52% said we’re going to stay flat. 14% said they’re going to decrease. You see 6% said we’re
going to basically leave the IBM platform and 7% said we’re going to bring on IBM
as new, we’re a new customer. So, if you take the people
that are spending more and new and subtract out the leaving
and the spending less, you get a net score, and
you get a net score of 8%. Now we’ve been sharing
with you this ETR data over the last several weeks and months. 8% is not great. IBM, according to ETR spending surveys, they’re losing share relative
to the overall market. We’ve covered this pretty extensively. We covered the Red Hat acquisition and talked about how that IBM intends to supercharge its cloud business. Specifically with Red Hat, I’ve said I’ve been on record,
saying this is largely a services play where they’re
going to basically take Red Hat as an application
development platform and help their customers
modernize their systems from using their large
services footprint to do that. And I want to talk for a moment about the IBM business overall. IBM is all about mission-critical work. The IBM Z, their high-end
systems, their related database, it’s all about mission-critical work. IBM shared some data
with analysts recently where they talked about,
if you look at IBM Z, IBM Security business, its database, particularly Db2, its middleware, its application management services, and its infrastructure, and
all that set of consulting work that goes around that. Add that up, it accounts
for 60% of IBM’s revenue. So this is why I want to spend
some time talking about IBM Z. I mean it’s kind of a
boring but important topic. It used to be the heart of IBM’s business. It used to drive entirely
their income statement, but in fact, today, it’s
still very critical. All those pieces that I mentioned account for 60% of that business. So Z is critical for driving
IBM’s Systems business and that gives air cover
for IBM’s business overall. So, Alex, if you bring up the next slide. What I’ve done is just pulled
out some quarterly data of IBM’s Systems revenue overall, and then juxtapose it
against IBM’s Z revenue. This is growth, this
is just percent growth. So the blue is IBM Z
percent revenue growth relative to the previous year. This is in constant currency, by the way. And as well, it excludes the elimination of IBM’s Systems X business,
the Intel-based business, so it’s normalized for that. And then the orange is
the overall Systems growth so you can see that the
blue grows virtually immediately after IBM
announces a new system. So, for instance, in January of 2013, IBM announced the Z13. We were there with theCUBE to cover it. We talked to a number of practitioners. What big banks and big
mainframe customers do and, by the way, 25 of the
world’s top 25 banks run on Z. Huge proportion of retail giants run on Z. Why, because it is the system of record, and the top system of record, along with Oracle, in the world. I’ll talk more about
that, but you can see here Z13, so we talked to
a lot of practitioners at the January launch, and they told me they’d buy this thing sight
unseen because they know it’s going to drive revenue for them. If they can get more power,
more performance, lower cost, it drops right to their bottom line. So you can see 2016, even though there was a kicker in there, of the next generation,
not next generation, but a kicker to the Z,
I didn’t show it here, but bad year in 2016 in terms of growth. And you can see the blue is
proportional to the orange. It drags it down in here. Z14 is announced and you can see, when the Z14 was
announced, in July of 2017, just right after that, boom. Big uptick in Z revenue and
proportional Systems revenue. So you’re on this two-year
cycle of Z announcements and you can see 2019, in the
first half, has not been great. IBM just announced the Z15 in September, so we fully expect that in Q4,
you’re going to see that uptick. So kind of wanted to share that with you. Next slide, I just want
to make a couple of points about IBM’s Systems business. It’s about an $8 billion business overall in terms of annual revenue. It comprises Z, power, and storage. It’s as they say, System
Z drags a lot of software. It drags storage, it drags services. It’s about a 53% gross margin business. The storage business is actually quite a good gross margin business, I think, probably higher than power. The server business is not
the greatest gross margin. I think mainframe is still pretty good. IBM and Oracle dominate the
business for systems of record: Oracle with Exadata and IBM with Z. Now you might say, hey, Exadata is growing and Z is, I just showed you,
the sort of fluctuation, but overall, it’s sort of flattish. Maybe it can eke out growth. It actually can show
good growth in one year, but if you normalize
it over a couple years, it’s pretty much a flat
business or declining business. So you might say, well,
Oracle X has grown, but that’s ’cause Oracle’s replacing its entire hardware business and much of its related
software business with Exadata, all that wood behind one arrow, whereas IBM has a more
diversified portfolio, and so that’s kind of
apples-to-oranges comparisons. Now the ETR data shows
that the storage intentions for the second half of 2019 really flipped to positive territory. Servers were still negative but improving. And so, as I showed you
in the previous slide, I definitely would expect
