Sebastian Hassinger (00:31)
Welcome to another episode of The New Quantum Era. I'm your host, Sebastian Hassinger. My conversations with people working on the frontiers of quantum information science and technology are motivated by my curiosity and my deep interest in watching this scientific field evolve into usable technologies for computation, communication, and sensing. From the very first introduction to quantum computing that I had at a workshop at IBM's TJ Watson Lab in 2017, I've been utterly hooked. But my fascination with emerging technologies didn't start there. In the early '90s, I founded an internet service provider while starting a master's program in communication studies at Simon Fraser University. By the mid '90s, I was leading a skunkworks project to build Apple's first customer support website, where we built our own search engine and editable web pages. And we're in the top 10 most visited websites in 1996. It was a lot of fun, and I made lifelong friends. Rest in peace to Mike Erwin. After the dot-com boom and bust, I shifted to mobile, and then open source, and big data, and machine learning, and so on and so forth. All along, I bounced between my own startup projects and internal innovation projects at large corporations like Apple, IBM, Oracle, and now Amazon. When you're bootstrapping a startup, it feels like you have to do everything with almost nothing. And you look at the resources of merging and new lines of business in large corporations with utter jealousy. However, all those resources inside a corporation don't come without a price. Without a doubt, some of the most challenging innovation goals to accomplish in my career have been those carried out inside corporations. Successful companies are proud of the things they've done that have led to their success and skeptical of any new thinking, for fear it might disrupt the secret of their success. Even when an internal innovation is needed to deliver new areas for growth or to stave off new competitive threats, it can be very difficult to get the freedom of action needed to put all those amazing resources to work. Quantum technologies are still in their very, very early stages, with one foot firmly in the experimental physics labs and theory groups around the world. However, for quantum technologies to eventually mature and evolve into truly useful tools that can create new value for humanity, they will need to tap into corporate level resources. Engineering techniques must mature, product feature sets need to be defined, operational models honed, and above all, the quantum industry needs to deliver. As exotic as quantum technologies are, they must be integrated into the tried and true tools of commercial success to realize their potential. Today's episode features an exciting conversation with a leader of a great example of corporate innovation. I spoke with Vijoy Pandey, GM and Senior Vice President of Outshift by Cisco. Outshift is the company's internal incubation engine, focusing on emerging technologies that target adjacent markets and personas. Their startup-like team's current initiatives span a number of AI and infrastructure topics, and most recently, have included emerging quantum technologies. Vijoy explained to me how the Outshift model works, and then we dove deep into the potential applications in quantum communications that are of particular interest to Cisco. I was really surprised actually by how near-term Vijoy's use cases are. I was aware of currently deploying technologies like quantum key distribution and post-quantum cryptography, but you'll hear him talk about relatively near-term practical use cases like network time synchronization, and even some applications in financial services like high-frequency trading. Even more exciting is Cisco's vision to build a scalable, interoperable platform for quantum technologies, which I believe will be crucial for progress in quantum across the board. By the end of my conversation, I felt like I understood Cisco's relevance to the quantum industry, and I was left with a real sense of excitement about what Vijoy and team will deliver. I hope you enjoy our conversation.

Sebastian Hassinger (05:15.07)
Vijoy, thank you very much for joining me today. I love to start with just giving you a chance to tell sort of your story of how you've gotten to, how your career has sort of taken you to the place you are right now. So can you give us a brief introduction?

Vijoy (05:31.16)
So if you really want to go back, Sebastian, I mean, I think I want to start from T0, where I got interested in computing in general. And surprisingly, it was in one of the spaces that people are pretty familiar with now, which is AI and chatbots. And this was back in the late, at the cusp of 89, 90s. I was in high school. dating myself here.

Sebastian Hassinger (05:49.11)
Mm-hmm.

Vijoy (05:59.74)
a friend of mine and I decided to build a chatbot. And this was on a BBC Micro, of all things, with 64 kilobytes of RAM. And we decided to build it using expert systems. And we tried to model it after Eliza, which was a chatbot developed in the 60s. said, yeah, maybe we can do one of those as a high schooler. But then we took it a step further and we said, let's make it in Hindi.

