KB [00:00:10]:
What’s up everyone? It’s KB and I’m on the go at Cisco Live at Mandalay Bay Resort and Casino here in Las Vegas. This week, some of the biggest conversations shaping networking AI in cybersecurity are all happening at once. This week is where operators, engineers and executives get into the details, what’s working, what’s not, and what’s actually changing inside enterprise environments at scale. We’re hearing a lot about AI driven networking, automation and visibility, but the real question is how does that translate into security outcomes, operational resilience and ultimately cost? Because behind every intelligent network claim, there’s still a human team trying to make sense of complexity. Across the next few segments, we’ll be speaking with leaders on the ground here in Vegas, unpacking what’s real, what’s noise, and where the industry is genuinely headed next. This is KB on the go from Cisco Live, Las Vegas. Let’s get into it. Joining back in person is Tom Gillis, SVP and General Manager, Infrastructure and Security Group at Cisco.

KB [00:01:21]:
And today we’re discussing threat mitigation and vulnerability shielding. So Tom, thanks for joining me back on the show and welcome.

Tom Gillis [00:01:27]:
Yes, thank you, thank you for having me.

KB [00:01:29]:
So I heard your presentation earlier today. I really like how animated you get and this is the third time I’ve interviewed you now, so I think it’s, you know, the same characteristic shines through. But I want to maybe start with every time a major vulnerability is disclosed, security teams find themselves in a race against tack it. So would you say the industry is losing that race or organizations simply overwhelmed by perhaps the sheer volumes of vulnerabilities they have to deal with?

Tom Gillis [00:01:57]:
Well, the best answer for you is both, which is yes to both points. It’s not an either or. So what we have seen as a steady trend, and I think this is self evident or clearly observable that the time from when a vulnerability is announced to when we see activity around that, that has gone from months to weeks to now, we see consistently within hours. And it’s AI automation that drives that. So vulnerability gets announced within hours. We see people exploiting that vulnerability because the AI tools will create the exploit. Okay, it is not reasonable to patch all your infrastructure in hours. Okay, so that’s, that’s an initial problem.

Tom Gillis [00:02:37]:
Now that problem is compounded with the current generation of models. Mitos is one of them, but not the only one. These models are so capable at finding vulnerabilities that in the previous world we would find in a Cisco device, two vulnerabilities, five vulnerabilities and these are, I’m not going to give you a particular number, I’m going to say a step function increase in the vulnerabilities. And you can see this in the public domain. These tools are finding vulnerabilities in Linux, they’re finding vulnerabilities in SSH. Microsoft announced 125 vulnerabilities in SharePoint. So combine that to the previous dynamic of it was hard to respond to one vulnerability in the time that attackers would be able to exploit it. We now have enterprise customers that are dealing with thousands of vulnerabilities.

Tom Gillis [00:03:32]:
And this is not a Cisco problem, this is an industry problem. And any company that builds software has this issue, which means the enterprise is going to be overwhelmed with every single vendor on the planet coming and saying we have an emergency security update, you need to update now. Starting, you know, now in June. So this is what I call the summer of turbulence.

KB [00:03:58]:
Okay, so then, okay, so the part that is interesting to me is over the years the advice has always been like patch, patch, patch. But then many organizations can’t really afford like downtime. I think one of your customers spoke about it today.

Tom Gillis [00:04:11]:
Correct.

KB [00:04:11]:
Reboots, service interruptions, etc.

Tom Gillis [00:04:14]:
Yep.

KB [00:04:14]:
So then talk me through. Has vulnerability management reached like a breaking point now?

Tom Gillis [00:04:19]:
Yes, yes. And it’s not just vulnerability management. The part that is, that has to fundamentally change is the operating model for infrastructure. So for as long as I’ve been doing this, which is decades, there’s been a philosophy of I’m going to design infrastructure, particularly data center, I’m going to validate that design, I’m going to deploy that design and then I’m not going to touch that design for as long as I possibly can think months, slash years, like customers will do a major upgrade in the data center infrastructure once a year and we as a vendor would release software like you know, every six months. Right. Big one and then a little one. That has to fundamentally change. So we are now entering a world where these models are discovering all these vulnerabilities that just simply patching is a real, real, real challenge.

