Everspin at Lytham Partners Summit: MRAM’s Strategic Future

On Tuesday, 19 August 2025, Everspin Technologies Inc. (NASDAQ:MRAM) participated in the Lytham Partners Consumer & Technology Summit. During the event, CEO Sanjeev Agrawal emphasized Everspin’s MRAM technology as a key solution for various applications, while also addressing supply chain resilience and competitive positioning. The company reported strong Q2 results, showcasing a promising outlook amid challenges.

Key Takeaways

• Everspin reported Q2 revenue of $13.2 million, surpassing guidance and analyst expectations.
• The company maintains a robust balance sheet with $45 million in cash and no debt.
• Everspin targets a $4.3 billion total addressable market (TAM) by 2029, focusing on AI and LEO satellites.
• New leadership appointments aim to drive business development and sales strategies.
• Everspin’s diversified supply chain mitigates potential tariff impacts.

Financial Results

• Q2 revenue reached $13.2 million, exceeding the midpoint of guidance and analyst consensus of $13 million.
• Non-GAAP EPS was $0.03, indicating profitability.
• The company boasts a strong balance sheet with $45 million in cash, remaining debt-free.
• Profit margins are on the rise, reflecting efficient operations.

Operational Updates

• Everspin’s STT-MRAM targets NOR Flash replacement in high-density applications.
• The XSPI family is seeing significant design wins, with revenue generation expected in 2025.
• Development of 1GB parts is underway, with the first high-density XSPI part launching in 2026.
• MRAM solutions are being developed for LEO satellite applications.
• Focus on MRAM applications in AI inferencing at the edge to enhance performance.

Future Outlook

• Everspin aims to capture a $4.3 billion TAM by 2029, focusing on data centers, data logging, and configuration memory markets.
• The company is capitalizing on the growing LEO satellite market with radiation-tolerant MRAM.
• MRAM’s advantages in AI inferencing at the edge are being leveraged for competitive positioning.
• The market anticipates 70,000 satellites in the next five years, presenting significant opportunities.

Q&A Highlights

• Everspin’s business model includes product sales, licensing and royalty agreements, and US government contracts.
• 90% of business is conducted through a global network of distributors.
• The company serves over 2,000 customers worldwide, including Fortune 500 companies.
• Competitors like NetSol and Avalanche Semiconductor focus on different market segments, with Everspin emphasizing its broader product range and domestic manufacturing capabilities.

Readers are encouraged to refer to the full transcript for a detailed account of Everspin’s participation in the Lytham Partners Consumer & Technology Summit.

Full transcript - Lytham Partners Consumer & Technology Summit:

Robert Blum, Managing Partner, Lithium Partners: Hello, everyone, and welcome to the Everspin Fireside Chat. My name is Robert Blum, managing partner at Lithium Partners. And today, I’ll be moderating a q and a discussion with Sanjeev Agrawal, the company’s chief executive officer. As a reminder, Everspin trades on the Nasdaq under the ticker symbol MRAM.

Alright. Let’s get started. Sanjeev, welcome.

Sanjeev Agrawal, Chief Executive Officer, Everspin: Thank you, Robert. Pleasure to be here.

Robert Blum, Managing Partner, Lithium Partners: All right. For those new to the company, provide an overview on sort of your mission and really how that sort of evolved to address this growing demand for advanced memory solutions.

Sanjeev Agrawal, Chief Executive Officer, Everspin: Yeah. So Everspin is a semiconductor manufacturing company. We manufacture magnetic based memories. And the claim to fame over here is that we can actually tune our technology to look like different kind of memories. It can be a data logging memory.

It can be a configuration memory, or it can be a data center memory. So we have basically shipped about two two hundred million plus units since our inception in 2008 to about 2,000 plus customers worldwide. And most of our applications are where, you know, there is a mission critical data involved, whether security is a concern, where data logging is important, and those are the applications that we have actually thrived on.

Robert Blum, Managing Partner, Lithium Partners: Well, what was sort of that that moment that, you know, the milestone of sorts in in this journey that you’ve been on here that have really shaped the position as as this leader now in in MRM technology?