Sebastian Hassinger (06:05.29)
Mm, yep.

Sebastian Hassinger (06:17.25)
Yeah

Sebastian Hassinger (06:25.16)
Right. Right.

Vijoy (06:33.74)
and Hindi required a Unicode extension that had not been invented yet. So we went through all kinds of loops to get Hindi going, but that's how I got into this stuff.

Sebastian Hassinger (06:27.51)
Hmm.

Sebastian Hassinger (06:38.77)
Wow.

Sebastian Hassinger (06:46.65)
That's fascinating. That's really interesting. Was there, you know, I played with Eliza myself and you know, was, were there challenges in sort of adapting the architecture of the chat bot, the Eliza chat bot to Hindi where there are linguistic challenges other than the Unicode?

Vijoy (07:02.74)
So it was expert systems based. There was a lot of hash tables and if then else's and things like that. But it also led to some very incredible, like interesting behavior because there were no guardrails. And the chatbot was actually abusing back at us, which is as a high schooler, that is incredible fun. which is what got us into this, which is what got us into where I am today.

Sebastian Hassinger (07:13.75)
Yeah, right.

Sebastian Hassinger (07:18.65)
Right.

Sebastian Hassinger (07:28.69)
That's fascinating.

Vijoy (07:28.97)
talking about historical reasons. And that got me interested in computing in general. But then I ended up doing distributed systems and software as my career. And that ended up with this role where I am right now, which is all about pushing boundaries and looking at distributed and scale.

Sebastian Hassinger (07:56.09)
Right, right. Yeah, mean, outshift is a fascinating sort of model for incubating corporate innovation. that, were you sort of in those internal incubation, internal innovation type roles prior to the formation of outshift?

Vijoy (08:07.56)
No, I wasn't. I this is my first internal incubation role. I mean, I've been a CTO in the past, so there's a little bit of that, but CTO roles usually tend to be primarily architecture and vision, or they end up being all of engineering. And this is a little bit different because we are structured like a bunch of within a large behemoth like Cisco. And so we...

Sebastian Hassinger (08:27.47)
Right. Right.

Sebastian Hassinger (08:34.04)
Right.

Vijoy (08:41.94)
Actually, we have capabilities all the way from engineering to product management to marketing to customer success and business development, just like a startup would, but at smaller scale. And we start ideating and then we start building products and looking at product market fit. We start looking at product company fit because whatever we build needs to be scalable by bigger Cisco.

Sebastian Hassinger (09:03.17)
Mm-hmm.

Vijoy (09:09.24)
product company fit is a big, big, big metric for us. And as we walk on that journey, we start involving larger Cisco. So we started involving the product views. We started involving sales, biz dev, and so on and so forth from the larger Cisco to then take it to scale. So it's a pretty interesting setup. And it's been working for five plus years now, which is a testament to that setup as well. haven't been at Google. I've seen various models.

Sebastian Hassinger (03:55.0)
Right.

Sebastian Hassinger (09:19.46)
Right.

Sebastian Hassinger (09:29.27)
That's great.

Vijoy (09:30.46)
I've been at Cisco for a while and Cisco has had various models, but this has been working.

Sebastian Hassinger (09:36.97)
Yeah. What I mean, there's, there's all kinds of risks that need to be managed in these turn type of internal incubation programs. you know, as you said, I think one of the key ones is, is, product company fit. Like you, can invest time and effort in something that sort of doesn't have a future in the company that's, that's actually investing. So that's great as a starting point. Were there other sort of things that you've encountered that you realized you'd needed to,

provide additional guardrails or additional sort of risk mitigation for the projects that you're fostering?

