Shielded: The Last Line of Cyber Defense
Your Hardware Refresh Is a Quantum Deadline with Mamta Gupta
May 22, 2025
You can’t patch your way out of quantum exposure, especially when your hardware refresh cycle is a cryptographic deadline. In this episode of Shielded, Johannes Lintzen sits down with Mamta Gupta of Lattice Semiconductor to explore how hardware-rooted security must evolve to meet the post-quantum challenge. Mamta unpacks the real-world threat of “Harvest Now, Decrypt Later,” explains why crypto agility is essential for long-lifecycle devices, and shares clear, actionable steps to build quantum resilience from silicon to software. Learn how FPGAs provide a reprogrammable, future-ready foundation, what CNSA 2.0 mandates, and why your next hardware refresh cycle may be your last chance to stay ahead. If your business touches critical infrastructure, long-life devices, or digital supply chains, this episode is your wake-up call.
Quantum threats are no longer theoretical, and waiting could cost you. In this episode of Shielded: The Last Line of Cyber Defense, Johannes Lintzen is joined by Mamta Gupta, a quantum physicist turned cybersecurity strategist at Lattice Semiconductor. Together, they break down how crypto agility, hardware-based trust anchors, and PQC readiness are now essential for any organization managing long-lifecycle systems. You’ll come away with practical steps to assess, upgrade, and future-proof your cryptographic infrastructure.

What You'll Learn:


Mamta Gupta is the Senior Director of Strategic Business Development for Security, Telecommunications, and Data Centers at Lattice Semiconductor. With a master’s degree in quantum physics, specializing in superfluidity, superconductivity, and critical phenomena, she brings a rare blend of theoretical insight and real-world engineering to the cybersecurity space. She has led key PQC strategy efforts at Lattice, helping develop quantum-resistant FPGA-based architectures and secure supply chain practices. A leading advocate for confronting the “Harvest Now, Decrypt Later” threat, Mamta is known for her pragmatic, cross-functional approach to building crypto-agile infrastructure.

Your Roadmap to Post-Quantum Readiness:

[00:06:05] Step 1: Understand the Real Threat: “Harvest Now, Decrypt Later”
The quantum threat is already operational, not hypothetical. Adversaries are actively collecting encrypted data today with the goal of decrypting it once quantum computers reach maturity. This tactic, known as "Harvest Now, Decrypt Later," means data with a long shelf life, such as health records, financial data, and national infrastructure, is already at risk. Waiting for quantum computing to become mainstream before acting is a critical miscalculation. Organizations must treat post-quantum cryptography as a present-day risk mitigation priority, not a future optimization. Key Question: Are you taking immediate steps to protect long-lived data from eventual quantum decryption?

[10:43] Step 2: Design for Crypto Agility Across Hardware and Software
The pace of cryptographic change is rapid; regulatory mandates, algorithm approvals, and threat intelligence evolve constantly. At the same time, the life cycle of deployed hardware can stretch over a decade. This disconnect demands systems that are flexible by design. Crypto agility ensures that organizations can upgrade algorithms, rotate keys, and adapt trust models without re-architecting infrastructure or replacing physical components. Without agility, today’s protections could become tomorrow’s liabilities. Key Question: Is your infrastructure architected to support cryptographic evolution across its full lifecycle?

[14:38] Step 3: Run a Three-Part PQC Readiness Assessment
Before launching a migration, organizations need full visibility into their current cryptographic environment. This begins with a cryptographic inventory, reviewing all firmware, certificates, keys, and the algorithms used to protect them. Next, evaluate your trust anchors, components like secure boot mechanisms, TPMs, or silicon-based keys to identify weak points or dependencies on outdated cryptographic methods like RSA or ECC. Finally, conduct a vendor readiness check across your digital supply chain. Determine whether suppliers of silicon, firmware, or software are equipped to support PQC or if they introduce downstream risk. This structured assessment turns uncertainty into a prioritized roadmap for upgrading your infrastructure with confidence. Key Question: Have you completed a full assessment of your cryptographic assets, trust anchors, and vendor readiness?

[17:21] Step 4: Build a Crypto-Agile Root of Trust Using FPGAs
The root of trust is the foundational layer of system security, and its resilience is essential in a post-quantum world. Field Programmable Gate Arrays (FPGAs) offer a key advantage over ASICs: reprogrammability. With FPGAs, cryptographic algorithms can be updated after deployment, providing the agility needed to respond to evolving standards and threats. They also enable secure firmware validation, hybrid cryptography, and alignment with upcoming compliance deadlines without requiring hardware swaps. Deploying FPGAs as the root of trust creates a flexible security architecture that supports long-term cryptographic adaptability. Key Question: Does your hardware architecture support post-deployment cryptographic updates without physical replacement?