Tom Gillis [00:05:10]:
And this isn’t a one and done type of thing. The models continue to find new vulnerabilities. So we need to not just react to this crisis. We the industry need to adopt a new operating model. And I will argue as an industry leader, Cisco is setting the direction there. So what we announced at the show and we demonstrated live to the audience, we have a thousand customers that have actually deployed this. So it’s in customers hands is what’s called Live protect. And live protect is a compensating control that can shield critical vulnerabilities between patch updates.

Tom Gillis [00:05:49]:
So remember the two problems you talked about right at the beginning where, like, oh, my God, more vulnerabilities, and they’re being rapidly exploited? This is the answer to that problem in that when we announce a vulnerability, it’s called a p cert, we will give you the patch, and where applicable, we will give you a shield. And a shield is designed to be deployed on a switch, a router or firewall without disrupting the switch or router or a firewall. No change control window, no reboot, no downtime. Bump, bump, bump.

KB [00:06:20]:
Shorter though, like, in between each patch.

Tom Gillis [00:06:22]:
Yes. So the moment we discover that we announce a vulnerability, the shield will be available at that time, and then it’s up to the customer to choose how to deploy them. But if you look at the way we’re engineering these shields, they are extremely targeted. They’re very, very specific. Don’t let this software process access this file. Which means they have a very, very low false positive rate and a very, very low performance impact. And so once git customers get comfortable that these shields can be deployed without disrupting operations, the intention is you let them happen in an automated fashion. So we announce a pserve vulnerability, a shield automatically gets deployed.

Tom Gillis [00:07:04]:
Customers have the ability to provide oversight over that. We’re never gonna determine that we let our customers do it. But the intention is as close to a real time protection, and that’s how we can stop the automated exploitation of these vulnerabilities.

KB [00:07:17]:
So do you think just given your experience, there’ll be more like shields than like, patches? Because it was how much? Things change quickly. Time frames, et cetera.

Tom Gillis [00:07:24]:
Let’s be really clear on that. A shield is meant to be a bridge between updates. Okay, so when I do an update, all the previous shields go away. You’re not going to have shields from the last update that you’re still running. So it’s not like firewall rules where you put a shield in and then it stays in forever, and then you forget who wrote the shield and why it’s there if it rods. Yeah, exactly. When you do an update, all shields are gone. And I’ve even contemplated putting some sort of an expiration point on shields because we don’t want customers to think, oh, I got a shield up.

Tom Gillis [00:07:55]:
I’m good. A shield is a compensated control. It’s an emergency measure. It’s a finger in the dyke. But a skilled attacker can find ways around a Shield, Right, right.

KB [00:08:05]:
How are you sort of approaching that if an attacker can find their ways around a shield?

Tom Gillis [00:08:09]:
More frequent updates. So it’s a one, two punch. Like we’re going to be introducing more frequent updates that fix the core vulnerability. And the shield is a bridge to get from where we are to where the customer needs to be. Now what your question was sort of alluding to is, well, the reason why customers update infrequently is because the updates are kind of hard and it takes hours and they don’t always work. And that also is an area that we’re attacking. So my colleague Anurag demonstrated digital twinning capability. So with a digital twin we can use.

Tom Gillis [00:08:43]:
This is the kind of a silver lining of all this AI, you know, sort of conundrum. AI allows us to create a model of your environment so that we can test the upgrade, the actual executable, against the policy configuration of a specific part of a customer’s environment and gain confidence in that upgrade. Put those two things together, we’re going to move into a world where updates just become low friction, high velocity. It’s not unreasonable that. Let’s give it a little time. Two years from now, updates are just automatic. It’s the same way if you’re an iPhone user. When’s the last time you updated your iPhone?

KB [00:09:24]:
When it asked me to.