Vijoy (10:18.16)
The way to think about this is, so we have a framework that we use to figure out which problems to go after. And one of the things that we first look at is, let's look at Cisco's core businesses, which is networking, security, observability, collaboration. Those are the four pillars. So look at problems that are adjacent to those pillars, both on a risk horizon as well as on a persona horizon. So can we look at...

networking but look at personas that are a little bit adjacent to who we talk to today. So that's a persona adjacency. And then we look at risk adjacency, which is, can we look at problems that are not in the platform side of the house, which is where Cisco products are, but problems that have technology risk or market risk? And two examples of those could be like technology risk is quantum. I mean, we know that there is

Sebastian Hassinger (11:10.11)
Right.

Vijoy (11:09.64)
an up and coming and promising piece of tech, but it's not solved yet. There's a lot of hard problems to solve for. And we don't know whether this thing will work out. So there's a little bit of that. So there's a little bit of technology risk. And then I would say on the market risk side, what we think is technology works, but it has not found a market yet. So we are looking at use cases. We're looking at stability of the market.

is not evolving every six months. I would put agentic AI in that space where the tech works, but how big is the market? We all believe it's big, but beyond coding, software development services, is there anything else that we've tackled? There are promises in science and drug discovery and so on, but we haven't solved it yet. So the market is still evolving, trying to figure that out. So that's the risk angle to it.

Sebastian Hassinger (11:58.08)
Yeah.

Vijoy (12:04.16)
There's also risks around process. How do we move fast like a startup? But also we are part of a large company. we actually took a framework with said, let's take a look at how Cisco acquires and integrate startups and let's flip that model to its head and really look at few things that we are paranoid about in terms of risks. So people and hiring, yes, as a large company.

Sebastian Hassinger (12:11.97)
Yeah.

Sebastian Hassinger (12:29.29)
Hmm.

Vijoy (12:33.74)
Revenue recognition, yes, as a large company. I those are things that you can't mess around with. As a publicly traded company. But everything else, yeah, you can bend the rules a little bit. So we did that and I think we have a pretty good rule book to follow at this point.

Sebastian Hassinger (12:54.55)
That's great. And it really draws on the history of Cisco as a very adept and high frequency acquirer of other companies. I was around during the dot com, I'm dating myself. I'm around during the dot com era. And I remember John Chambers being sort of the champion of growth through acquisition at that stage. That's very smart to draw on that of corporate cultural learning. You mentioned, know, quantum already.

I, let's jump into that because one of the reasons I really wanted to talk to you is as far as I know, like there's something really compelling about quantum in the context of a corporate incubator of this nature because, you know, because there's so much uncertainty about the timelines, because there's so much uncertainty about the use cases in some ways it feels like, you know, an old school, almost, know, Bell lab style, incubation.

model is what's needed because there are sort of fundamental challenges to address this. Do you sort of see this as more like foundational research or basic research or is this still in the model of a startup but with sort of more uncertain runways and timelines than the other projects you're working

Vijoy (14:07.68)
So one of the hats that I wear, Sebastian, is also Cisco Research. So there's Outshift, which is the incubator. And just sitting beside that is also Cisco Research. And also I've got Cisco Developer Relations, which is DevNet. Because they somehow, all of these three functions, and open source, by the way, as well. So all of these things are sort of, they all need to come together in some of these incubation efforts that we have been doing.

Sebastian Hassinger (14:12.03)
Mm. Right.

Sebastian Hassinger (14:30.96)
Right.

Sebastian Hassinger (14:39.11)
Yeah.

Vijoy (14:37.28)
Quantum actually started within Cisco Research and still sits primarily within Cisco Research. But what we've realized over the past year is we are spending more and more of outshifts, resources, skills and time on this problem, which basically is telling us that we are pivoting from the physics research part of it.

Sebastian Hassinger (14:50.83)
Right.

Vijoy (15:11.68)
to a systems operability deployment part of the equation. And the one thing that I wanted to just point out right off the bat is, Cisco, since we talked about the four core businesses or four pillars that Cisco has and the PCF, the product company fit, we are looking at quantum computing from the quantum networking and quantum security side of the angle, as you can imagine. And one of the surprises that we ran into, good surprises,

Sebastian Hassinger (15:31.93)
Of course, yeah, yeah.

