Episode Transcript
Transcripts are displayed as originally observed. Some content, including advertisements may have changed.
Use Ctrl + F to search
0:00
Unknown: Metcalf, the founder of Ethernet once once said at a
0:03
talk he said, I don't know what comes after Ethernet, but I'm
0:06
pretty sure it's going to be called Ethernet. Antone Gonsalves: Hi, and welcome to Tech News this week,
0:12
I'm your host, Antonio and solvus, editor at large for
0:16
techtarget editorial, Cisco recently released earnings that
0:20
spooked investors. What convinced them to push the sell
0:23
button on Cisco stock, which fell 10% in a single day, was
0:28
the company's lower than expected revenue and earnings
0:31
for the rest of the fiscal year. Here to help us understand
0:35
what's happening with Cisco is Bob Albert, Networking Industry
0:40
analysts at Tech targets at an Enterprise Strategy Group.
0:44
Unknown: Welcome, Bob. Thanks, Anton. Great to be here.
0:47
Antone Gonsalves: Yes. Nice to have you. All right. So Cisco
0:50
beat analysts estimates last quarter, but it's forecast for
0:54
this quarter and the rest of the year convinced analysts to lower
0:57
their price targets. What happened? Or what did? Why did
1:02
everything changed so rapidly? Yeah,
1:05
Unknown: I mean, I guess the The interesting part is, I don't
1:08
know that it's so rapidly, I think this is the, you know, the
1:11
event that we've been seeing, building up to this for quite a
1:15
while now, with all the supply chain shortages, and
1:20
organizations wanting to overcome that. And pre ordering.
1:23
And, you know, across both the distributor level, and the end
1:27
user customers, I think what we're just seeing is a bit of a
1:30
natural correction. And it's not just a Cisco phenomenon. If
1:34
you've seen if you've been following the earnings announcements all fall, you've seen a lot of the networking
1:38
companies do the same thing, like, hey, we beat this quarter.
1:41
But we're just letting you know, down the road, things are going
1:44
to slow down as both the customers and the distributors
1:50
get to consume the product that they already have. So it's just
1:54
I honestly, I look at it as a minor correction in the market.
1:58
And I think, you know, it's not anything impactful of any of the
2:03
vendors products or anything like that. It's just more of a
2:06
there's a supply and demand issue. And right now, there's
2:08
plenty of supply, which is going to weaken the demand for new
2:12
orders over the next couple of quarters. And that'll correct
2:15
itself. And then we'll be back back online, as
2:18
Antone Gonsalves: we all know, Cisco is is a bellwether in the
2:20
tech industry. You know, CEO, Chuck Robbins, you know, he
2:26
insisted that the earnings that disappointed analysts, it didn't
2:32
reflect a slowdown in enterprise IT spending in general. Right.
2:39
Now, investors, he said it was strictly what what you're
2:43
describing, you know, customers dealing with the backlog? Yeah,
2:50
the spending? Okay, well, but obviously investors didn't
2:54
believe them. They had doubts. This is so so why why the
3:00
disconnect? Is are is is is it? Is it enterprise IT spending? Is
3:07
it going to slow over the next quarters? Or isn't it? Well,
3:11
Unknown: I think like you said, for all those products that
3:14
people have been that were suffering from the backlog,
3:19
right, and pre ordering and ordering as much as they could
3:21
to make sure that they could meet their demands have now find
3:24
themselves in a little bit of surplus. I think that coincides
3:28
from also from a, you know, Wi Fi perspective, there were a lot
3:32
of organizations over the last couple of, or at least this year
3:35
anyways, who spent a lot on reimagining their campus
3:38
environment, and preparing for the return to work. And I think
3:42
we're seeing that a lot of those that maybe is a mature
3:45
environment as well. But like I said, from an overall
3:48
standpoint, we don't see anything slowing down to that
3:53
rate from a distributed environment, right. So at the
3:57
Enterprise Strategy Group, we refer to the distributed cloud.