Tom Gillis [00:09:25]:
Yeah. Or it just happens, right? You wake up in the morning and it’s just updated. I don’t even know. Right? Yeah, right. And so the world we’re moving into, which quite frankly, that’s how the cloud works, right? That’s how Kubernetes works, is it’s this idea of lots of little tiny changes on a continuous basis. The industry calls it cicd, Continuous integration, continuous development. We’re bringing those principles to infrastructure and it’s a huge step.

KB [00:09:49]:
So then just for a moment, staying with the digital twin, do you think because of that, if people can do this modeling or forward modeling and then will reduce potential attacks?

Tom Gillis [00:09:57]:
Well, what reduces the attacks is having an up to date infrastructure because we know where the holes are and we plug those holes with new code.

KB [00:10:04]:
Doing it in advance though, so that’s like better than what’s currently sort of happening.

Tom Gillis [00:10:09]:
Well, the advantage of the digital twin is that it makes it easier to do the upgrade more frequently. Instead of once a year, we’re going to do once a quarter, maybe once a month even. Like it just lots of little tiny steps. You know, you can’t break a vulnerability that has been fixed in the code. Run, run and Then live protect is the bridge for what?

KB [00:10:26]:
Oops.

Tom Gillis [00:10:26]:
We discovered a new thing. We’ll give you that kind of like, near real time coverage that can plug the gap until your next upgrade.

KB [00:10:34]:
And then with the next upgrade, do you envision that those windows will just get shorter and shorter and shorter, just the way in which the world’s moving now?

Tom Gillis [00:10:40]:
Yeah, I mean, there’s some natural equilibrium where, you know, today customers upgrade a lot of times. Once a year. That’s clearly not enough. What’s the right frequency? Is it once a quarter? Is it once a month? We don’t know. We’re going to figure that out. And that will just kind of evolve based on what the models find and how the software works and what customers tell us. So, you know, I often use the word we. What I’m referring to when I say we is I was like, we, the industry, like, our customers are in this with us as partners, trying to, like, sort through this giant conundrum.

Tom Gillis [00:11:14]:
Yeah. Which is it’s a crisis right now. Like, the scale and scope of the vulnerabilities that every customer is dealing with right now is unprecedented. And I would say advantage Cisco for having this capability, but we together have to mature it, we have to propagate it out in the market, and we have to kind of drive human change, which is adopting this new operating model that is dramatically different than the operating model that people have been working on for decades. Decades. Right. And so change like that overnight is hard to do. But when we have really good media, podcasters that are helping us get the message out, that goes a long way.

Tom Gillis [00:11:51]:
Yes.

KB [00:11:52]:
So final question, because I know you’re on a roll.

Tom Gillis [00:11:54]:
Yeah.

KB [00:11:55]:
What do you think sort of now, do you think customers know, like, what they are dealing with in terms of scale and scope, or now’s your time? Really articulate that.

Tom Gillis [00:12:03]:
It’s one of these things where I get customers that ask me like, hey, how bad is bad? And my answer back is bad, you know? You know it’s bad. But customers still ask. So I think there’s, like, I did a podcast on this and we were expecting 5,000 users. We’ve got like 40,000 users. So the appetite to learn and be educated on this new operating model is very, very high. And again, as a market leader, that’s a big part of what Cisco does and does well is we educate everyone. Right. The CCIE and the Cisco Academy.

Tom Gillis [00:12:36]:
Like, we’re super good at pushing that education out there. There’s a lot of education that has to happen really, really fast.

KB [00:12:45]:
So joining Me now in person is Ramana Kompella, head of Cisco Research and Cisco Fellow. And today we’re discussing quantum networking. So, Ramana, thanks for joining me and welcome.

Ramana Kompella [00:12:54]:
Thank you so much, Karissa.

KB [00:12:55]:
Okay, so I’ve been at the conference, Cisco Live conference, and I’ve heard the word quantum actually starting to come through the conversation more. And perhaps when people hear the word quantum, they immediately think of, like, quantum computers. But Cisco was focusing on quantum networking. So why do you believe the network, not the computer itself, will be the key to unlocking the next sort of era of quantum technology?