Vijoy (15:37.66)
of the past year or so is that quantum networking is actually not that far out. Whereas quantum computing is still being figured out, quantum networking is more like here and now than quantum computing.

Sebastian Hassinger (15:58.31)
Well, it's so interesting to me because you've really got a split horizon in a way because post quantum cryptography, quantum key distribution, those are existing technologies where there's more of the more of the work is on product market fit and productization and finding the right way to deploy those technologies to solve customer problems. But actual entangled photons communicating like flying qubits, sometimes people call them.

That one's probably even further away in some sense at scale than quantum computing. Would you agree?

Vijoy (16:33.14)
Well, I would say that flying qubits at scale is still some time away. Though there is a tremendous amount of pressure. It is one of those burgeoning sort of rising tide of innovations that are happening right now. Where I believe that there is a transformer or a chat GPD moment in quantum about to happen. Because if you just look at the number of

innovations in the quantum space over the last two years, compare that to the last two decades or three decades, I mean, we are seeing this exponential curve. I think any moment now we're going to hit this point where we're like, man, this is it. And it's suddenly now practical to use this piece of technology and the risk, the tech risk has just gone down. But having said that, I mean, yes, flying qubits at scale, maybe a little bit further out, but flying

qubits not at that much scale, but in a way that can be leveraged for classical use cases is where I think the sweet spot is, which is what is giving us a lot of excitement where if you can do entanglement based quantum networking, but use that quantum network to solve for some classical use cases, like you pointed out a few, there are a whole bunch of security use cases.

Sebastian Hassinger (17:34.66)
Right.

Sebastian Hassinger (17:48.65)
Hmm.

Vijoy (17:55.42)
the whole bunch of coordination and synchronization use cases that a quantum network, is few qubits, is good enough to solve for them building out this large hundreds of thousands of qubits quantum environment.

Sebastian Hassinger (17:59.42)
Hmm.

Sebastian Hassinger (18:05.63)
Right, right.

Sebastian Hassinger (18:11.23)
That's interesting. when you say synchronization coordination, do you mean literally like, you know, keeping all the clocks on a network all in sync that that sort of synchronization?

Vijoy (18:28.74)
That is a great use case. So one of the use cases that we came across is, let's say you're building these really large global databases. And who's doing that? they got it. My previous life, I was at Google, they had Spanner. And the way you synchronize clocks, Spanner is like catching your nose the other way around. There's so many.

Sebastian Hassinger (18:28.56)
Hmm.

Sebastian Hassinger (18:35.71)
Who would be doing that? Everybody.

Sebastian Hassinger (18:50.47)
Yeah.

Vijoy (18:52.38)
So many infrastructure elements that go into just making sure that the clocks are synchronized, that's a clock synchronization use case. The other use case that we just came across in the clock synchronization space is when we were talking to ESNet, this is the energy sciences network, we were talking to Inder and team at Lawrence Berkeley. And they were describing this use case where you have all of these telescopes, radio or visual.

Sebastian Hassinger (19:20.17)
Hmm.

Vijoy (19:21.54)
scattered across the globe, taking pictures of space, the same little segment in space, the same little square, but from different angles on planet Earth, so that you can see movement. So you can sort of say, yeah, there is a meteorite coming our way. In fact, I can imagine there are many, many more things that they are searching for, which are human-made.

Sebastian Hassinger (19:39.44)
Right.

Sebastian Hassinger (19:41.12)
Hey, yeah.

Sebastian Hassinger (19:48.73)
Yeah.

Vijoy (19:48.22)
which are not meteorites, but let's not go there. But to do that, you would need all of these clocks to be synchronized across the globe. And so how do you do that? So there are these large global synchronization problems. So that's clock synchronization, but beyond clocks, there is in fact one use case that has been talked about quite a bit so I can talk about it, but decision coordination is a really, really good problem space. And so think about high frequency trading and...

Sebastian Hassinger (19:53.62)
Right. Right.