4:00
So applications moving to multiple public clouds moving to
4:03
the edge being on private data centers, employees still very
4:07
much engaged in hybrid work activities. And then right, and
4:12
so they're working from home, they're working remote, they're
4:14
working in the office. So the one thing that's tying all that
4:18
together is everything needs to be connected. So is there going
4:21
to be a slowdown for the network? I don't think so. I
4:25
think as organizations continue to adopt hybrid and even return
4:29
to office, it's really going to continue to drive perhaps some
4:33
of that upgrade. Because what's happening is the legacy gear
4:36
that they have can't handle the collaboration bandwidth, right?
4:40
That video and voice traffic go into every desktop, right? Think
4:44
about pre pandemic, you had a video room. It was time for a
4:48
meeting and everyone was in the office and then you went to the
4:51
video room to talk to either another company or someone who
4:53
was remote. The reality is today's every desktop is a video
4:56
room. Right? And so and the way Everyone connects that desktop
5:01
is wireless now, right? No one runs wired to the desktop in any
5:04
new building, right prior to the pandemic, everyone was talking
5:07
about reimagining the office space and completely wireless.
5:10
And so to enable that organizations are going to have
5:13
to move to Wi Fi six, ie Wi Fi seven, right to get that
5:16
additional capacity to support that, that video,
5:21
Antone Gonsalves: and they are going to need that kind of an upgrade, given the fact that the return to Office is not full
5:26
staff. I mean, we're looking at, Unknown: it's not but like I said that because it's not full
5:31
staff is exactly why they have to do it. Because that means
5:34
what any, anytime they're going to collaborate with somebody,
5:37
half of the people might not be in the office. And so again,
5:41
it's not. So if everyone's having this in the office, you
5:43
don't have enough video rooms to accomplish that. So from what
5:46
I've heard, like said that the talk track is typically around,
5:50
how do we reimagine the campus environment to enable those
5:55
collaboration apps to be used from anywhere in the building
5:58
and not just specific video rooms, and then how this plays
6:02
out for the networking companies, that means new wireless access points. And because these new wireless
6:07
access points accommodate the 2.5, the five and the six
6:11
gigahertz range, they require more power, which means they
6:14
have to replace the wired infrastructure that's supporting
6:18
it, right. So they can get P OE plus and things like that. So
6:20
they have sufficient power to power the access points. And
6:23
they're multiple antennas, etc. So, so that's where that's where
6:27
you start seeing that change. Now, the question is where we
6:30
are not transition. Right, and whether that's going to continue
6:33
to drive more, or people have already bought for that. But
6:38
there's like said there's there's a few other things that are that are driving it, as well. I don't know if you want
6:43
to go there now or if you want to stay ahead
6:46
Antone Gonsalves: what because yeah, what's so my strategy
6:49
group's research showing Unknown: us? Well, it's there. It's not only the research, but
6:54
also the I've been on a lot of calls with the networking
6:57
companies Lately, they've been having analyst days and things
6:59
like that talking about their plans going forward. And there's
7:02
two little two little letters that keep cropping up. And all
7:05
those that happens to be a an AI. That happened. And so with,
7:11
with all of the, with all of the interest around AI and all of
7:16
the activity that's going on, there's there's something that
7:20
we refer to it's like networking for AI. So how do you enable
7:23
those back end infrastructures for these AI environments that
7:28
these hyper scalars are putting together these large enterprises
7:31
are putting together? And if you look at it, you know, based on
7:35
who you talk to, it could be anywhere in that range of an
7:39
additional six to $10 billion dollar Tam, for these network
7:42
providers. So here's something new that's going to drive
7:46
growth, Antone Gonsalves: what tam stands for, what a
7:48
Unknown: target addressable market. Right. So that's, I
7:52
mean, they're going to make six or 10 million, that's just that's what it is to be split up amongst all the all the vendors.
7:58
And you know, a lot of them will, you know, in that lower
8:00
end, one tends to be more toward just the back end networks, or
8:03
the high performance back end networks. These are the ones
8:05
that will compete against the InfiniBand stack, did invidious
8:09
put together for that space. And this is where things like I know
8:13
if you're familiar with the ultra Ethernet consortium, I am.