Ramana Kompella [00:13:24]:
Great question. Yeah. So if you’ve been following quantum computing, so there’s a tremendous amount of promise to the kinds of problems that quantum computers can solve that are extremely hard for classical computers. So quantum computing has that promise and has had the promise for a while now. Now, if you look at the quantum computing industry, we still are at hundreds to maybe thousands of qubits. Okay. So these are the qubits instead of classical bits, you have the quantum bits. Right? So the capacity of a quantum computer is not sufficient to solve some of the hard problems that it holds promise.

Ramana Kompella [00:14:06]:
Right. For things like drug discovery, optimization problems in the logistics and financial world, weather modeling, new materials, molecular simulations, all kinds of stuff. Right. So you need orders of about hundreds of thousands of qubits to maybe even millions. So There is a 100x to the 1000x gap between the number of qubits that we have versus the number of qubits that we want.

KB [00:14:33]:
Okay.

Ramana Kompella [00:14:34]:
And of course, you know, quantum computing vendors are not sitting idle. They’re continuously evolving their roadmap to add more capacity and so on. But our strategy at Cisco is to build a quantum network that can truly interconnect all these different quantum computers so that just like how classical computing has scaled out, we want quantum computing to also scale out with the help of, like these, I don’t know, hundreds and thousands of small to medium sized quantum computers interconnected with the help of a quantum network to build one logical distributed quantum computing platform. So that’s where, like, you know, that’s why we’re investing, and we feel that strategy is probably a lot more faster and practical compared to just keep building bigger and bigger quantum computers.

KB [00:15:24]:
So that’s, that’s what you meant by the practical side of it, rather than just keep building bigger and bigger computers?

Ramana Kompella [00:15:30]:
That’s right. That’s right.

KB [00:15:32]:
And then even like this on the qubit stuff, I know that I’ve done a couple of interviews around it, so I’m not some pro. But what do you think? How would you like describe that for people that are still wrapping their head around, like, quantum stuff?

Ramana Kompella [00:15:45]:
Right. So the way to think about it is in classical world, you have, you know, the lowest level of information is encoded in zeros and ones. Right. Either it’s a electric pulse that is, you know, high or low. In the quantum world, it’s actually a superposition. It’s not 0 or 1. It’s actually 0 and 1 at the same time. The closest analogy I can give is think of it as a spinning coin.

Ramana Kompella [00:16:11]:
Okay. Is it heads, is it tails? Not sure. Right. It’s a combination of heads and tails. Okay. So quantum mechanics actually, like quantum particles, exhibit this kind of a superposition of states, which allows you to do certain forms of computation much, much faster compared to classical. Okay. Because it’s a superposition of states, you.

Ramana Kompella [00:16:37]:
You can actually explore a lot more computational space using quantum computers compared to, like, you know, classical computers. So a qubit, it’s actually much more powerful than a single bit because it’s actually representing probabilities of you being in zero or one state, essentially. And that’s actually like what unlocks the power of quantum computing.

KB [00:16:59]:
So then Ramana Cisco recently announced the universal quantum switch. So perhaps someone who hasn’t heard about it doesn’t know what it is for the first time, maybe share what sort of the breakthrough here. And then why do you think the industry should pay attention? And I asked that last part because someone who works in media across this space, like every person I interview, it’s like, oh, but they got to pay attention to this and then they got to do this. So, like, what draws the attention for people out there, given your experience in space?

Ramana Kompella [00:17:30]:
Yeah. So I think switching is foundational to scaling. So if you look at the Internet, Internet has scaled to billions of devices because of switches and routers. These are the things that interconnect billions of nodes. So it’s a scaling problem that is often solved with the help of a switch. Right. Otherwise you’ll have to have point to point connections between every pair of nodes. And that doesn’t scale really well.