[25:33] Step 5: Architect Hybrid Cryptography with Strategic Intent
Transitioning to post-quantum cryptography doesn’t mean abandoning classical algorithms immediately. Hybrid cryptography, running both classical and quantum-safe algorithms in parallel, offers a way to maintain current protections while building future resilience. However, this approach adds complexity in key management, execution order, and performance optimization. Systems must be designed to support multiple key types and enforce clearly defined policies on trust precedence and key retirement. A well-architected hybrid model ensures security and agility without introducing operational friction. Key Question: Have you developed a hybrid cryptography strategy that balances performance, policy, and long-term resilience?

[31:13] Step 6: Launch a Cross-Functional PQC Pilot
A pilot is the fastest way to move from theory to execution. By testing PQC readiness in a contained environment, organizations can surface critical constraints, such as firmware signing limitations, key size restrictions, or vendor gaps. Pilots also help unify teams across engineering, compliance, and leadership, creating a shared understanding of what the PQC transition entails. Running a pilot allows for experimentation, measurement, and iteration before committing to large-scale deployment. It’s a low-risk, high-leverage way to build momentum and organizational buy-in. Key Question: What would it take to launch a practical PQC pilot inside your organization in the next 90 days?

[36:56] Step 7: Build a 12-Month Action Plan Aligned to Regulatory Timelines
PQC migration is not a weekend upgrade; it’s a multi-year journey that requires strategic pacing. CNSA 2.0 mandates post-quantum protections in all new systems by 2030, and enforcement has already started impacting procurement. ENISA, the EU cybersecurity agency, has also issued guidance with specific requirements for hybrid cryptography. Organizations must build an actionable 12-month plan that includes cryptographic audits, vendor engagement, proof-of-concept evaluations, and measurable internal milestones. Aligning to regulatory timelines now ensures you don’t lose compliance or customers later. Key Question: How are you structuring your next 12 months to show measurable PQC progress aligned with global regulations?

Episode Resources:


Want exclusive insights on post-quantum security? Stay ahead of the curve—subscribe to Shielded: The Last Line of Cyber Defense on Apple Podcasts, Spotify, and YouTube Podcasts.

✔ Get insider knowledge from leading cybersecurity experts.
✔ Learn practical steps to future-proof your organization.
✔ Stay updated on regulatory changes and industry trends.

Need help subscribing? Click here for step-by-step instructions.


Shielded: The Last Line of Cyber Defense is handcrafted by our friends over at: fame.so

Transcript

Mamta Gupta 0:00 I would always say, please please please go and take a look at cnsa2.0 guideline. It is going to affect you, and this threat is real, and you need to protect your data. So that's what got me started. And this Harvest Now Decrippleator is the engine driving the whole PQC train right now. The regulation, the implementation, and the kind of solutions that are being done, the whole thing is driven by these threats.

Johannes Lintzen 0:31 Welcome to Shielded, the last line of cyber defense by PQShield, the podcast where we are moving the conversation around post quantum cryptography from the why to the how. I'm Joe Lindzen, your host, and I'm here to guide you through the challenges, innovations, and solutions that will help your organization stay secure in the quantum era. Let's future proof your defenses together. Today on Shielded, the last line of cyber defense, we're joined by a strategic leader in the cybersecurity world with deep expertise in both hardware and software solutions for the post quantum era. Mamta Gupta, senior director of strategic business development for security telecommunications and data centers at Lattice Semiconductor. Lattice is a powerhouse in the semiconductor world, providing innovations that power everything from telecommunications to data centers. As a seasoned expert in this space, Mamta has been at the forefront of ensuring that Lattice's hardware adapts to the quantum revolution, playing a key role in developing quantum resistant technologies for their field programmable gate arrays, FPGAs. In today's conversations, we'll talk about how the semiconductor industry is tackling the transition to quantum ready security, how digital supply chains need to evolve, and why it's crucial for businesses to start preparing for the quantum shift today. Mamta, thank you so much for joining us on Shielded, and welcome to this podcast.

Mamta Gupta 2:03 Thank you so much. It is great to be here.

Johannes Lintzen 2:05 So, of course, it's not often the case, but it is raining in California today. So we are just chatting about well, of course, it creates havoc on the road because no one knows how to drive when it's wet, But it's also really good for our fruit trees. So, Mamta, I hear you have some fruit trees in your yard. What are you growing?

Mamta Gupta 2:26 Yeah. Gardening is my passion. That is my detox. That is everything I do apart from work is really all geared towards gardening. So, yes, I have quite a few fruit trees in my backyard, and the rain is really helpful. I have one particularly prolific loquat tree. I don't know if you know that fruit. So it's, like, laden, and squirrels are having fun. We are having fun. And then I have these soccer ball sized lemons. They're called ponderosa lemons or Turkish lemons. That tree is also growing. I'm sorry. I don't know. I have 500 lemons, soccer ball size hanging on that tree. So, yeah, a lot of produce. I am really enjoying this fruit season this time. re having fun. We are having fun. And then I have these soccer ball sized lemons. They're called ponderosa lemons or Turkish lemons. That tree is also growing. I'm sorry. I don't know. I have 500 lemons, soccer ball size hanging on that tree. So, yeah, a lot of produce. I am really enjoying this fruit season this time.