8:17
So something recently, Antone Gonsalves: Cisco and Arista are members of that
8:21
consortium. And their purpose is to make Ethernet instead of
8:27
InfiniBand. The number one net Correct? Yeah,
8:32
Unknown: HPE. And there's a few other companies on it, you can
8:34
go and see the whole list of them. And yeah, the idea is,
8:37
look, do you really want to introduce another technology.
8:40
And we've all known if you follow networking long enough,
8:43
there's been a lot of technologies that have been introduced, but Ethernet is still going. And so remind Bob
8:49
Bob Metcalf, the founder of Ethernet. Once once said at a
8:53
talk, he said, I don't know what comes after Ethernet, but I'm
8:56
pretty sure it's going to be called Ethernet. And so and
8:59
that's what you're starting to see with this ultra Ethernet,
9:01
right is that hey, how do we, how do we have a tech? Why
9:04
don't? Why would we try and put something completely new in
9:07
where people don't have the resource, the skill sets, etc.
9:10
When we could just modify and leverage Ethernet to do the same
9:14
thing, right? Let's keep the cost down, leverage the same
9:17
boxes, the same methodologies, there's a little bit of
9:20
difference, right? Because the performance is required. And so
9:22
there's some new technologies and that's what the that's what
9:25
the the UEC, the ultra Ethernet consortium is all about how do
9:29
we modify Ethernet? How do we modify the rocky stuff to make
9:33
it simpler to make it non blocking all this kind of good
9:36
stuff to enable organizations to use that? And so, you know, like
9:40
I said, with Arista, with Cisco and HPE, and you can imagine
9:44
others will join as well. There's a lot of other companies in there, that they're really going to be focused on. How do
9:49
we how do we incorporate this and build it out? So that's a
9:51
big opportunity for a lot of the network companies to expand
9:55
their Tam and get back to a growth area right through Do
10:00
something that doesn't exist didn't exist six months ago 12?
10:03
Well, it did, but not in a way it is now with everything
10:06
blowing up and all the attention being spent on it.
10:09
Antone Gonsalves: Okay, so who's going to who's going to be the
10:11
buyers of this ultra Ethernet? When and if it comes, I mean, we
10:17
have I mean, when it comes to AI, enterprises, these based on,
10:24
on my reporting that enterprises are not this will try to figure
10:30
out the return on investment on AI, they're still trying to
10:32
figure out what's required. And because there's so much more
10:36
than even beyond the technology, what's required technology wise,
10:40
but to get the education of employees, how do you use it?
10:43
How do you secure it, I mean, these are issues that they're
10:46
struggling with. And as a result, if they're under
10:48
pressure to do something right away, if a CIO is under pressure
10:52
from the board, or the CEO to do something right away, they're
10:56
just turning to Microsoft to Salesforce, and use AI. All
11:02
right, so you're talking about something about Ultra Ethernet?
11:05
I mean, is that really, for networking companies to sell
11:08
more to service providers and cloud, but then cloud providers
11:13
hyper scalar, scalars, as opposed to the universe? So
11:17
Unknown: I think this is going to follow a path, right?
11:19
Initially, it will be those hyper scalars, right, the
11:22
service providers, the, you know, really large enterprise,
11:27
who will be wanting to build their own training and
11:30
inference, malt, right environments. That's who's going
11:33
to adopt it. But like most technologies we see, and even
11:36
hyper scalar technologies make their way down to the
11:39
enterprise. So the initial market is yes, it's about
11:43
enabling these hyper scalars that are using 10s, of 1000s of
11:47
deep of GPUs, and like just massive environments, huge power
11:50
consumption, that's who will be the initial customers, you'll
11:54
have some of the really large enterprises doing that as well.