Sanjeev Agrawal, Chief Executive Officer, Everspin: Yeah. So when we were developing the technology, we realized that there is no other technology that can actually read and write as fast as MRAM. So basically, just to give you an example, nanoseconds read and write versus a microsecond write with NOR flash memory. Okay? So we realize that that is a very critical parameter for many of the applications such as programmable logic computers in automation industry.

The idea is that the robots are actually manufacturing on the floor or manufacturing different things. And when they need to write back to the PLC, constantly informing the PLC as to what they’re doing. So if there is a power loss in this situation and you are using a traditional memory, like a NOR flash, all the WIP on the floor would actually be scrapped. But with our memory, you can actually record with the nanoseconds what the robot was doing. So in that sense, the power is lost, we don’t actually have to scrap the whip on the floor.

When the power comes on, the robots can start at that very moment or at that very location or that very action that they were doing. So the bottom line is very fast data logging. The other attribute is automotive temperatures, minus 40 to 125 degrees C. So very robust, very reliable, and fast data logging. When we realized these three things, we basically understood that it has a wide range of applications.

It has a application wherever you’re thinking of data logging. For example, battery management system and EVs, PLCs like I just said in automation, casino gaming or in any medical equipment, for example. So that’s where we actually took off and started off with our first product targeting the data logging industry.

Robert Blum, Managing Partner, Lithium Partners: You touched on a couple there, but maybe walk specifically through product offerings, Everspin’s product offerings and then sort of how they integrate into some of the customer systems that you just mentioned here.

Sanjeev Agrawal, Chief Executive Officer, Everspin: Yeah, sure. So if you look at Everspin’s data product portfolio, we actually categorize them into three different categories. The first one is persist, which basically stands for performance memories for systems. The second one is UNESYS, which is unified data and code for systems. And the last one is Agilus, which is basically agile memories.

And that’s basically looking into forward looking applications like AI and as such. So we started off talking about persist, which is basically where you’re looking for very robust memories at extreme temperatures, minus 40 to 125, for example, or even for aerospace applications. And the idea is you can record very quickly as to what’s going on. So like the battery management system that I was talking about. Unisys is where you’re actually trying to combine data and code.

So what’s the idea over there is if you think about FPGAs, the configuration for an FPGA is actually stored in a memory which is off the system. So on power up, the memory, basically the configuration is downloaded from the memory into the elements in the FPGA and then the FPGA executes or the logic elements execute the action. Now this NOR flash memory takes a microsecond to write. So we’re trying to download a 64 megabit memory, for example, it’s going to take several minutes. But if you were to replace that with an MRAM, it can only take nanoseconds.

So for example, if you’re familiar with Tesla’s, what happened in early, I believe in 2024, when they were going through a configuration, all these Teslas were actually parked on the side of the road waiting for this configuration to go through. And many a times it failed as well when it had to be rebooted. If you were using an MRAM instead of that NOR flash memory, it would have been done by the time you hit the power button to actually start the card. And if it failed, all you had to do was start it up again, and by the time the configuration would have updated. So that’s the power of MRAM, very fast data write and reads, very fast data logging and robust temperatures.

Robert Blum, Managing Partner, Lithium Partners: All right, very good. I wanna sort of jump into the business model here a little bit. Describe what the business model is. You’ve got revenue sort of from product sales, licensing partnerships. Sort of provide a little bit of an overview on each of those, if you could.

Sanjeev Agrawal, Chief Executive Officer, Everspin: Yeah, sure. So our business model basically has three different revenue streams. One is product sales. The second one is licensing and royalty. And the third one is, and I basically put it out as separate because it’s a US government and contracts.

So we are a memory that is mission critical for various US government applications. Okay? So there is a revenue stream from there where somebody can come and license our technology. So you get licensing revenue from there. And then there is a foundry model.

So on-site technology is being qualified and the same contractor or the prime would actually use us as a foundry to build those parts. And then when the part ships out the door, then you get a royalty based on the part that uses Everspin’s technology. That’s one. The second one is basically product sales. So we have three different generations of memory over here.

One is the field switched or toggle MRAM. And then we have the perpendicular MRAM or spin transfer doc MRAM that we are actually shipping parts today into the data center, as well as into the configuration memory solutions. So product sales, licensing revenue and foundry revenue, those are the three categories that we recognize revenue from.