Sebastian Hassinger (19:58.09)
Hmm.

Sebastian Hassinger (19:59.52)
Hmm.

Sebastian Hassinger (20:14.47)
Right.

Vijoy (20:17.13)
The way high frequency trading works today is there is a guarantee of fairness that an HFT provider gives to their customers. And the way they can provide that fairness is by putting all of your customers in the same physical environment where the speed of light is not a problem. Now you and I know this, that how do you scale? I mean, you fit all your customers in one location, especially in a location like New York.

Sebastian Hassinger (20:25.57)
Right.

Sebastian Hassinger (20:35.45)
Right. Yeah. Yeah.

Sebastian Hassinger (20:42.06)
Yeah.

Vijoy (20:45.24)
where you have a premium when it comes to space, how do you expand and how do you grow that domain? Well, you create a data center in New Jersey. But speed of light matters for algorithmic trading across New Jersey and New York. So how do you synchronize actions across New Jersey and New York or for that matter anywhere within the US or anywhere else?

Sebastian Hassinger (20:57.26)
Interesting.

Vijoy (21:15.36)
without incurring the speed of light tax. And you do that by entanglement and teleportation. And so the number of qubits you need is not so many.

Sebastian Hassinger (21:15.17)
Right.

Sebastian Hassinger (21:19.04)
Interesting.

Sebastian Hassinger (21:22.89)
Right. I was going to say it's, it always feels like there's this challenge at this stage we're at even as much progress is happening as you said, with finding the use cases that we've been so set on use cases that drive massive volumes of data at very high frequency. And this is almost the exact inverse, right? It's small numbers of, of, data points with very high dimensionality. So it's like the interior space has to be very large to fit.

the quantum use case, but the actual number of bits that you're moving in and out of the device has to be really low. it sounds like, you know, that's an interesting one where that may be the case. that sort of, were you looking for those characteristics in particular and sort of hit on this use case?

Vijoy (22:07.99)
So we're looking at it as a spectrum. if you talk to us and you ask us what's our vision for, what is Cisco's vision for quantum computing? And our answer is we are in it for the distributed quantum computing game. We are a distributed computing company because through our network, you can take many similar or dissimilar computing devices, connect them.

put a layer of software above it and enable scale out distributed computing. And we've done that with classical computing and yeah, all the hyposcalers build out computing that way. Why should it be any different in the quantum space? And to do that, you need the quantum network. So the North Star vision is to build out the quantum network for the data center so you can scale out and you can get to hundreds of thousands or million qubits of usable qubits.

faster than any singular vendor would do. So that's the not star. The other thing that allows us to do is you can put disparate technologies on the same fabric. So instead of betting on trapped ion or whatever, all of the eight or nine technologies that exist there.

Sebastian Hassinger (23:34.93)
Yeah, seems like more of them every day actually, or more variations. Yeah.

Vijoy (23:31.68)
more than every day, right? So don't bet on them. And in fact, what you and I hear about is like, each one of those technologies are probably good for certain things. So they will all coexist. And how do you make them coexist in the same fabric? By making them connect through a common fabric. And so that's our North Star vision. And then connecting more of those data centers, connecting sensors and building up the economy. But that's the North Star vision.

Sebastian Hassinger (23:41.46)
That's right.

Sebastian Hassinger (23:52.83)
Right.

Vijoy (24:04.74)
But what we realized is as we building those pieces of technologies out, hardware, protocols, applications, software, as we were building those out, those same protocols and software and hardware pieces can solve these classical problems today. The decision coordination, the clock synchronization, the eavesdropper proof security, the secure position verification. So a bunch of security problems, bunch of synchronization problems.

Sebastian Hassinger (24:24.17)
Hmm. Hmm.

Vijoy (24:28.51)
And so what excites us is the here and now of using that stack or solving it in the classical domain today. And that's where we're seeing a lot of activity as well from the hyperscalers, from financials, from energy companies who are all coming in as design partners. Yeah.

