00:00Gary great to have you here on the program 40 percent revenue growth in the most recent quarter. The guidance
00:05a little bit more measured if you will.
00:07And I'm wondering if you can start there. When you listen to that conference call you talk about an addressable
00:12market potentially doubling by 2029 to 50 billion.
00:15You're sitting on a backlog of seven point seven billion dollars. Why didn't you raise your forecast maybe a little
00:21bit more aggressively.
00:23Well we raised our forecast for the full year which we've done every quarter so far. We've grown the business
00:3140 percent so far this year year on year.
00:35And we've quadrupled basically our earning power from this time last year. So I think you know from our point
00:42of view it's full steam ahead and we feel very strong that this is a very durable demand.
00:48I think networking generally has been under invested in from an AI ecosystem perspective and we're seeing enormous amounts of
00:58demand from the hyperscalers and the service providers basically to facilitate the infrastructure of AI.
01:05Nothing happens on AI until it comes out of the data center. Right. And it needs the network.
01:10Absolutely. And to a large degree it needs Sienna as you still remain the dominant player there.
01:15But when we start to talk about the shifts in the way that these data centers are being built and
01:20making sure that these latency issues are addressed.
01:23You've seen NVIDIA start to develop its own switching protocols that are more directly tied to its chips.
01:29You've heard of Microsoft and other companies some of the hyperscalers also trying to do the same.
01:34Are these companies still going to need Sienna by the time that TAM opens up by 2029?
01:39Absolutely is the answer to that. What you're talking about really is is optimizing distributed compute.
01:45What we're all about is actually the physical connectivity. So you think about these dynamics around compute.
01:52I mean particularly when you've got constraints on compute that actually requires more and more networking to actually connect these
02:00data centers.
02:01So we're you know super confident that over the next few years more and more optical connectivity will be required
02:09to facilitate the the AI infrastructure and that you know optimizing that compute really just requires more networking.
02:17Gary I I get to hang out with Romain a lot and he's always talking about optimizing distributed compute.
02:24How is what we're looking at now different from what we saw in the tech bubble because back then investors
02:33were paying so much to lay miles and miles of fiber that didn't turn profitable for years after that.
02:40And all of the big stocks yours Cisco and the others took a huge hit you know went pretty much
02:47to zero for many for many years again investors were burned for so long.
02:51What's the big difference. Well I unfortunately do recall that and I have the scars from the from that period.
02:58So I empathize with that and I'm suitably paranoid around it as you'd expect. Yes.
03:03This is a very different set of dynamics. I describe that as sort of the first coming of the Internet.
03:08You know we built it and no one came. And then that really enabled the sort of models that we're
03:13very familiar with now the Amazons and the Ubers etc.
03:16This is a very different dynamic. As soon as we build it you know we can't build it fast enough
03:22for traffic to come on the network.
03:25And you think about all of the ecosystem that's gone into the AI world you know GPUs power space etc.
03:34language models. Now it's about the networking to really not strand all of that massive amount of investment.
03:42You need to connect these data centers with high speed low latency technology.
03:48And that is what Sienna does best in the world.
03:50But one of one of the pitfalls that we saw back during the late 90s and early 2000s was this
03:55idea of the build out of this capacity.
03:57At a time when a lot of the end clients didn't really have a business model in place to effectively
04:02use that capacity or at least I should say more importantly to monetize that capacity.
04:06You don't see parallels to that today. You think your end clients have a business model ready to go that
04:12is actually going to sustain their demand for your services.
04:15The answer to that is yes. And that is largely right now driven by the hyperscalers. And they have a
04:21business model that is scaling up on AI is profitable.
04:25And for every dollar they spend they get more than a dollar back for it. So I think it's it's
04:29absolutely got substantial business models that are that are proving out.
04:34And that's before you really get into the enterprise adoption which I think is largely all in front of us.
04:40So I mean where is that going to play out the biggest. That's what Jensen's been talking about as well.
04:45I think you know he talks about 60 trillion dollars worth of enterprise companies that are going to be using
04:51this.
04:52Who benefits first. Who benefits the most. Well I think right now it's the hyperscalers making making the big investment.
05:01I mean beyond that you know what industries we've heard so much about how health care is going to benefit.
05:05Finance should benefit. But we don't see that in the returns. Where where where are enterprise customers going to actually
05:12be profitable going to be monetizing AI.
05:15Well I think we're still in the early stages of that adoption frankly and it's taken longer than than anybody
05:21would would typically imagine because you have to change workflows.
05:25And that you know take takes a little bit of time. But the progress that you are seeing across the
05:30enterprises finance health care industrials etc.
05:34You are beginning to see that now. And companies like Sienna is adopting AI. We're seeing tremendous productivity gains in
05:41our services business and in our ability to get product out quickly.
05:45So that will begin to translate into economic benefits for the general enterprises as well as the hyperscalers.
05:53Gary do you have any concerns about some of the concentration risks with regards to your revenue sources right now.
05:58You said in the most recent quarter basically about 34 percent of your revenue came from two companies.
06:03I know you don't name them but anyone with the Bloomberg terminal can figure out pretty quickly who those two
06:08companies are.
06:08If they pull back in any sort of meaningful way from some of those CapEx commitments what does that do
06:14to your forecast.
06:16Well I think from our point of view it's not necessarily about the total CapEx that's spent.
06:21It's what are they spending it on. And increasingly to my point the the constraint is around the network.
06:28So to enable all of the optimization of their investment they really need the connectivity.
06:35And so you know from our point of view it's not about what you know how much absolute CapEx.
06:40It's what they're spending it on and percentage of that for networking.
06:43And that continues to increase.
06:45And our exposure is across all of the major hyperscalers, neoscalers and the large service providers around the world.
06:54Just to wrap this up though Gary and I'm just wondering if you could kind of just address less the
06:59issue of revenue growth and the growth of the company overall.
07:02And more importantly the idea of maybe potentially expanding margins.
07:05That expansion has been relatively flat.
07:07And I understand there are some cost inputs and price inputs that are sort of factoring into that that are
07:12beyond your control.
07:13But does this become more of a profitability and margin expansion story longer term if at all.
07:21Absolutely. And you're seeing a four times increase in our profitability year on year so far.
07:28And we think there's an enormous amount of operating leverage that we have in this business.
07:33We've demonstrated it over the last 18 months.
07:36We have an enormous backlog that we're confident will ship out in 2027 and beyond.
07:42And we're dropping that to the bottom line.
07:45You know we've had an incredible improvement so far.
07:48And our unique operating model.
07:50We're the only focused optical systems player in the world.
07:54And we're able to drop that to the bottom line because of our leading technology, customer relationships,
07:59and the unique model that we bring to the market.
08:02And we're going to drop that to the bottom line.
Comments