Robert Blum, Managing Partner, Lithium Partners: Let’s talk a little bit more about the customer base, right? Sort of how do you balance large companies, smaller customers, specialized sort of situations, more broad situations? Talk about the customer base in general, if you could.

Sanjeev Agrawal, Chief Executive Officer, Everspin: Sure. And so, you know, we actually have a very diversified customer base. We have 2,000 plus customers worldwide, and these include customers like Siemens, Schneider, IBM, Juniper Networks, you know, the big Fortune 500 companies as well. So the idea is that we serve most of these customers through our distributors worldwide. Okay?

So the idea would be, for example, let’s say we have a large customer like IBM. In the early stages, we would actually work with the customer to help them understand our part, work out any issues that are there. And then basically we hand over, and this is the customer’s choice, we hand over the sales or the supply chain through a distributor. So 90% of our business actually is through distributors worldwide. We have distributors in Asia Pacific, in Europe, as well as in The Americas.

Robert Blum, Managing Partner, Lithium Partners: When you sort of put them into different verticals here, right? You talked about some of the medical applications, some of the automotive applications. What are the areas of focus that you’d maybe see the greatest growth opportunity from going forward here?

Sanjeev Agrawal, Chief Executive Officer, Everspin: Yeah. So to date, most of our revenue comes from our Persist family, which is basically industrial automation, the PLCs that I was talking about with the Snyder’s and the Siemens of the world, casino gaming. So one thing that I forgot to talk about was today, they’re using this traditional memory NOR flash in casinos. So every action on the casino machine has to be recorded three times or in three different places. And if you’re using an OR flash, it takes forever to record that action, and that limits the number of times you can actually play on the casino machine.

If you replace that with MRAM, you get a reliable memory, you get a faster memory, and now you can get more actions on that same casino machine as opposed to an RF flash memory. So that’s another steady revenue or high segment that we have revenue from in the Persist family. The third one I would highlight is the aerospace and defense industry. So we ship a lot of parts into automotives. For example, I was talking about battery management systems.

I was talking about Boeing and Airbus, for example, and then various applications through our primes like Honeywell and Frontgrade, where we have actually gone into space on our way to Jupiter on the NASA mission with Lucy. And then we are also actively collecting data from Mars today with our Toggle MRAM over there. So that is our existing revenue. So going forward, where do we see growth? Where we see growth is our NOR flash actually stopped scaling at 40 nanometer CMOS.

And also the density is basically that you can buy our NOR flash today, but basically around two fifty six megabit is the highest density you can buy. So we are now targeting our STT MRAM product to actually scale, sorry, to enable the roadmap for higher densities but replacing flash with our MRAM. So we are actually shipping today anywhere from 64 megabit to 128 megabit MRAM that looks like NOR. And in the future, we are actually designing a product that takes that same density from two fifty six megabit to two gigabit. So we’re going to be able to address this market.

This has a huge TAM of about $4,300,000,000 and there’s no other company that is actually building MRAM today that can address this space. So we are very excited and we are actually bringing out a first part with the high density in 2026 timeframe.

Robert Blum, Managing Partner, Lithium Partners: Good. Let’s talk about the sort of the competitive environment here. Who are some of your primary competitors? How do you position yourselves against them?

Sanjeev Agrawal, Chief Executive Officer, Everspin: Yeah. So if you look at the competitive landscape, there is only one company called NetSol, like Network Solutions. They’re based out of South Korea and they use Samsung as a foundry, just like we use GlobalFoundries as our foundry. But they only have one offering. They have a 16 megabit part that they actually sell down to four megabit as well, but there’s only one native die, 16 megabit, and that’s it.

And the next part that they’re actually designing is all the way up to one gigabit, which is about three years out. So if you compare that to us, we have an offering all the way from two fifty six kilobit to two fifty six megabit, and then bringing out a one gigabit part in the next year or so. So much broader range of parts that we can offer in industry standard package parts, BGAs, T SOP, etcetera. So I think NEXOL is there, but I think they offer only one small density and therefore we don’t see them too much in the space with our customers. So because of the offering that we have, the customers actually like to see our roadmap they would rather design us in compared to our competitors.