Sebastian Hassinger (24:49.81)
Right. was going to ask, you mentioned positioning and that brought the topic of sensing to mind. feels like quantum sensing is again, sort of ahead of where computing is because you can have these sophisticated but relatively simple devices as compared to like hundreds of qubits working in a single system, but you still have the problem of getting that quantum data.

into your classical infrastructure so that you can process it so you can use it in your application. Is that one of the categories you're sort of looking at is connecting quantum sensing applications? Yeah.

Vijoy (25:32.26)
Right, so to us the roadmap on the quantum networking for quantum use cases side goes from building out this fabric, quantum network fabric, within a data center which actually allows for heterogeneity and operability and distributed scale out computing, then connecting multiple of those quantum data centers, almost like a data center to data center quantum network, and then connecting out to even

quantum sensing that enables the quantum internet. So that's for the quantum computing side. And then in parallel using that same stack for the here and now problems, which is, I solve for these coordination, synchronization and security use cases in the classical domain today, or leveraging the same quantum networking stack? So to us, that is the entire gamut of what we're going after. But the here and now in the classical cases,

Sebastian Hassinger (26:13.57)
Mm-hmm. Mm-hmm.

Vijoy (26:30.02)
use cases are actually the ones where we're seeing a lot of traction. Because people are coming to us trying to solve these high frequency training problems, trying to solve these security problems and position verification problems. Because you're not sending quantum data per se, but you're leveraging the problem, the properties of entanglement and teleportation, or the collapse thereof in the security side.

Sebastian Hassinger (26:33.63)
Hmm. Mm-hmm.

Sebastian Hassinger (26:50.67)
Right, right.

Sebastian Hassinger (26:55.53)
Right. Right.

Vijoy (26:53.84)
to solve these problems and so you can do it way more simply. You don't need these mega cryo chambers, you don't need these big detectors. The problem just simplifies dramatically, which is what makes it pragmatic today.

Sebastian Hassinger (27:00.89)
Right, right.

Sebastian Hassinger (27:06.49)
Yeah. Yeah. That's interesting. And you mentioned, you know, people coming from HFT and, you know, asking for help with these problems that sort of begs the question about collaboration. mean, these, is it, you know, you've got sort of in these kinds of internal incubation efforts, you've got three potential sets of collaborators, right? There's sort of customers. there's, I there's startups, there's actual like pre seed or seed stage startups and other.

partners, then there's the academic research world, which in quantum in particular, but also in AI and ML, because there's still very early stage technologies, there's a lot of activity going on in the research universities, in the national labs. Are there sort of preferred modes of collaboration for outshift?

Vijoy (28:02.24)
It's all of the above. we are part of, so OutDrift is an internal incubator at Cisco, but we sit within the Chief Strategy Office. And the Chief Strategy Office, of course, has strategy for the company. It's got the corporate development arm, which is investments and acquisitions. And then it's got incubations, which is my group. And we work pretty closely together. So for example, on quantum,

Sebastian Hassinger (28:01.65)
Okay.

Vijoy (28:28.58)
We have investments in some companies and startups because like we discussed, mean, the physics is still not solved yet. There's some parts which are solved and some parts which are not solved. And so we need the partnerships and we need a thousand flowers to bloom because we don't know which one will go where. But we are building out on our vision on the hardware. We're building on our vision of the software. Then we're also working with larger vendors on the computing side, you can imagine, because

A network needs computing and computing needs networks. There is a symbiotic relationship there. So we're working with some very large vendors, which I can't talk about right now, but there's some interesting announcements coming in that space. And then we are working with these design partners because they're not customers yet. Like I said, we are on the quantum side, we're bridging between research and incubation for now. So everything that we've announced is research prototypes. We are very careful to say that because things are not available at scale yet.

Sebastian Hassinger (29:28.09)
Right.

Vijoy (29:23.99)
And so they're design partners and most of the design partners that we discussed are high frequency trading or financials or not just as such as FD, but other security use cases, the federal agencies, the energy companies where security synchronization at scale, et cetera, matters to them. I mean, it doesn't matter to a person like me buying socks on Amazon. I mean, it's not going to matter to me, but there are these really.