Ramana Kompella [00:17:58]:
It scales quadratically. Essentially, if you have 10 nodes, you need 100 connections. If you have a million nodes, you need 10 to the 12 connections. It’s a lot. It scales much faster. So if you have a switch, you can actually build interconnect many different nodes with much smaller number of switches and routers. So that’s actually why scaling is enabled with the help of a switch. Now, for Cisco, you know, Cisco is analogous to switching in some sense.

Ramana Kompella [00:18:24]:
Right. So switching is really what was the foundation for Cisco. So we have done that for the past 40 years in the classical arena.

KB [00:18:32]:
Yes.

Ramana Kompella [00:18:32]:
And we want to be rightfully building this infrastructure in the quantum arena. So for us, therefore, the universal quantum switch is actually such a important milestone and a breakthrough. And what makes this switch actually quite unique is there’s a word universal there, which is actually quite interesting. If you look at the quantum computing world, right, you have quantum computers of different modalities, like superconducting trapped ions, neutral atoms, photonics, and so on and so forth. And every vendor is, of course, betting on different technologies, and we don’t know who the winner is going to be. Right. So we are building a switch that can talk to pretty much all these different types of quantum computers. So that’s why it’s universal.

KB [00:19:19]:
Got it.

Ramana Kompella [00:19:19]:
And they use different encoding types, like, you know, something called frequency bin, time bin polarization. These are different ways of encoding quantum information. So our universal quantum switch removes any kind of limitation on the quantum processor to choose whichever modality. So if A wants to talk to B and A and B are of different types, A chooses this modality, B chooses something else. Our quantum switch can actually translate between the two, and that’s actually pretty powerful. And it becomes kind of a universal switch that can interconnect arbitrary types of quantum processors out there.

KB [00:19:56]:
And then on that note, because even earlier today, due to Patel, it’s talking about how it’s integrating with all these partners. So I’m obviously seeing a very common theme in this space. Even other major vendors are saying the similar sort of thing because now customers want to have options or they have that portability and stuff like that. So would you just say as a trend overall, that’s where the market’s sort of heading. People want to be able to have the variety.

Ramana Kompella [00:20:22]:
Yeah, absolutely. And, you know, different quantum computers have exhibit different properties, Right? So. And some algorithms are actually shown to be better suited for some modalities compared to the others. So because of that, right, like maybe the future, I mean, I don’t know for sure this is as a prediction, but maybe the future, we’ll have a data center which will comprise many different types of quantum computers. And if you don’t have a universal network fabric that can bring them all together, you would be looking at investing in parallel quantum networks for different clusters of homogeneous machines. And that’s not a very easy in terms of operations and cost efficiency and stuff like that.

KB [00:21:07]:
Okay, so then moving on now a little bit. So today’s quantum systems are being built using very different technologies. I was just sort of discussed sort of the flip side of that would be are we heading toward a fragmented quantum ecosystem? Because you said before some are better suited to others algorithms, et cetera. And then to add to that, is interoperability becoming the biggest challenge then that the industry needs to solve? That’s something that I’m often hearing. Observability, interoperability, all that sort of stuff. That’s like the main thing I’m hearing as an undertone in recent times.

Ramana Kompella [00:21:39]:
It’s the nature of evolution, right? Like you start out when you’re building your own product, you’re not thinking interoperability from day one. You’re thinking, I need to perfect my system so I can showcase some value. Right. But Cisco has had a long tradition of building open interoperable networks, right? That’s the glue that brings them all together and actually make it powerful than any one entity. Right? We unlock the collective effect, right? So the same way I think, you know, and rightfully like, you know, all the big players in the market are focused more on perfecting the art of building a quantum computer. And a bigger quantum computer. And a bigger quantum computer, right. So yes, at some point they will want to do the networking part, but that’s where like, you know, we have taken a substantial leadership role in establishing the foundations for an open interoperable network.

Ramana Kompella [00:22:35]:
So that becomes second nature to how we interconnect all these different quantum processors.