Johannes Lintzen 3:14 Well, now I'm jealous because lemon trees are those are my favorite. And first, it was an orange tree that we planted, beautiful orange tree, but the gophers got into it. So, like, there's always that threat in California soil as well. You get the gophers. You get the cartoonish, the plant starts to wiggle and it disappears. So,

Mamta Gupta 3:35 yeah, that is so true.

Johannes Lintzen 3:37 But, of course, we're not only here today to talk about plants and the weather in California. That would be a pretty boring podcast, I think. Instead, we're here to talk about, of course, the migration towards post quantum and the threats that come with it. So maybe to start with, Mamta, maybe you can tell us a little bit how you got started in the world of both hardware security, but also quantum technologies.

Mamta Gupta 3:59 It's actually funny. My background is in quantum physics. Lot of people just take and double take. What do you mean by that? I'm like, I did my master's there. So I have a master's in quantum physics, super fluidity, superconductivity, and critical phenomena. And that directly ties into the work of quantum computers. Superconducting materials are what are required in quantum computers. So starting with that kind of quantum math, physics, and then my first foray into technical landscape, I was involved in a lot of aerospace and defense and high reliability projects, and that's their security and reliability is very important. And slowly, I got exposed to cryptography. And, of course, because of my background in superconductivity, I always kept an eye on what was happening in the quantum computer space. And from there, platform security threats and the solutions, they were so important in my past experience that I just kept on building on that and learning about it and grew my expertise. And then I started leading the strategic marketing, uh, and strategic business development for security, data center, and telecommunication. And all these three segments are extremely aware of what is happening in the quantum computer space and how it is going to threaten our current security protections.

Johannes Lintzen 5:38 And, of course, one of the big threats that is being discussed a lot is the harvest now decrypt later. And as you shared with me in preparing for this podcast, you did mention that that certainly also was and still is one of the main driving factors for the journey that you're taking towards, uh, quantum safe solutions with Lettuce. So maybe you can elaborate a little bit on the HarvestNOW decrypt later and why it plays such a crucial role in your approach. for this podcast, you did mention that that certainly also was and still is one of the main driving factors for the journey that you're taking towards, uh, quantum safe solutions with Lettuce. So maybe you can elaborate a little bit on the HarvestNOW decrypt later and why it plays such a crucial role in your approach.

Mamta Gupta 6:06 Absolutely. As I said, I always had an eye on where quantum computing was going. And it always is like, okay. It'll come somewhere towards middle of the century. That's where the whole thing was. That's somewhere around twenty, thirty five, 40 quantum computers will come. So keep an eye. It's interesting. But then I was in a conference, and there was this presentation from NIST and CNSA. And they were talking about this attack. They called it steal now, decrypt later. And they said, look. We are getting very concerned that, you know, certain actors and especially some nation state backed actors are actually harvesting data. And the data that is going to be sensitive ten years, twenty years, thirty years from now, and they just sit on it. And once quantum computers come online, they plan to break the encryption and actually steal that data. And that was the moment for me that, oh, this is not about how fast quantum computers can get on my desktop if they can ever in my lifetime, but it is about protecting data now so that it is safe for future and certain kind of data. And a lot of people said, oh, well, thirty years, who cares? No. Twenty years, ten years, who cares? But immediately, you think about it, it's about your Social Security number, your health records, your financial records, your retirement fund. It's all relevant. Ten years from now, twenty years from now, thirty years from now, depending on your age, it's all relevant. So if that data is going to be compromised, that needs to be protected today. And that is what is driving the urgency. That is what is driving the regulation. And that's what woke me up, and I started advocating in our company that, hey. We need to have a very aggressive post quantum crypto strategy and rollout so that we can equip our customers with these solutions. And I also started evangelizing almost that it was almost like a customer advisory. Wherever I would present or I would meet with customers, I would always say, please, please, please, go and take a look at cnsa2.org pipeline. It is going to affect you, and this threat is real, and you need to protect your data. So that's what got me started. And you're right. This harvest not a trip later is the engine driving the whole PQC train right now. The regulation, the implementation, and the kind of solutions that are being done, the whole thing is driven by these threats. , and this threat is real, and you need to protect your data. So that's what got me started. And you're right. This harvest not a trip later is the engine driving the whole PQC train right now. The regulation, the implementation, and the kind of solutions that are being done, the whole thing is driven by these threats.

