Welcome to GSI Technologies Fourth Quarter and Fiscal Year 2025 Results Conference Call. At this time, all participants are in listen-only mode. Before we begin today's call, the company has requested that I read the following safe harbor statement. The matters discussed in this conference call may include forward-looking statements regarding future events and the future performance of GSI Technology that involve risks and uncertainties that could cause actual results to differ materially from those anticipated. These risks and uncertainties are described in the company's Form 10-K filed with the Securities and Exchange Commission. Additionally, I have also been asked to advise you that this conference call is being recorded today, May 1, 2025, at the request of GSI Technology. Lee-Lean Shu, the company's Chairman, President and Chief Executive Officer, will be hosting the call today. With him are Douglas Schirle, Chief Financial Officer; and Didier Lasserre, Vice President of Sales. I would now like to turn the conference over to Mr. Shu. Please go ahead, sir.

Good afternoon, and thank you for joining us to review our fourth quarter and the fiscal year 2025 financial results. Let's start with a few highlights from fiscal year 2025. We closed the fourth quarter with solid revenue growth, significantly reduced net loss and a meaningful improvement in cash burn, finishing the year with $13.4 million in cash and a more disciplined operating structure. Revenue for the fourth quarter increased by 14% year-over-year and 9% sequentially to $5.9 million, driven by strong demand for our SRAM chip as we exit this year. This revenue growth and lower operating expense resulted in a sharp reduction in quarterly net loss and a material decrease in cash usage. For fiscal year 2025, while annual revenue declined 6% compared to the prior year, we meaningfully reduced our net loss by 47% from $20.1 million in 2024 to $10.6 million, driven by a 35% reduction in operating expenses. This structural cost improvement is central to our goal of preserving cash and extending our runway. We expect to maintain our quarterly operating expenses at the current levels to minimize our cash burn until we secure new funding sources. In the fourth quarter, we made good progress across multiple fronts to advance our technology roadmap and the commercial strategy. Notably, we secured an initial order for radiation-hardened SRAM from a North American prime contractor, a key validation of our product. We anticipate follow-on orders this fiscal year. This chip carries a significantly higher gross margin than our traditional SRAM. EDA is the point person with this customer and will expand further on the opportunities. Our ongoing SBIR program with government agencies are progressing well, and we are successfully meeting our milestones. Today, our SBIR has general payment totaling $1.6 million, and we anticipate receiving an additional $1 million once we complete the program. This quarter, $870,000 was booked as a reduction to R&D expense further helping to lower operating expenses. We are especially excited about recent enhancements to Plato, adding the integration of camera interface directly into the chip. This new feature paired with other connectivity enhancements allows the chip to interface with a wide range of sensors. This makes Plato particularly well-suited for AI agents requiring object recognition. The new capability has increased strategic interest in Plato and we are currently in preliminary discussions with multiple parties to scale partnership and asset for the next phases of development. Didier will provide more detail on this exciting development. As we look ahead to fiscal year 2026, we plan to build on the progress of our APU development, drive continued growth in SRAM sales and advance execution of our strategic initiatives across both commercial and government markets. At the same time, we remain committed to maintaining operational efficiency. In parallel, we continue to explore strategic alternatives with a primary focus on securing funding to support the next phase of operator development. We are also working with our banking team to explore other options that could provide new sources of cash to execute our AI strategy. With that, I will now hand the call over to Didier.