The other one is Avalanche Semiconductor. They’re based out of California, and they are actually focused now only on the aerospace and defense industry. So they talk about one gigabit STT MRAM that they’re planning to bring to product over the next year or so. We’re targeted towards USD applications.

Robert Blum, Managing Partner, Lithium Partners: When you sort of take a step back here and you look at some of the key market trends, and I think you’ve hit on a number of them here, but what are the biggest sort of trends that are shaping the memory industry going forward here? And how are you sort of ideally positioned to capitalize off these?

Sanjeev Agrawal, Chief Executive Officer, Everspin: Yeah, good question, right? So at least two or three different areas. One we just talked about, nor flash. There isn’t another solution because nor flash stops scaling at 49 meter CMOS. So we are positioning ourselves to look at that market.

Now more recently, we’ve also realized that this low earth orbital market or the LEO market is actually growing pretty fast. One of the reports actually said that you’re going to have 70,000 satellites over the next five years or so. And that’s greenfield for hourspend and because MRAM is actually radiation tolerant, it can actually serve the LEO markets very well. And examples of that is you might have seen our press release with Blue Origin, where we are actually designing over there, and then also with Astro Digital. So most of these guys are looking for reliable memory that is radiation tolerant, and that’s where Everspin’s MRAM comes into play.

And there is, like I said earlier, there is no other player that is actually offering MRAM for these markets today. So I think it is another good opportunity for us. And we are making sure that we collect all the data to show radiation tolerance, and then we are working with the right people or the right vendors to make sure that we get designed in as these LEO markets evolve over the next few years. And then if I didn’t mention AI, I would be remiss, right? So today, if you look at the data centers, they are heavily focused on very high density for compute applications.

In fact, slowly now they’re actually turning towards inference. It was all about high density, it was all about compute, and now they’re actually looking at inferencing on the edge as well and in the data centers. That’s where MRAM comes in. If you’re looking at if you’re using SRAM today, for example, for inferencing, what are the challenges with that? The challenge is that it’s actually very leaky, and you have to keep refreshing it.

And then you also have to have a backup in case you lose power. So if you use MRAM over there, you can actually you’re as fast as SRM and you can actually compute or execute in place. And then it’s non volatile, so you don’t have this large leakage that is associated with SRAM. So you can then start replacing all this SRAM or NOR flash that you have on the edge with MRAM for faster AI because you don’t have to go to the cloud every time you need to make decision. You’re losing a lot of power going back and forth.

You can make the decision on the edge, you can execute on the edge. And that’s where I think MRAM has a role to play in the coming future.

Robert Blum, Managing Partner, Lithium Partners: Big topic conversation here of late has been sort of supply chains, tariff, manufacturing locations. Talk a little bit about the company’s operations, any restraints, any issues that you’ve had of late.

Sanjeev Agrawal, Chief Executive Officer, Everspin: So in Q2, what we reported, we did not report any impact or even in Q1, we did not report any impact on tariffs. So we haven’t seen any yet, but it’s a changing landscape. Every day there are new rules, so we have to keep in mind all that. But as far as the supply chain is concerned, it’s pretty diversified. For example, the MRAM that we manufacture in Chandler in Arizona, we actually buy the front end wafers either from TSMC in Washington State, or we actually get them from Taiwan from Fab six.

And then we finish the MRAM over here, and then we do our packaging and final test in Taiwan, and from there we distribute to our customers. For Sprint TransferTalk MRAM, we actually use GlobalFoundries as a single source. By that, I mean as a single turn. So they do the front end as well as the back end. The front end is done in Germany, the back end is done in and by the back end, I mean MRAM, that’s done in Singapore.

And then the parts are shipped again to Taiwan for final assembly and test. And from there, they actually ship to our customers. So in that sense, don’t have a very you can’t point a finger and say, Okay, 100% of our products are going to get impacted by this change in tariff policy. I think it’s pretty diversified and we keep looking at opportunities to diversify our supply chain.

Robert Blum, Managing Partner, Lithium Partners: All right, very good. You have a you just reported, I believe it was second quarter financial results here over the last earlier this month, I guess. Provide some highlights of of the quarter, customer wins, milestones, etcetera.