Johannes Lintzen 8:54 So a clear focus on that threat being very imminent already today, um, obviously, with nation states already starting to collect that data, waiting for that moment when they have a cryptographically relevant quantum computer available. So, obviously, getting ahead of that time line. It I think that was the other part of CNSA two point o to also communicate that time line, to acknowledge that a migration over to Quantum Save is gonna take its time. It's not something that's gonna happen overnight.

Mamta Gupta 9:25 Actually, you've heard of a good point. This change, even if you say, okay. I'm completely woken up to this threat, and I want to make a change. It will take time. This is not a simple change. The whole industry to protect its data trove, it will take years. Last time we put that little lock icon on our websites, it took us ten years to get there, that we had it on majority of our websites. So it's the same kind of timeline it'll take for and now we have, like, multiple times of data that we have to protect than what we had in 02/2008 or 02/2006 when we started this whole security h t t p s movement, and it took ten years to do that. So, yes, please keep in mind, it is going to be a long haul, so it's better to start early.

Johannes Lintzen 10:12 Right. Better to get ahead of it. And, of course, you point the term to say, the question is not if, it's when such a cryptographically relevant quantum computer will become available. And we are quickly moving closer to that reality. So from where you sit in the supply chain as a semiconductor company and hardware industry enabler, How do you see the challenges in adopting PQC, and what are the implications for businesses that might be your customers going forward? um computer will become available. And we are quickly moving closer to that reality. So from where you sit in the supply chain as a semiconductor company and hardware industry enabler, How do you see the challenges in adopting PQC, and what are the implications for businesses that might be your customers going forward?

Mamta Gupta 10:44 Very right, Sue. And the biggest challenge that I see is the disconnect between the speed of cryptographic evolution, which I just mentioned, is very slow because the data is so much that we need to protect. And also the life cycle of hardware, that is also very long. And then you have this threat that is evolving very fast. So implementation time lines, life cycle of hardware are on a completely different time line or a different scale, years and tens of years kind of scale. While the threat evolution and the requirement evolution is happening literally every two quarters. Every two quarters, either a new update comes on the threat or a new regulation comes or a new algorithm is identified or one leading algorithm is disqualified. So there is so much churn happening. So this dichotomy between or this disconnect between what needs to be done and the time it takes to do it versus how the, uh, requirements are changing and the threat is changing is the biggest challenge, and people are still not comprehending it. So it gives you a twofer. Right? So first of all, if your device is going to be in market for next ten years and you are subject to CNSA two dot o requirements, you should have started yesterday because that device should have some PQC capability. Because by 2030, every device fielded a subject to CNSA two dot o mandate must have PQC protection. So that is one. That first, you've got to start very quickly if your device life cycles are longer. And then the other thing is that because it is evolving so fast, the threat is evolving so fast, the requirements are evolving so fast, you cannot afford to lock in something rigid. And that's where crypto agility comes in. That, okay, I will put in a PQC requirement, but if you lock it in an and that algorithm gets evolved or broken and a new version comes out or a new kind of threat comes out, then you're locked in with a useless protection. So this is where the traditional ways of doing things have to be rethought, and you have to think about crypto agility. Lot of people find agility and security, but this is what the need of the hour is.

Johannes Lintzen 13:22 They need to merge. They need to come together. re locked in with a useless protection. So this is where the traditional ways of doing things have to be rethought, and you have to think about crypto agility. Lot of people find agility and security, but this is what the need of the hour is. They need to merge. They need to come together.

Mamta Gupta 13:24 Exactly. So that means something. You have to put in protection in something that will allow you to upgrade it later. And then I would also mention one more thing that the requirements and algorithms, their compliance and certification framework has not been decided. So where is this need for agility coming in? Why are we so worried about that these algorithms will be broken? It's because their compliance and certification frameworks are still evolving. They have not been released. So now you have algorithms that have been released that you are now mandated to follow, but you don't know whether they will pass their own certification or compliance or not. During that framework development, we might find vulnerabilities. So there is a widespread recognition in industry that we will see some evolution, hence, the need for crypto exhibits.

Johannes Lintzen 14:17 So if we go a little bit up the stack, obviously, we've been talking about the supply chain a little bit. So if I put myself into the shoes of a financial organization now or maybe a provider of health care solutions, what can I do today to get prepared for the quantum threat? Like, how do I engage in making sure that those steps are being taken? the stack, obviously, we've been talking about the supply chain a little bit. So if I put myself into the shoes of a financial organization now or maybe a provider of health care solutions, what can I do today to get prepared for the quantum threat? Like, how do I engage in making sure that those steps are being taken?