KB [00:22:41]:
If you look back over history, it was everyone outsourced it to like one provider, then they went to sort of point solutions. And now I’m sort of seeing again, it’s like we’ve spoken about earlier today, yes, the reduction of tools, we know that. But I’m sort of moving more towards people now wanting to have a single player that is going to look at all of these things because, I mean, it’s exhausting for a lot of people. And then the interoperability isn’t even there. And just looking back over the evolution of time, it’s like we’re coming full circle again. And history somewhat repeats itself, of course, all the time.

Ramana Kompella [00:23:13]:
History always repeats itself. But yeah, we are in the quantum networking revolution where we were like 40 years back in the classical networking revolution. So a lot of things need to be rethought about how we build quantum networks. And mind you, quantum networks are not the same as classical networks. They work very differently. They work on not transferring information directly but by transferring something called an entanglement, the end nodes. So entangled photons, entangled particles. So these entangled particles are used to teleport information from one processor to another processor, which is not the way classical works.

Ramana Kompella [00:23:52]:
Classical works by literally you taking a bunch of information, putting it in the network, and that information slowly crawls through all the switches and routers and eventually gets to the destination. Right. In a quantum world, that’s not the way it works. So the way it works is you distribute entangled photons to A and B that want to talk to each other. And once you have these entangled photons, they. That use teleportation. Teleportation. The actual quantum information is going through a quantum channel that like, you know, nobody can tap into.

Ramana Kompella [00:24:22]:
It’s a. It’s. It’s entangled states, essentially.

KB [00:24:25]:
So going back to come around 40 years ago, we were in like classical computing. I mean, I’ve spoken to people over the years and they’re like, Fontom will never get to a point where it was like, with like PCs and stuff like when, you know, Windows 95 came out. Do you think it will though? What are your thoughts there?

Ramana Kompella [00:24:42]:
Anybody’s guess. But I feel like the way to think about quantum is it’s a specific technology for some really hard problems. Okay. Ordering socks off of Amazon is not the hard problem that quantum computer needs to solve. You’ll still do using your classical CPUs and GPUs, and maybe your AI workloads will run pretty well on GPUs, maybe some combination of GPUs and quantum processors. But the bottom line is quantum is not a replacement for classical infrastructure.

KB [00:25:14]:
It’s an and.

Ramana Kompella [00:25:15]:
It’s. It’s an and. Exactly.

KB [00:25:18]:
Okay, that’s really interesting because I think that’s probably where people’s mind goes as well. They think like, oh, like quantum is going to be like PCs and stuff like that. But then going back to your point around this really specific use cases and hard problems. Can you give me an example of like, what does that look like? I know, like, for example, the whole geopolitics thing. Sure. Is there anything else in terms of the specific solution that this would help? Like, moving forward?

Ramana Kompella [00:25:43]:
Yeah. So a lot of times you need to, like, do things like in order to figure out whether a particular drug is going to work. Right. You need to run simulations at the molecular level. Okay. So now, of course, you can simulate molecular dynamics on a classical computer, but that takes forever and ever. Why? If it’s molecular behavior, you need to Tap into the quantum mechanics of the nature. Right.

Ramana Kompella [00:26:08]:
As opposed to simulating using inefficient means on a classical computer. So that’s a use case, where the use case is so important because it allows you to discover new materials. Like, for example, aerospace industry is always in quest for lighter, stronger materials to reduce the amount of carbon emissions and reduce, go long distances on lesser fuel, fossil fuel burning and stuff like that. Right. You have pharma industry always looking for new drugs and you know, for like cancer and other diseases. So you have industries that are always like doing these kinds of molecular level simulations. And quantum computers are really purpose. They’re actually the closest to how you simulate nature.

Ramana Kompella [00:26:53]:
So that’s actually a class. And the other class of problems is optimization. So optimization problems have been shown to have some. There are some very hard optimization problems that can be solved using a quantum computer. Because quantum actually allows you to search for a large space. It’s like optimality is basically a search problem. You’re searching in a vast space to figure out what is an optimal solution for your particular problem. And quantum computers are efficient at exploring the multiple dimensions of your search.