the Systems business to have an uptick in Q4 and it’s dragging storage with it. IBM synchronized the storage
announcement, the DS8000… I am not great with model numbers but the recent storage announcement with the mainframe announcement, I’ll make some more comments about that. But it seems that IBM’s
trying to do a better job of synchronizing that. IBM’s also going to smooth out
its Systems revenue, I believe. I mean, right now, it’s very cyclical, but I’ll make some comments
about that in a moment. So IBM system Z and Exadata are unique in that their IO is tightly integrated. These are purpose-built systems and the storage and the
IO are purpose-built for the systems of record, so they’re very, very low latency. Give you an example, Oracle Exadata, in recent announcements
at Oracle OpenWorld, I think 18 microseconds latency. IBM, with its recent Z announcement,
I think, is even lower. I want to say 15 microseconds
but don’t hold me to that. Whereas, if you compare
that to traditional systems, you’re talking about
maybe 200 microseconds, in other words, those systems
that aren’t purpose-built for systems of record with integrated IO. The IO is hardwired with
custom silicon and A6 and so it’s ultra, ultra-fast IO which means you can push 10
times the IO through the system. So very, very high performance relative to what you saw
in previous generations. Why do I spend so much
time talking about this? Because this is a harbinger for
future Systems developments. Talking, within two to three years, you’re going to see the mainstream systems with this type of low latency. So now, you might also say, wow, that means that the IBM
and the Exadata business are in big trouble. No, these systems are
not going to be replaced, they’re not going to be migrated. It’s too risky, it’s too expensive. We’ve talked about that a lot on theCUBE where it just doesn’t make business sense for people to convert off the Z mainframe. There’s too much custom code. They’d have to freeze that code for many, many months, maybe even longer. They’d risk their business. They’re going to make much more money purchasing the next generation of system as long as the Z mainframe continues to add function, which it’s doing. Same thing with Oracle Exadata. Years ago, IBM announced
support for Linux. Obviously, Red Hat is now
another key piece of that. In the recent Z15 announcement,
Encryption Everywhere. They announced a hybrid cloud so basically bringing the Z to cloud. Really strong security focus. This cloud piece is interesting. We talk a lot about Cloud 2.0. Bringing the Z and the
systems of record to cloud is something that IBM has said that it intends specifically to do. That will begin to potentially smooth out IBM’s Z revenue. You know, it’s ironic. In the latter part of the 1980’s, kind of a financial game
that IBM was playing, they converted their rental base, which was a monthly income
stream, to purchase. When they did that, it created the effect of showing up on the income statement and kind of hiding the trouble
that IBM was really in. When that transition
ended, IBM really tanked, and that’s when IBM got into big trouble, the whole downsizing trend. Gerson had came in, they bought PWC, and really transformed the company. But the Z as the system of record, or the old 3090, has lived on. Now we’re seeing that
dynamic come full circle where, over time, IBM can shift from an upfront pay to a subscription, which is, as they say, coming full circle. It’s going to be interesting to
see how that transition works. The other point again. The storage seems to be
synchronizing its product cycles with Z at least at the high end. And so this is likely
to carry through the Q4. We see, from the ETR spending data, that storage intentions are up. I think the net score was up 5% versus a negative from the previous quarter. Servers overall were still down. They don’t have a question specific to Z, but I would fully expect
that Q4 this year, you’re going to see a
nice uptick in Z revenue. And, as I pointed out
before, with that 60% number, this is going to provide
another halo effect for IBM’s overall business. Will it be enough to propel the stock? Probably not, this stuff is factored in, the analysts understand
these product cycles, but it’s something that I
wanted to shine a light on because, again, it’s one
of these important topics that not a lot of people talk about. People roll your eyes when
you talk about the mainframe, but the mainframe is
here, it’s alive and well, and what I call “mainframe”,
Oracle Exadata and IBM Z, are really the two companies that are prominent in that space. And, while they might
compete, to my earlier point, you’re really talking about each company having its own install base and, as long as they
keep investing in R & D and keep those product cycles coming, I would expect that this business is going to be healthy yet
cyclical for a long, long time. This is Dave Vellante for
theCUBE Insights powered by ETR. We’ll see you next time,
thanks for watching. (upbeat electronic music)

Ray Krug, NETSCOUT | Unified Communications


>>From the SiliconANGLE media office in Boston, Massachusetts, it’s theCUBE. Now, here’s your host Dave Vellante.>>Hi everybody, welcome
to this CUBE conversation. I’m Dave Vellante.\ We’re going to talk about