Mamta Gupta 14:39 So number one is visibility. Look at what you got and understand what's protecting your data and devices today. That means look at the firmware. This is the low level code that will run on your device before your even OS comes out, and see whether it is cryptographically signed, what algorithms are being used. And then look for in security industry, we call these things trust anchors. These are foundational components like your keys that you use for signing or authenticating and the certificates that you use to establish trust in the system that, yes, this is the right device. This is the right firmware version. So you have certificates for those. So you have to see how are these certificates and keys being generated? What algorithms are you using? If you're using RSA and ECC, they're going to be broken. And CNSA2.0 has decommissioned them. So it's not like, oh, I'm at RSA two k today. I'll move to four k, so that'll give me more protection. Or I'm at ECC two fifty six. I'll move to 384 or higher. No. They are completely, fundamentally broken. I call that bullet to a paper. The quantum computer bullet to these ECCRSA sheets of paper. Whether you put two sheets of paper in front of a bullet or five sheets of paper, it's not gonna protect you. So we have to see so visibility, assessment of your trust anchors, and then your vendor readiness. You must have a supply chain. Right? You're using vendors. Are they ready for PQC enablement so that when you use their components and they are in the path of your either establishing trust or are they supplying firmware or critical data protections for you? Are they enabled? Are their products enabled? So these are the few things you've got to look at. How is your firmware being protected? How are your trust antennas being established? Which algorithms are you using? And are your vendors ready?

Johannes Lintzen 16:53 So in that entire digital supply chain, of course, we're talking from a more of a bottom up approach where we start with the silicon level and then protect everything up the stack all the way to the application layer. And obviously, Lettuce, your products are being used globally. Can you give an example on what such a migration project looks like? How you engage with your partners and customers up the stack? And what is required to update ultimately the application? otect everything up the stack all the way to the application layer. And obviously, Lettuce, your products are being used globally. Can you give an example on what such a migration project looks like? How you engage with your partners and customers up the stack? And what is required to update ultimately the application?

Mamta Gupta 17:22 Yeah. So Lattice provides hardware security. We are often the root of trust for our system. So think of it like building a skyscraper. So you need to start with a strong foundation. And in security, that foundation is normally referred to as a root of trust. And in most critical systems, the root of trust is anchored in silicon. So it starts from a piece of silicon or a hardware, could be a hardware security module or a root of trust silicon device. And then you create a chain of trust on top of that. And subsequent firmware and software, you bubble up the trust from there, and you have certificate exchange, and you have attestation. You check each other's credentials, and you establish trust in the whole system. So first of all, that tamper resistant root of trust, that is PQC enabled. That is your starting foundational block in supply chain so that now you can put the firmware layer on top, and that has to be validated using cryptographic signatures. So today, what people mostly use are RSA or ECC. That has to move to CNSA approved algorithms like LMS, XMSS, or MLDSA. And your signing, your authentication has to be done using these new algorithms. And if you use ECDH, which is to create a secure key exchange protocol where you exchange your encryption keys, that has to move from ECDH, which will be completely broken and is the foundation of those harvest now, decrepit attacks, you have to move to MLKEM. So this is the new algorithm that has been approved by cnsa2.org. So you have to upgrade your root of trust. You have to upgrade the methods of signing your firmware, and then your software stack must also be upgraded with PQC algorithms. And in every stack or every layer of your supply chain, you got to build in crypto agility. And if you create a chain of trust, then if your root of trust device has crypto agility, then you can bubble it up to all the layers. So what we recommend is that have your root of trust device be crypto agile so that it can always be upgraded with the latest, greatest algorithms, and then it can transfer the trust up the chain. Then you can have as long as your firmware and software stacks have certain amount of receptiveness to that crypto agility, you are good. But your root of trust has to be crypto agile. So then you can build layers on top. be upgraded with the latest, greatest algorithms, and then it can transfer the trust up the chain. Then you can have as long as your firmware and software stacks have certain amount of receptiveness to that crypto agility, you are good. But your root of trust has to be crypto agile. So then you can build layers on top.

Johannes Lintzen 20:25 So, of course, with discussing cryptographic agility, a lot of changes, updates in firmware software are gonna be needed throughout the stack. And, of course, it's relatively simple if we talk about software. But if we talk about hardware, it becomes a little bit more of a challenge. Earlier on, you used the term ASIC, and we've been talking about FPGAs a little bit. So maybe you can elaborate a little bit on that for our listeners who might not be so familiar with those terms and where the benefit of one lies above the other. if we talk about hardware, it becomes a little bit more of a challenge. Earlier on, you used the term ASIC, and we've been talking about FPGAs a little bit. So maybe you can elaborate a little bit on that for our listeners who might not be so familiar with those terms and where the benefit of one lies above the other.