Sebastian Hassinger (29:51.17)
Yeah, not yet. No.

Vijoy (29:55.38)
That's not the first use case, but there are use cases where this matters today. So those are the kinds of design partners coming in.

Sebastian Hassinger (30:02.19)
Right, right. And the academic world, do you collaborate with researchers at universities? Yeah, yeah, yeah, yeah. That's my normal guest. That's the normal pool I draw from from guests. That's why I'm asking the researcher.

Vijoy (30:05.82)
sorry, I forgot to mention that, yes. There's a lot of academic...

Vijoy (30:15.35)
Yeah, no, no. So there's a lot of work happening with academia. In fact, the one, the first announcement, almost like the coming out party that we had on quantum from Cisco, was this piece of silicon that we developed in collaboration with UC Santa Barbara. And this is a photon entanglement source chip. And what it does is actually generates 200 million entangled pairs a second in chip.

Sebastian Hassinger (30:47.67)
Wow.

Vijoy (25:36.8)
Yes, when you deploy, there are other reasons why you will get less than that. But within the chip is going to generate 200 million entangled photon pairs a second. But it does that at room temperature, at telecom frequencies, and at really low power. And these three things are important because you can throw thousands of these chips at a problem. It's low power, which implies low cost.

Sebastian Hassinger (31:03.61)
Right, right.

Vijoy (31:21.74)
It's at room temperature and it's at telecom frequencies which you can deploy these in current environments without ripping and replacing your current fiber infrastructure, which is what allows us to do these classical use cases without specialized equipment in place.

Sebastian Hassinger (31:27.37)
That's incredible. Yeah. Yeah.

Sebastian Hassinger (31:43.85)
Yeah. Do you have, I mean, there's a number of sort of pilot quantum networks up and running in Illinois, in New York, in Europe and a bunch of places, in Asia and Singapore has got a quantum safe network. Do you view those pilots as potential collaboration test beds or do you have your own test network or a combination of both?

Vijoy (26:52.8)
It's a combination of both. So we are actually collaborating with quite a few of those who joined up a of consortiums in this space. But the one thing that we are also realizing is that they all, quite a few of them started out from, let's solve QKD, quantum key distribution. And we are of the viewpoint that yes, security is a massive problem to tackle. But instead of going all the way to QKD, you can actually do quite a bit.

Sebastian Hassinger (32:08.86)
Right.

Sebastian Hassinger (32:29.59)
Yeah. Right.

Vijoy (32:31.05)
before that and before that in the technology sense. So a simple thing like, since our chip again sends entangled pairs at telecom frequencies, you can send those entangled pairs on a regular classical fiber alongside classical IP. So somebody is actually eavesdropping and tapping into your network. In any classical network, I can tell you provably,

Sebastian Hassinger (33:01.79)
Really?

Vijoy (32:59.99)
there's somebody in your network. And so you can leverage standard TLS, the NIST compliance, PQC, quantum cryptography standards or QKD. I mean, you can go anywhere in that spectrum. Even with standard TLS, I can tell you, man, there's intruder in your network. And so we are a little bit broader and we are trying to meet people where they are.

Sebastian Hassinger (33:11.13)
Yeah.

Sebastian Hassinger (33:23.95)
That's fascinating.

Sebastian Hassinger (33:26.63)
Hmm. Mm-hmm. Mm-hmm.

Vijoy (33:29.82)
And that's why we have this strategy which says, this chip is going to give you quantum networking on the infrastructure that you have today. And so we are also building out a software stack which says, you can deploy the software stack alongside your current networking stack to turn on quantum networking capabilities alongside what you already have.

Sebastian Hassinger (33:59.55)
Well, let's talk about future proofing.

Vijoy (34:00.34)
We all know Sebastian that the world is not Greenfield. So if anybody comes to us and says, let's start from scratch, that is not the answer anybody wants to hear.