Thank you, Lee-Lean. As Lee-Lean mentioned, this quarter's primary revenue driver was the continued strong demand for our high-density SRAM. Our SRAM has been deployed in critical systems used in chip manufacturing, and the recent uptick in business with KYEC is being driven by surging demand for our next-generation AI chip from a leading GPU provider. Despite the ongoing tariff negotiations between the US and its trading partners, we currently anticipate the demand from this customer to continue in fiscal year 2026 at a similar level to what we experienced in 2025. With that said, we may have some variability in the timing of the shipments but importantly, the demand is still anticipated to remain consistent. The big news this quarter is an initial order for our radiation-hardened SRAM. While waiting for the forecast from the prime contractor, we anticipate follow-on orders in fiscal 2026. In addition, we are actively working with this customer to secure heritage status. Gaining this status would enhance the market acceptance of our radiation-hardened SRAM and unlock access to new high-value sales channels. It is worth noting that radiation-hardened SRAMs carry a gross margin well above those of our traditional SRAM chips providing a strong financial lever as we work to reduce our net loss and cash burn. Let me switch to Plato and elaborate on Lee-Lean's earlier comments. By integrating a camera interface directly into the chip alongside enhanced connectivity features, Plato significantly broadens its addressable market. Plato processes data locally without relying on cloud infrastructure; it's now optimized for edge devices and ideal for agents performing object recognition. To clarify when an agent is, it's helpful to look at how the approach to AI is shifting to Agentic AI. These AI systems don't just analyze data but they also act independently, for example, generating motor commands for a robot or a drone. This involves multiple capabilities that a single-purpose GPU is not well suited for. Plato, on the other hand, can manage a combination of computing tasks that involve more than just a single number of crunching or graphic workloads. Put another way, Agentic AI goes beyond basic data analysis. It must make decisions, process inputs from sensors like cameras and microphones, respond in real-time, and take actions in the physical world. In this context, Plato's capabilities position it at the forefront of sectors preparing for significant growth driven by the increasing demand for intelligent autonomous systems or agents across various industries at the edge. Thus interest in Plato has grown among the strategic partners we've engaged with over the past year. Pivoting to our ongoing SBIRs as Lee-Lean stated, these projects are on track and we are meeting the milestones. As a reminder, we are currently working on a Phase 2 contract from both the Space Development Agency and the Air Force Research Labs along with our most recently announced Phase 1 contract with the US Army. As planned, we delivered a server with a Leda-2 board to the Air Force Research Labs and will shortly deliver another Leda-2 board to the Space Development Agency. The Phase 1 SBIR for the US Army contract is evaluating the use of Gemini-II in edge computing AI solutions, and we are on track to meet all expectations with this partner. This quarter, we also delivered a YOLO algorithm for the Air Force Research Labs including the benchmarks for a real-time object detection application. We continue to increase the performance of the YOLO algorithms which can immediately determine the exact placement and identify the type of objects. We plan to deliver the improved YOLO3 and YOLO5 algorithms this summer. Lastly, an update on our SAR projects. We made further progress with an offshore defense R&D customer which ordered a Gemini-II system to evaluate the chip's capabilities for low-power in-flight applications. We will be shipping this system this quarter. This organization is also a potential funding partner for Plato. In addition, a US aerospace company continues to evaluate our Gemini for onboard satellite applications. Taken together, these activities support the use of Gemini-II for integrated edge applications such as SAR generation and drones, with subsequent object detection and actionable decisions. Now I will move on to the customer and product breakdowns for the fourth quarter. In the fourth quarter of fiscal 2025, sales to KYEC were $1.7 million or 29.5% of net revenues compared to $544,000 or 10.6% of net revenues in the same period a year ago and $1.2 million or 22.7% of net revenues in the prior quarter. In the fourth quarter of fiscal 2025, sales to Nokia were $444,000 or 7.5% of revenues compared to $694,000 or 13.5% of net revenues in the same period a year ago and $239,000 or 4.4% of net revenues in the prior quarter. Military defense sales were 30.7% of fourth quarter shipments compared to 35.5% of shipments in the comparable period a year ago and 30.0% of shipments in the prior quarter. SigmaQuad sales were 39.3% of fourth quarter shipments compared to 42.4% in the fourth quarter of fiscal 2024 and 39.1% in the prior quarter. I'd now like to hand the call over to Doug. Go ahead, Doug.