Mamta Gupta 20:56 Absolutely. So we normally deal with two kinds of silicon devices. One are application specific, ASICs, application specific integrated circuits. And they do a certain set of functions, and that is hardened in silicon. That means you have taped out and you have actually made the silicon which is hardened. It cannot be changed. That personality is built into the system into that hardware. So they are very function specific. And if the function changes, they have some maneuverability if they have, uh, CPU in there or a SOC in there, which is a processor. And they can bring some maneuverability with that software that is running, but pretty much they're fixed in function. And then you have another class of devices that are reprogrammable. That means they come to you as blank slates. They are FPGAs, field programmable database. That means they are blank slates. They're still piece of silicon, but you program them. You actually create the design in them based on your application. And this is changeable. You can program it if you're happy with your application in it. You can lock it, and then it just acts like any other silicon piece. It could almost act like a application specific piece of silicon and mimic an ASIC. But you have the ability to now reprogram it, and it'll take on a different personality. And this is ideally suited for a crypto agile hardware root of trust. So Lattice builds these FPGAs, which are root of trust, and we have built in crypto agility. So some crypto agility is built in by the nature of FPGAs themselves because they are programmable. So you can put an algorithm in there that will be hardened into the FPGA fabric. So it's not sit in software, and you will have physical isolation of keys. You have physical isolation of code. So that is very secure environment, and that gives you that hardware root of trust. These are some of the tenets of hardware root of trust. Security architects listening in our audience would understand what I'm saying, but you need that kind of code and key separation for a good hardware root of trust. And that's what FPGAs can offer, but they also offer reprogrammability. And this is your insurance against any changes in either requirements or the algorithms. You field an FPGA as a root of trust that has your PQC algorithms built into it. They will be able to be reprogrammed, and you can update your algorithm. So you can upgrade without swapping out the hardware. With ASICs, you have to swap out the hardware. You have to tape out a new device, upgrade it, and then you have to take it out of your board, the old one, and put the new one there. With FPGAs, you just have to upgrade the bit stream. Over the air, you can bring in a new bit stream and reprogram, and the FPGA will now wake up with the new algorithm. So there is another benefit of using FPGAs in such security anchor or root of trust role. You can do hybrid cryptography. So if you have classical today, you can add on post quantum crypto algorithms that can be implemented in your fabric. So FPGAs have a reprogrammable fabric, which is the area where your logic gets implemented. And you can then have both of these algorithms run-in parallel. So hybrid cryptography is critical during transition because today, we are still protecting against classical computers, and they are against classical computers. ECC and RSA are fairly robust. So we need to have what's working today, and then we also need to have additional algorithms for tomorrow. So currently in this transition period for at least next five years, we see that hybrid cryptography is extremely important. And FPGAs can carry that together because they can be reprogrammed. So if you have classical in there, you add PQC, and your FPGA is ready to go. need to have additional algorithms for tomorrow. So currently in this transition period for at least next five years, we see that hybrid cryptography is extremely important. And FPGAs can carry that together because they can be reprogrammed. So if you have classical in there, you add PQC, and your FPGA is ready to go.

Johannes Lintzen 25:19 You can phase one out and phase the other one in or replace entirely. Exactly. Talking about hybrid, are there any you mentioned the benefits which are obvious, But are there any challenges that also come with hybrid technology in this context?

Mamta Gupta 25:35 Indeed. Nothing is free. So if you are going to run two algorithms, of course, they will run one by one. So you could authenticate with one, and then you can authenticate with the other. So sometime, the performance will suffer a little bit. Key lengths. So if you have finite key storage, which is geared towards classical. And now if you want to fit in your post quantum keys, you could get into a challenging situation because key storage is normally finite in FPGAs and ASICs alike. And that's where you will have to carefully architect your hybrid solution to see that you have enough space.

Johannes Lintzen 26:19 Yeah. Of course, with the hybrid there, you know, I think there's room for optimization of within those challenges, be it space required performance. And then, of course, use case specific, choosing that right algorithm.

Mamta Gupta 26:34 Right balance of hybrid and when to switch over or which takes priority. So if you're doing layered key management, can PQC keys retire classical keys? Can classical keys retire PQC keys? It's everything has to be thought through that what do you do in your architecture. So the other thing would be that the device manufacturers, beat FPGAs or ASICs, they also must keep hybrid in mind as we are rolling our devices every day so that you have enough key storage in there and you have enough flexibility in key management that both kinds of keys can be handled.

Johannes Lintzen 27:17 Well and also, because you've been talking a lot about regulations and sort of the global landscape, and some of the regulations that are coming out are putting more emphasis on hybrid than others. We've seen a little back and forth by some of the work that has been issued with the topic of hybrid. ] Johannes Lintzen: Well and also, because you've been talking a lot about regulations and sort of the global landscape, and some of the regulations that are coming out are putting more emphasis on hybrid than others. We've seen a little back and forth by some of the work that has been issued with the topic of hybrid.