Sanjeev Agrawal, Chief Executive Officer, Everspin: Yeah, sure. So we actually reported a solid quarter one more time. We actually came in above our midpoint of guidance and above the consensus of the analysts. We came in at 13.2 versus 13. And once again, our non GAAP income earnings per share was $03 So again, another profitable quarter, which is obviously great for Everswed.

But the highlights that I would mention from that quarter is about three things. One is we actually starting to see a depletion of the inventory overbuild at our customers. So for the longest time, we actually didn’t see many orders coming in because of the inventory overbuild past the pandemic. So we’re starting to see the bottom. We’re actually starting to see orders grow.

And that was basically in three different applications. One, for example, in the automation, which is what I talked about as one of our primary business. The other one was in data center. And the third one is in the FPGA applications. And here I would like to point out that the second highlight was the new parts that we have brought to market over the last two years, the XSPI family.

We’re actually starting to see significant design wins in this family, and we expect to start generating revenue in the ’5. And last but not the least, in preparing for our growth, we actually have expanded our leadership team. We actually separated out our position of VP of Business Development and Sales. So we now have a dedicated VP of Business Development, David Schrenk, and we brought on Sean Dougherty as our VP of Sales. He just joined us from Intel, he was actually the chief revenue officer for the Ultera business over there.

So very excited to have him on board and looking forward to our growth over the next few years.

Robert Blum, Managing Partner, Lithium Partners: Alright. This is an investor conference, obviously. Talk about, I guess, what you would see as the key highlights investors should be looking at and why you think this is an attractive opportunity here for shareholders.

Sanjeev Agrawal, Chief Executive Officer, Everspin: So a few things. One is a solid balance sheet. We are debt free and we have about $45,000,000 of cash in the bank. And we continue to increase our profit margins and we continue to run a profitable business. So that’s number one.

Second is we are actually targeting a pretty large TAM of about $4,300,000,000 by 2029. So we actually have products that address the data center market, products that address the data logging market, and also the configuration memory or the NOR flash. So I think we are well positioned to address these markets and are well along the way. The other thing is we are the singular domestic supplier of MRAM for mission critical applications. So we are designed into several US government contracts or projects where we actually provide radiation hard MRAM.

The front end is actually hardened by our defense primes. We only do the MRAM portion, which is the same, whether it’s commercial or the US government. But because MRAM is naturally radiation immune, it actually helps with the USG applications once the front end has been hardened by our primes. The other thing is we have a pretty diversified customer base like we were talking about, Probably 2,000 plus customers worldwide. So it’s not like if one customer had a downturn, it’s going to impact our revenue or profitability in a big way.

So pretty diverse business, pretty sustainable, profitable business that’s growing over the next few years with a large TAM.

Robert Blum, Managing Partner, Lithium Partners: A strong balance sheet, as you Final mentioned question here, as you sort of look ahead, what is your vision for this company over the next decade? And maybe a decade’s too long here, but go as far out as you feel is appropriate. What’s sort of that legacy that you hope to leave as CEO here?

Sanjeev Agrawal, Chief Executive Officer, Everspin: So MRAM is the future of memory. I think that’s what I would like to be my legacy once I’m done from this role. And the way to get there or the way to make the mark over there is the journey that we have started. I think the MRAM can be tuned to look like any memory for any application, whether it’s a data center, whether it’s AI, whether it’s an or flash, whether it’s data logging. So it’s such a versatile memory and it’s so reliable.

I think is just a pleasure to see it get adopted more and more for these different applications. And I think last but not the least is the environment and the climate. There is so much power being consumed and people are planning for so much power consumption with these AI applications. I think MRAM is a beautiful solution. It can actually reduce the power consumption, increase the speed and make life easier for the common person as well.

So I think, like I said, MRAM is the future of memory and I think we’ll get there.

Robert Blum, Managing Partner, Lithium Partners: All right. Very good. Well, Sanjeev, thank you so much for your time today. I want to thank everybody obviously for continuing to watch throughout the day here. If there are any questions for Sanjeev that I can help direct to him or to the investor relations team there, please let me know.

Again, that’s Bloom, blumlithympartners. We have additional presentations, fireside chats coming up here, so please stick around. Again, Sanjeev, thank you so much for your time. Greatly appreciate it.

Sanjeev Agrawal, Chief Executive Officer, Everspin: Was a pleasure, Robert. Thank

This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.