Beginning with the results for the quarter, we reported net revenues of $5.9 million for the fourth quarter of fiscal 2025 compared to $5.2 million for the fourth quarter of fiscal 2024 and $5.4 million for the third quarter of fiscal 2025. Gross margin was 56.1% in the fourth quarter of fiscal 2025 compared to 51.6% in the fourth quarter of fiscal 2024 and 54% in the preceding third quarter of fiscal 2025. The year-over-year and sequential increase in gross margins was primarily due to higher revenue and product mix. Total operating expenses in the fourth quarter of fiscal 2025 were $5.6 million compared to $7.2 million in the fourth quarter of fiscal 2024 and $7 million in the prior quarter. Research and development expenses were $3 million compared to $4.8 million in the prior year period and $4 million in the prior quarter. Research and development expenses in the fourth quarter of fiscal 2025 were reduced by $870,000 reflecting government funding under the SBIR programs. Selling, general and administrative expenses were $2.6 million in the quarter ended March 31, 2025 compared to $2.4 million in the prior year quarter and $3 million in the previous quarter. Fourth quarter fiscal 2025 operating loss was $2.3 million compared to an operating loss of $4.5 million in the prior year period and $4.1 million in the prior quarter. Fourth quarter fiscal 2025 results included interest and other income of $52,000 and a tax provision of $6,000 compared to $108,000 in interest and other income and a tax benefit of $85,000 for the same period a year ago. In the preceding third quarter, net loss included interest and other income of $70,000 and a tax provision of $44,000. Net loss in the fourth quarter of fiscal 2025 was $2.2 million or $0.09 per diluted share compared to a net loss of $4.3 million or $0.17 per diluted share in the fourth quarter of fiscal 2024 and a net loss of $4 million or $0.16 per diluted share in the third quarter of fiscal 2025. Fourth quarter pre-tax stock-based compensation expense was $512,000 compared to $693,000 in the comparable quarter a year ago and $429,000 in the prior quarter. Turning now to the full year results for fiscal 2025. We reported net revenues of $20.5 million for fiscal 2025 compared to $21.8 million for fiscal 2024. Gross margin for fiscal 2025 was 49.4% compared to 54.3% in the prior year. The decrease in gross margin was primarily due to product mix and the effect of lower revenue on the fixed costs and our cost of revenues. Total operating expenses were $21 million in fiscal 2025 compared to $32.3 million in fiscal 2024. Research and development expenses of $16 million compared to $21.7 million in the prior fiscal year. Selling, general and administrative expenses were $10.8 million compared to $10.6 million in fiscal 2024. The decline in research and development expenses was primarily due to cost reductions announced in August 2024. Research and development expense in fiscal 2024 included preproduction mask costs of $2.4 million related to our APU2 product. Research and development expenses in 2025 and fiscal 2024 were reduced by $1.2 million and $440,000 respectively reflecting government funding under the SBIR programs. Operating expenses in fiscal 2025 include a gain on the sale of assets of $5.8 million from the sale of the company's headquarters building in Sunnyvale California and a sales and leaseback transaction. The operating loss for fiscal 2025 was $10.8 million compared to an operating loss of $20.4 million in the prior year. The fiscal 2025 net loss included interest and other income of $326,000 and a tax provision of $130,000 compared to $414,000 in interest and other income and a tax provision of $70,000 in the prior year. For the fiscal year ended March 31st, 2025, we reported a net loss of $10.6 million or $0.42 per diluted share compared to a net loss of $20.1 million or $0.80 per diluted share in the prior fiscal year. On March 31st, 2025, we had $13.4 million in cash and cash equivalents compared to $14.4 million at March 31st, 2024. Working capital was $16.4 million as of March 31st, 2025 versus $24.7 million at March 31st, 2024. Stockholders' equity as of March 31st, 2025 was $28.2 million compared to $36 million as of the fiscal year ended March 31st, 2024.