Mamta Gupta 27:33 Yes. I would actually say ANISA in Europe, they have actually required now it's required in Europe that if you are using MLDSA in chem, which is a brand new math. It's called lattice based math. It's brand new math. And when I say brand new, it's been there for years, but in relatively brand new math. Because in cryptography, most of the math is, like, 40 or old older. Right? Battle destined. But this is new math, relatively new math. So when you use MLDSA and ML chem, Anees actually requires you to use hybrid because these are the algorithms that are most considered most vulnerable to evolution or changes. If you are using hash based cryptography, which is more stable because it's based on a fairly well known map of hash cryptography, like LMS and XMSS, these are also CNSA two dot o approved algorithms, then the requirement doesn't ask you to do hybrid. But the new ones, MLDSA and CAM, which everybody is racing to adopt, they must keep hybrid also. And if Europe is your focus market, then you must think about hybrid.

Johannes Lintzen 28:50 Now do they see it as a just a migration step, or do they foresee it being a long term solution as well?

Mamta Gupta 28:57 That's a good question. Right now, Aneesa has not put a clear date that, okay, after this, you will switch only to PQC. Like, CNSA is very crystal clear in that. They have a cutoff date, 2030. End of '20 '30, every route of trust or every trust anchor shall move to PQC. There is no hybrid. Actually, CNSA doesn't prefer hybrid. They say no. No hybrid will be used in protecting national secrets. So they are not hedging the bets at all. They are going one shot, just move forward, full close quantum. But EU is taking a more measured approach. And they believe in hash based, but for the new lattice based, they are asking for hybrid. And as far as I know, there is no timeline to end this. That, okay, here is a cutoff date to only PQC.

Johannes Lintzen 29:53 Probably, uh, also different means of enforcing regulations where some are more of a best practice suggestion and others like CNSA two point o. I mean, they have a clear enforcement capability. Just not gonna buy your products anymore after the date. It's quite simple.

Mamta Gupta 30:12 And we are seeing that in the market. We are seeing that in the market. Because when I used to approach customers in '22 that, hey. You got to look out. This mandate is coming, and '25 or '26 will be the horizon when you will be worried about it. And they will say, oh, no. I'm big enough. I'll get exceptions. And the same customers have come back to us because their boxes got rejected. et. Because when I used to approach customers in '22 that, hey. You got to look out. This mandate is coming, and '25 or '26 will be the horizon when you will be worried about it. And they will say, oh, no. I'm big enough. I'll get exceptions. And the same customers have come back to us because their boxes got rejected.

Johannes Lintzen 30:35 Well, and we're in that hardware replacement cycle. Right? The life cycle of products in the field that you did mention earlier. I mean, if I don't move now, when is my next window going to be? If I procrastinate for another five, seven years, likely, I'm gonna miss that train that, you know, has left the station a long time ago, so to speak. Now besides product related challenges, of course, getting an organization ready for such a massive migration is also a huge organizational challenge. Maybe you can share a little bit the steps that you took within Lettuce to get the focus on that migration from an organizational perspective. ong time ago, so to speak. Now besides product related challenges, of course, getting an organization ready for such a massive migration is also a huge organizational challenge. Maybe you can share a little bit the steps that you took within Lettuce to get the focus on that migration from an organizational perspective.

Mamta Gupta 31:14 Yes. That whole presentation about how it's not a crypt later literally was a tipping point for us. I heard that, and very quickly, the management was equally convinced that, look, this is real, and this mandate is real. This is not a guideline. This will be a mandate from cnsa2.o. So we immediately started inventing the protections we offer for firmware and trust anchors. We are a platform root of trust. Lattice FPGAs are root of trust capable, and we normally are positioned as platform root of trust in data center, telecommunication, industrial applications, where we are not only detecting our keys or our bitstream, but we are also establishing trust on the entire board. We can go and authenticate clashes and all that stuff. So wherever we are issuing those certificates or we are authenticating firmware on the board or something, we started inventing those that okay. Where are the use cases? Because every customer has a slightly different use case. They follow different protocols. We started inventing those. And then we also started doing a hybrid pilot program, and it was like a feasibility project. And we said, hey. Do we have enough memory or flash? So all of our root of trust devices come with flash onboard. That means there is flash dedicated to the chip. It's with the chip. So your keys and all are stored in there. So, normally, we have ample space because it is dedicated flash just for security functions. And we started inventuring that. Hey. Do we have enough space for these keys? And this pilot was very revealing because we realized, oh, we could fill it up very quickly because these PQC keys are big, very resource hungry. And the whole certain signing operations could be resource hungry, and they need to be accelerated intelligently. Otherwise, you could hit into these challenges. And then if you do hybrid, then we had to think through all our key management capabilities that we offer. As I mentioned, that who takes precedence? Can classical attack PQC or PQC always takes precedence? And when do you switch over or you treat them as equal? So this pilot revealed a lot of those issues that we had to think through. And this could almost become a template for anybody who's trying to start on this journey. They take an inventory of your firmware and trust anchors, how are they being protected, and take do a hybrid crypto pilot. See where you would sign with which keys, and is your system capable of handling multiple keys? Your system may be capable of handling one or two keys with certain length. But now if you're putting in three or four or five keys in there, is it capable? And then also, we had to then talk to our vendors. We have a very big manufacturing supply chain where we manufacture our devices. We had to put in lasting security protocol. We have a very robust lasting security protocol. We have HSM in our manufacturing, and we had to inventory our vendors. Hey. What is your PQC roadmap, and how are you doing it? And, again, this could be a template for our listeners as well. n where we manufacture our devices. We had to put in lasting security protocol. We have a very robust lasting security protocol. We have HSM in our manufacturing, and we had to inventory our vendors. Hey. What is your PQC roadmap, and how are you doing it? And, again, this could be a template for our listeners as well.