unified communications and its role in digital transformations. Ray Krug is here. He’s a solutions architect at NETSCOUT. Ray, good to see you,
thanks for coming on.>>Hi Dave, good to be here.>>So talk a little bit about NETSCOUT. You guys are into a lot
of different things, but give us the overview.>>Yeah, NETSCOUT, what
they’re primarily focused on is providing the visibility to assure digital business initiatives, to provide availability assurance, performance assurance, as well
as security assurance as well and we do this using our smart data and smart analytics platform. We kind of do this for, okay,
got a huge customer base, we do this for over
90% of the Fortune 500, 95% of the carrier service providers, so we scale to these large enterprises, sophisticated service providers, providing the visibility they
need to assure their services.>>So as a solution architect
what specifically is your role?>>Probably worthwhile
giving a bit of history because I know we’re talking
about unified communications. So I have been with NETSCOUT
now for about eight years it’s been and I came from an acquisition. The acquisition was
from a British company, a spin-out of British
telecom called Psytechnics and we specialized, this is
eight years, well 10 years ago, in analyzing the IP network
for voice and video traffic and actually being able to
understand how we can take impoundments in the network
and how that translates to impoundments in voice
quality over a voiceover IP. So that was the original
data transformation project, the so-called digital transformation from TDM networks to IP. So yeah, we took those analytics and basically figured out how to do that.>>So deep understanding of actually what’s going on in the network?>>Yeah absolutely, and what was exciting, and back to NETSCOUT, is when
they acquired Psytechnics, they took this technology and put that into their pro-technology, they did that within three or four months. Our technology was in their
probe monitoring the voice, both voiceover IP networks,
and then what was interesting, within 12 months, all our
workflows that we created for insured performance
of voiceover IP networks got embedded into the NETSCOUT
portfolio of products. And since then, eight
years, winding on forward, we’ve been embedding
more and more technology into our InGenuis One platform
to give you better and better voice, video, and unified
communication analytics.>>I love that story, Ray,
because the vast majority of mergers and acquisitions fail to meet their original objectives,
they take too long to integrate so some companies are good at
it, some not so good at it, so it must’ve been pleasing
to see that happen, and see your baby
actually scale like that. All right, lets talk about big picture. What are the big trends
that you see sort of driving unified communications today?>>Yeah, unified communications is getting more and more complex,
and perhaps on one accord, sophisticated, but you
kind of think, okay, most common used case for
us is to be a contact center because at the end of
the day, contact center, the customers are demanding
more and more ways to interact with the business,
traditionally it was voice but now they want web, web chats, video, whatever it might be, so
contact centers a big consumer of unified communications. And then there’s the
different technology trends like, of course, Microsoft
Skype for business, evolving into Microsoft
Teams, or Cisco Jabber, unified communications and
all that sort of thing. A whole bunch of other topics going on, again, part of digital
transformation initiatives, SIP trunking, we’re still
seeing that going on. So I was talking about TDM to IP, so that was back in my day in Psytechnics, now it’s taking those and
transferring IP to SIP trunking to save costs, that’s the main thing, but it is a change and it is
more, not instrumentation, but more appliances on a network, like session border controllers in order to add your SIP trunking, and of course there’s
also other technology, migration to the cloud
as well, which ends up, from our perspective what we’re seeing is in very hybrid environments. So now you’ve got a lot of on-prem stuff and some cloud stuff, it’s
all going to work together in order to make voice, video, unified communications successful.>>Isn’t another sort of challenge, I’ll call it give the
people what they want, you talk about contact centers
being a primary source, people want to communicate
in different ways. Young people maybe want to use chat, some people like me want to
pick up and talk to a human. Is that part of the challenge, is bringing all those together to service all these
different constituents?>>Yeah, absolutely, because
at the end of the day, it’s a contact center, you
want to make sure you provide an engaging experience to your customers, however that might be. Omnichannel or whatever
word you want to do it. The longer and happier the customer is dealing with your business,
perhaps the more money they’ll spend with your business, perhaps the better brand awareness they have of your business as well.>>So double click into
some of the challenges of actually bringing this stuff together, making it work, is it cost, you
mentioned complexity before, is it understanding the
analytics, who’s using what, predicting, double click on that.>>That’s a big topic, but
we talked about new features and immersive experience