Sebastian Hassinger (34:00.73)
yeah.

Sebastian Hassinger (34:08.29)
Yeah, yeah, yeah, yeah. I always think how unique this point of evolution of quantum computing and quantum technologies in general is because in one sense, we've reset the clock to the 1940s. And in the other sense, we have the cloud and open source and open standards and GPUs and massive amounts of compute and storage to help us. And also the pattern, right?

Vijoy (34:27.22)
Right.

Sebastian Hassinger (34:34.93)
The history doesn't repeat, but it does rhyme. It feels like we're able to draw on the lessons of the past. And this is actually where we can end on this question, because I've just been wondering about this as we've spoken from the beginning. You mentioned distributed quantum computing and the fact that there are these multiple modalities that likely are going to persist over time and have different roles in these distributed computing infrastructures.

Do you, since your focus is on networking communications and security, do you think it's possible that you'll expand into computing in the context of transduction, right? mean, taking a microwave qubit or a trapdion or an atom-based qubit and connecting it to another physical qubit somewhere is going to require photons. It's going to require flying qubits. It's going to require telecom bandwidth.

you have all the ingredients to be the glue in these distributed systems. Is that sort of in your long range thinking is getting into computing in that sense?

Vijoy (35:43.58)
the way we think about the quantum computing world and you hit upon something which is the way we think about quantum computing is the supercomputer of the future is going to be it'll comprise of cpus and gpus and dpus all kind of xpus accelerated pus on the classical side as well as qpus or quantum processing units on the quantum side and a combination of all of these

Sebastian Hassinger (35:58.57)
Yep.

Vijoy (36:13.18)
is what's going to drive the drug discovery of the future or the material discovery of the future and make your or my life much, much, much better to where we are today. So that is the viewpoint. But the way we like to phrase our problem is by saying quantum computing has a scaling problem and classical computing has a physics problem. And the quantum network is actually the solution to both of those problems.

Sebastian Hassinger (36:34.85)
Hmm. Hmm.

Vijoy (36:41.68)
And so to us, the quantum network is actually the bridge between all of these problems that we're facing today in scaling out supercomputing of the future. But remember how we started? This whole thing is in the tech risk bucket. And one of the biggest problems to solve to make this the grand challenge here is transduction. And transduction

Sebastian Hassinger (36:59.05)
Right.

Vijoy (37:11.98)
And like I joke within the team is it's always the goddamn NIC card. So really think about it at a 10,000 foot level is a NIC card. It is the thing that connects your computer. Really it's not everyone, but it is a NIC card. So it's always the NIC card. I'm sorry, but it is the bridge between computing and the network. And it is a grand challenge.

Sebastian Hassinger (37:13.82)
Hey!

Sebastian Hassinger (37:19.67)
Yeah, a really exotic one. Yeah.

Sebastian Hassinger (37:29.25)
It kind of put in those terms, it makes a lot of sense that Cisco would be right at the center of this grand challenge, doesn't it? So thank you. That's really a fantastic framing from it. any sort of last sort of perspective, any sort of the goals you have for the near future or the types of things that you see in the next 12 to 18 months for the industry or for outshift?

Vijoy (37:36.64)
That's right.

Vijoy (37:54.18)
I think I believe that we are moving from a very physics based approach to quantum to where we are heading into a systems based approach for quantum, which means thinking about operability and interoperability and a Brownfield environment, not just within quantum, but across quantum and classical. And we are at the cusp of solving those problems. And once those get solved, the transformer or the chat GPT like moment.

for quantum is about to happen and it'll be one day and everybody's going to wake up and realize there is a step function that just happened, what do I do about it? And unlike other step functions in computing where most of them have been software based and you could catch up, this one is going to hit you hard. So you better prepare, is would be my message to everybody listening.

Sebastian Hassinger (38:51.61)
Yeah. Yeah.

really, really wise words. So thank you very much, Vijoy. This has been terrific. I really appreciate your time.

Vijoy (38:58.48)
It was great fun. Thank you, Sebastian.