Johannes Lintzen 34:38 So really back to Harvest Now Decrypt Later being the trigger for your product migration. It was also the big trigger, it sounds like, that got your c level attention and immediate triggered all these really collaboration issues and, you know, projects work within the organization. Because I'm sure that the product team and technology and software development, hardware design, I mean, there's a lot of collaboration that needs to happen to even get to a point where you can run these POCs. I think it's much easier said than done given that

Mamta Gupta 35:10 Indeed. See, one thing I'll say, house now, the crypt later. Once you focus people's attention to it, it has the lure of almost a horror movie plot.

Johannes Lintzen 35:21 Nice. Yeah.

Mamta Gupta 35:22 It actually pretty easily catches attention and makes sense to most people. But as you said, it's easier said than done.

Johannes Lintzen 35:29 So you think it won't be long until we see the next Hollywood movie in

Mamta Gupta 35:34 I said it here. Sales gets it. Investors get it. Everybody gets it, that this risk is real. And you just have to focus on it, but doing the work is a lot of work. And the systems are used to doing security in one way. There's a lot of dependence on certain silicon pieces, you know, ASICs, TPMs, trusted background logic. And all of those TPMs rely on traditional cryptography like ECC RSA, and that will be broken. So every TPM on this planet has to be upgraded or replaced. So lot of people have that Oh, my trust anchor is actually a TPM. What do I do now? And this is where we sit down with them, and sometimes they have to re architect their security strategy. Or sometimes they start looking at a hybrid that might create a PQC as a entry gate protection, and then I transfer trust internally where my system is now protected. And internal things can be done through classical. So that's that. So for

Johannes Lintzen 36:44 those organizations that are maybe a little bit behind or just waking up to the threat, what's your recommendation for them? Like, what should they be focusing on for, let's say, the next twelve months to really get started for good on this migration journey? ssical. So that's that. So for those organizations that are maybe a little bit behind or just waking up to the threat, what's your recommendation for them? Like, what should they be focusing on for, let's say, the next twelve months to really get started for good on this migration journey?

Mamta Gupta 36:58 So the first thing would be get visibility into your systems, into your use cases, into your customer needs. That if you're a supplier, then look at what you are supplying and how you're protecting it in a customer system. And if you're a customer, look at how your vendors are protecting their deliverables to you. So get visibility about your firmware protection, your trust anchors. Start a pilot. As soon as you get started on it, start looking at it. Start implementing something. There are very many open source versions available. Most of the vendors' IP security, I've even just like yourself. You have those evaluation phases that you do with your IP so people can engage with you and get evaluation licenses. Start evaluating this in your system and start looking at what upgrades would be required and start mapping it out. It's gonna take some time, effort, and money to upgrade your system so that they can be B2C compatible.

Johannes Lintzen 38:03 Excellent. Yeah. All of those are very good hints and tips and tricks. Now before we wrap up, and thanks, Mamta. This has been a fascinating discussion. I certainly learned a lot. Where can listeners follow your work? Like, where would you like to point people to?

Mamta Gupta 38:18 So definitely visit us at latticesemi.com. We have a security page. You will see a lot of informational material there, blogs, white papers. And then I post regularly on LinkedIn, especially around quantum readiness, FPGA cyber resilience, and secure infrastructure. So follow me like me. I do a lot of security seminars, webinars, and they're posted on our latest semiconductor YouTube channel. You can find us there as well.

Johannes Lintzen 38:49 Great. Well, thank you so much, Mamta, for sharing your insights. I know this has been very valuable for many of our listeners. So thank you.

Mamta Gupta 38:56 Thank you for having me.

Johannes Lintzen 38:58 Appreciate it.

Mamta Gupta 38:59 It was a pleasure.

Johannes Lintzen 39:00 A quantum error is already here. With the right tools, the right knowledge, and the right mindset, together, we can face it head on. To find out more about PQ Shield and how our cutting edge solutions can help you secure your data against the quantum threat, Visit pqshield.com. Make sure to search for Shielded, the last line of cyber defense in Apple podcasts, Spotify, or anywhere else you get your podcasts. Don't forget to click subscribe so you never miss an episode. On behalf of the team here at PQShield, thanks for listening.