from unified communications, so that’s all brilliant. The trouble is, high quality is key. You got to make sure that it’s successful, so any migration project,
you need to be successful to make sure that you’ve succeeded. Okay, so that’s number one. Quality is key, but also in terms of cost, sometimes these initiatives
about cost savings, so SIP trunking is a good example of that. I want to make my service
the same as it was before, have some sort of future
upgrade capability, but kind of make it cheaper, that’s what SIP trunking
does for you as well. So those are some of the
reasons for doing it, but then that introduces more components in your infrastructure to
make all that stuff work and it’s not just about voice and video, it’s all about the other
backend servers as well to make it all happen
whether that’s mail or chat or presence or whatever it might be. Lots of components now
that have to work together, stuff that you control but also
stuff that you don’t control like SIP trunks is a good example, or gateways out to the PSTs, things that you don’t control, and that makes it kind of
really tricky to deal with. There’s a bunch of other stuff as well that’s important, network convergence, you’ve got all these
applications converging onto that one network infrastructure, how do you manage that?>>Quick tangent. So you mentioned SIP trunking, explain what that is for our audience so they don’t have to google it. (Ray laughs)>>Yeah, so SIP trunking, basically, if you think about
gatewaying out to the PSTN in terms of making your
plain old telephone calls, dialing a number and sending out, SIP trunking does that all
from an IP perspective. So the idea is, you don’t
necessarily do a conversion to TDM, traditional phone
systems, it all goes IP. So basically, you then
send everything out, IP, over the network, it
gets to the other end, and the whole purpose of that,
it’s a service that you buy from your service provider and it’s cheap.>>Okay, you talked
about these challenges. Generally, how does the industry approach solving these problems and specifically how does NETSCOUT solve them?>>Great question. So traditionally, let’s
sort of rewind a little bit, I talked about a lot of components that need to work together to make your unified communications experience. Lots of servers, lots of
network infrastructure, firewalls, session boarder
controllers and all that. Traditionally, what you do is
monitor each of those devices. Take a look at their CPU utilization, or take a look at how the
servers are performing, and often, very little
is taken into account about the network and how that’s behaving, because again, I’ve said
it’s a converged network. So you end up with a picture saying, all my servers are working fine, but then you end up with the problem, but users are complaining
because they can’t dial, users are complaining
because the quality is bad. So that’s kind of the
problem with trying to bring all those together
using the different metrics and coming up with some
sort of conclusion.>>And then it’s finger pointing, right?>>Oh yeah, classic.>>Which mole to whack.>>Yeah, in constant use
cases, war rooms, okay, all my lights are green for
every person in that war room but the people are still
complaining, absolutely.>>Okay, so talk more about
how NETSCOUT approaches this.>>So, the name gives it away, really. We always focus on what’s
going on in the network, wherever that network may be, so we’re taking a look at
that, we call it Y data, it’s packet data, and we’re
able to translate that. Whatever’s going over the wires, whether it be an application
going over the wires or whether it be unified
communications going over the wire like voiceover IP, RTP, or signaling, SIP as an example of those. So we’re able to get that
picture of how everything is communicating with each other, and we’re being able to raise that level. So packets are notoriously
hard to interpret, but we’ve cracked it, we’ve
got a sort of technology, it’s a patented technology called ASI, adaptive service intelligence,
we call it smart data, but it’s converting that Y data into meaningful keeper points metrics. So you name it, you name the application, we’ve got performance metrics. So whether that be voice, voice quality, mean opinion score, we’re
taking that from the Y data. Whether it be application performance from a database that might be running, or a mail server that might be
running, we have performance. Whether it’s this signaling
that goes on to get data and all that, we have
performance metrics about that. So we’re using the same
data set, the Y data, bringing it up to our
analytics, our ASI layer, and then we have an understanding of what component’s failing. Is it the voice that’s failing? Is it this part of the
network that’s failing? And then, for voice, there’s a whole topic on how we understand that,
remembering my background and the analytics behind that.>>So, your secret sauce is
you’ve got this deep probe into the network, you’ve got this ASI, this patented technology, and
you’ve got an architecture to leverage that capability,
and that is really your big differentiator from
a technical perspective? Is that right?>>Well, from a technical
perspective, absolutely. And from an obvious perspective, we solve, in the easiest way, the
most complex problems. It’s kind of where it’s coming, ’cause these are tricky problems to do, they sometimes go unseen for ages, but because we’ve got