At this point, we'll open the call to Q&A.

Can you talk a little bit about the market for the Plato chip and the Gemini-II chip? How large are these markets? How do they scale over what time period? Just give us a sense for the market conditions here. Thank you.

We haven't disclosed the total addressable market numbers yet, but regarding the markets, Gemini-II can be viewed as an enhancement of Gemini-I, specifically designed for edge computing. Gemini-I showcased our capabilities in search and high-performance computing applications like synthetic aperture radar. With Gemini-II, we aim to extend those capabilities to the edge. Unlike Gemini-I, which wasn't designed for low power use, Gemini-II gets us closer to edge applications. While Gemini-I focused on SAR applications at ground level, Gemini-II targets SAR applications on drones or satellites at the edge. Thus, we are emphasizing search and high-performance computing at the edge with Gemini-II. As for Plato, it is aimed at the large language model market rather than the search market. While many associate large language models and generative AI with data centers, we are bringing that capability to the edge. So, consider generative AI and large language model applications more oriented towards the edge.

Thank you.

Thanks, Michael.

Michel, does that answer your question?

Yes. It does. Thank you. Yeah. That's it. Thank you very much.

Thank you. The next question comes from the line of Robert Christian with Acoustic Technologies. Please go ahead.

Yes. Thanks for taking my question. I was wondering is the company experiencing any interest in the Gemini-II stand-alone chip from commercial companies other, say, the military? And is the company still working with the hyperscalers? Thank you.

So good question. Honestly, the majority of the early interest has come from more of the military defense type of applications and they are looking at it on a component level as well. So we do have what we call the Leda-2 board, which I discussed a little earlier that we have delivered and will be delivering to some of our partners. And so with the board we're delivering it with a SAR algorithm or some kind of a YOLO algorithm that's been developed, which is why they're getting a card. But some of the folks that we have been having discussions with they are looking for a chip only, because they will be mounting it on a drone or in a satellite. One of our other integrating partners that we've discussed in the past, they're actually going to be developing their own miniaturized board. And so they will be procuring just the Gemini-II chips from us to put on their proprietary board. So the answer is yes, we are seeing interest on the chip level, but there's also some board level interest as well.

Okay. And how about the hyperscalers?

Yes, we are currently focusing more on the edge with the hyperscalers and have had discussions with them, but the process tends to take longer. In contrast, we find that working with military clients leads to a much quicker path to revenue.

Okay. Can you share with the shareholders a little more detail if you can on what Needham is bringing to the table?

Well, really nothing is off the table at this point. It could be sale of assets. It could be funding into the company. It could be helping us with opportunities for R&D funding and help with the development of products that we're looking at. It really could be just about anything.

Okay. But you can't share any specifics at this time then?

There's nothing specific to talk about at this point or at this time yet.

Okay. Thank you.

Thank you. Our next question comes from the line of Michael Cooper, who is an investor. Please go ahead.

Actually, sorry, the last part of that last question, answered my question, which was what has Needham actually presented as options or opportunities for you and it doesn't sound like they've done an awful lot there.

We have explored some options, but nothing has come to fruition or is ready to discuss at this time.

Okay. Thank you.

Thank you. The next question comes from Jeff Bernstein with Silver Bernstein Capital. Please proceed.

Yes. Just a quick one, could you give us what cash flow from operations was in the quarter and what your CapEx was?

I don't have it for the quarter, but I have cash flow for the year. For the year, cash used in operating activities will be about $12.9 million.

Okay. And what was CapEx for the year?

Very little. Fixed asset additions during the year were like $45,000, very little very minor. I'm looking at fixed asset acquisition right now.

Thank you.

Okay. That's great. Thank you.

Thank you. As there are no further questions, I would now like to hand the conference over to Mr. Shu, for closing remarks.

Thank you all for joining us. We look forward to speaking with you again when we report our First Quarter Fiscal 2026 results. Thank you.

Thank you. This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.