11:58
And then you'll see over time, right, that you'll see other
12:02
enterprises wanting to have their own environments wanting
12:04
to build things out. But you're right, if you're, if you're an
12:07
SMB, or a smaller enterprise, and you can leverage, you know,
12:11
Ignite was this week. So it's top of mind, right? If you can
12:13
do co pilot for 365 and Microsoft graft and use their
12:17
studio to integrate with your, your SharePoint, for data,
12:21
right? Because it's all about the data for AI, right? How do I
12:23
get the data? And most importantly, for all the
12:26
enterprises like you're saying, from a security perspective,
12:29
it's about how do we make sure that none of our data gets put
12:32
into a public API environment? So by building out those
12:36
private, we commonly refer to them as SLM is right, the set of
12:40
the large language models and the small language models, just
12:42
based on their own data, it's that they own and control. So
12:47
you're right, it's going to be a process. But definitely the
12:50
larger organizations, the hyper scalars are absolutely gonna
12:53
lead the charge on this. They're obviously they're already there.
12:55
Right? They're already buying this. And so the question is,
12:58
how much of it goes to? How much of it goes to InfiniBand? And
13:02
how much of it goes to Ethernet initially, and then once like
13:05
UEC, is out? How much starts shifting over to to Ethernet? I
13:09
mean, the, the hyperscalers have certainly shown that they don't
13:14
want to use proprietary technology, right? Everything's
13:17
white box, or some slightly modified white box type of
13:21
technology, right? They want to be able to reduce cost, they
13:23
want everything to be standardized. So there's a lot
13:26
of, there's a lot of pluses for why they would want to use
13:30
Ethernet. Over InfiniBand.
13:33
Antone Gonsalves: Okay, so let's circle back to Cisco here. Were
13:36
one of the because Cisco, Cisco, if Cisco is going to get back on
13:43
track in terms of revenue to satisfy their investors, it
13:48
seems like Splunk is, is a critical piece of that they're
13:53
spending $28 billion for this company. How will Splunk change
13:59
their business? And, you know, give them their next round of
14:04
revenue? Unknown: Yeah, I think that's another it's a really
14:08
interesting one. Obviously, this has been talked about for a couple of years, right? Whether it's people have been waiting
14:12
for this to happen. So it obviously it presents a huge
14:17
opportunity for them from the observability space. As I
14:20
mentioned earlier, these distributed cloud environments,
14:24
with applications everywhere, with workers everywhere, it
14:27
becomes increasingly difficult, right? The complexity increases,
14:31
the ability to have the ability to have visibility, essentially,
14:36
right across these highly distributed environment becomes
14:39
really challenging. So they've been doing things on the network
14:42
side, like with 1000 eyes to look at what's going on in the
14:46
across the internet, right? Being able to tie that in with
14:49
their devices, being able to get all that telemetry. While Splunk
14:53
has been working on how do I collect all the log data from
14:56
everywhere? How do I collect application data? So there's
14:58
obviously there's a little bit of old up there, right? Because
15:00
Cisco has AP D and things like that. But the overall strategy
15:04
here is how do we take all the tools that we have to collect
15:08
all the data, we have to provide that end to end visibility
15:11
across that environment. So that we're able to very quickly and
15:16
easily detect when there's an issue or to be able to collect
15:19
enough information and correlate it so that we can then you know,
15:23
proactively optimize the traffic. And so what Splunk is
15:27
doing is really, you know, the the addition of Splunk is
15:31
enabling them to have that, basically that full stack of
15:36
observability solution that collects from the was it melt,
15:42
right, the metrics, events, the logs, the traces from the, from
15:45
the applications and from the network. And I don't know that
15:47
anyone's got a better network visibility story than Cisco
15:52
does. Between all the you know, all the acquisitions, they've
15:55
made, all the AI focused acquisitions, they've made all
15:58
the observability acquisitions they've made to be able to feed
16:02
all that data up, and then be able to, you know, crunch it and
16:05
provide ultimately what everyone wants, right? It's not data for
16:08
the sake of data. It's data for the sake of getting actionable
16:12
intelligence, right? Being able to do something with that data
16:16
that drives a better process that drives better quality that
16:20
drives a better experience is ultimately what the goal is
16:23
right. Antone Gonsalves: Bob, thanks a lot for joining me. It's always
16:27
a pleasure to hear your insight and Cisco and everything else
16:31
networking. And for our listeners, that wraps up today's
16:36
podcast,
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More