Ramana Kompella [00:27:33]:
that helps you basically get to more, better solutions for your, I don’t know, path assignment problem that airlines have, or figuring out your portfolio for financial trading or logistics and supply chain. Where exactly to procure parts. What should be the route, what should be the cost? There’s a bunch of these problems that can be formulated as optimization problems. So they can be solved with the help of a quantum computer.

KB [00:28:00]:
So in switching gears now. So people talk about quantum breaking encryption.

Ramana Kompella [00:28:05]:
Yes.

KB [00:28:06]:
So, I mean, this is a pretty big conversation, but what is the biggest story here? Because what are the real world applications that become possible when quantum systems can actually communicate and work together across a network? And also, I think maybe about a year ago, I interviewed someone, I think it was like a Q day. Like companies are now scared, like in the next five, 10 years of all this encrypted data, could that then be broken?

Ramana Kompella [00:28:30]:
Oh, yeah, yeah, yeah. So, okay, so let me start with like a perfect trifecta of like, you know, things that are shaping up in the industry, right? So, number one, quantum computers are getting bigger and bigger. Enter Cisco. We have a quantum network that we are racing towards that can actually make it even bigger. Right? Because we can stitch them all together and build a distributed quantum computing platform. And then algorithmic advances, Right. Recently somebody came up with a modification of Shor’s algorithm, which is the one that breaks RSA and cryptography, the number of qubits that are required has gone down by a factor of 10 using this new algorithm. Okay, so what that means is basically that we are in this perfect storm, racing towards breaking cryptography in the next like two to three years.

KB [00:29:22]:
Wow. Okay.

Ramana Kompella [00:29:23]:
So now, I mean, I don’t know if exactly like whether it’s two years or three years or, you know, but it’s very soon actually if you look at the government side, they’ve already acknowledged that this is going to be a problem. Right? So 2027 onwards, all new infrastructure should be post quantum resilient. So that’s actually CNSA 2.0 regulation around. All the systems need to be like, you know, the new systems need to be CNSA 2.0 compliant. Then there’s Google and big companies actually like you know, going out there and telling everybody that your data that is traveling Today, actually in 2029, your data that travels Internet is actually going to be broken. Okay, but 2029 is not that far away. It’s only three years away. But backtrack, if you have data flowing today, that is also at risk.

Ramana Kompella [00:30:15]:
If your longevity of the data is more than three years. Let’s say you have a tax record, Tax records have a seven year, whatever statutory limitation, they usually don’t need to keep records of more than seven years, but three years is not. Those are still relevant and important. So which means basically all that information, right, three years down the line is going to be potentially visible. I mean again this is a prediction, so take it with a pinch of salt. But my point is basically that if you are running any kind of sensitive information, you’re transferring any kind of sensitive information on the Internet today. Harvest now, decrypt later is a big, big threat. Time to migrate to quantum safe infrastructure is today.

KB [00:31:01]:
Okay, so Ramana, final question. Given everything you said, what do you think happens now? What do you think moving forward the rest of 2026? I know you shared a couple of years beyond, but curious to see in

Ramana Kompella [00:31:14]:
the next six months, I mean from Cisco’s point of view, I mean what we are doing is we have unveiled the Cisco quantum switch. Our goal now is to actually take all the pieces. Like we have announced the entanglement source last year. We have a switch that we announced recently. We have some software stack that we have been brewing entanglement swapping algorithms and stuff like that. So all these pieces now need to come together in the form of proper networking solution. And the moment we actually started announcing and stuff like that, we have had tremendous amount of inbound demand around, hey, when can we get our hands on this technology? Right? So we are actually working with some providers to actually figure out a way to get our solutions in hands of them, deploy them, understand the operational challenges. Right.

Ramana Kompella [00:32:05]:
It will take time to soak and mature the technologies, but we are frantically working towards it. So in the next six to nine to 12 months, we will be working hard on getting these solutions at least deployed in a restricted setting, but, but in a provider environment. That’s where we want to, like, really go towards.

KB [00:32:28]:
And there you have it. This is KB on the go. Stay tuned for more.