that overall visibility, we get to that root cause very quickly.>>Okay, let’s talk about
the business impact. Maybe you can give us some
examples, customer examples, and how it affected their business?>>Yeah, so that’s important. A couple of things, let’s
imagine you’re contact center, a service company, so
I’ve got one in mind, and the one that I have in
mind, six contact centers, they take up to about
100,000 calls in a day. So it’s important. They’re a service company
so people phone them up to have their service. If you can’t make contact
with your service company, maybe the impact of that is,
okay, that service is rubbish, I’m going to go to a
competitor, as an example. Or you don’t get your
service that you require. So there’s huge implications. In this example, we’ve found
that calls were dropping, as an example, so people are
connecting with their agent, calls are dropping, okay, hopeless. It’s really problematic. And it’s interesting that you
pointed out about war rooms and finger pointing, and
that’s exactly what happened. What they’d done, they’d engaged
in a SIP trunking project to deploy SIP trunking
they were going to save a million dollars a month by
implementing this SIP trunk. So that’s huge, okay yet,
when they deployed this, they were having a bad
experience, so that’s critical, so they needed to achieve
that successful migration, so they had tours but
nothing that could spot what was going on with
these calls dropping. So along come NETSCOUT,
we deployed our probe, and very quickly, it’s just amazing, very quickly we were to able to analyze the reason for the call dropping. Turned out it was a firewall issue, complex network so it’s kind of difficult to know where the traffic is routing. We were able to figure that out, give it the evidence to say the signaling, the SIP, was dropping, and
we were able to pinpoint that and they got that fixed very quickly.>>Which meant that they
were able to realize that million dollar a month savings.>>Precisely, yes exactly. Let alone that any business
that might’ve been affected by the fact that people couldn’t call in.>>Any other examples you can share?>>Yeah, I’ve got a really great one, probably closer to a
lot of people’s hearts, and relates to a hospital, and they were going through
a digital migration project. It’s as simple as changing
their phone handsets from one vendor to
another in some respect, about 2,000 phones that
they were replacing, so it’s kind of interesting. So I’ve now got a nice new
shiny phone on my desk, when I pick up the phone
I get very bad quality and stuff like that,
and just blame the phone and all that sort of thing. Sometimes that’s change,
people don’t like change, they like all the buttons
on their old phone, and sometimes it’s real, but in a way, the business impact for that
one is, if I’m a customer, a patient, I’m phoning up
my doctor for some records, and the phone quality is bad, then I’m not going to
have that much confidence that the doctor’s going to be able to cope with my ailment that I might have. So it’s really important to have quality, and when it’s about your health, then it’s really
important that it’s there.>>Awesome. Let’s end on some advice that
you would give to customers. So you got people trying to
do digital transformations, they’re trying to pull all these different communication systems together, trying to understand
where the exposures are, the performance issues. What advice would you give
to people that are struggling with these problems,
where should they start, and what should their journey look like?>>In some respects, I
think visibility is key, both before pre-migration, during migration and afterwards. So in my example before, having visibility of the
performance of the phones before, in this migration issue,
and then as I go through the migration, being able
to just check that when they deployed the new
phones, everything’s working. And then of course, once,
if there were any problems, so in my example, it was QOS problem. QOS, quality of service, so
that’s a networking problem and it goes back to, because
we’re in the network, we’re looking at the network, as much as that’s the most
complex problem to solve, and it’s everywhere, QOS
problems are everywhere, it’s the simplest thing for us to fix. So monitoring during migration, seeing what the behavior
of the phones are, during that process,
correcting everything quickly, so that the migration
project is successful, and then post-migration,
business as usual, monitoring, so if there are any problems
you can quickly react to it.>>Got it, okay, so you’re going
to through a business case, you’re going to make this part
of your digital transformation, you’re going to bring
together all the stakeholders but I think your point is,
if you don’t have visibility on what’s going on in the network, there are going to be some blind spots that you potentially run into. If you have visibility in the network, you’re going to be able
to remediate those, and the example you gave
of the services company, you’re going to be able to
achieve your expectations and your ROI results and have confidence that you’re going to be
around for the next project. So Ray, thanks very much for
coming on and sharing with us. And thank you for watching everybody, we’ll see you next time. This is Dave Vellante with theCUBE. (bright synth music)