NASDAQ: OSS

On August 31, 2018, the Company acquired Concept Development Inc. (“CDI”) located in Irvine, California. CDI specialized in the design and manufacture of custom high-performance computing systems for airborne in-flight entertainment, flight safety equipment, and networking systems. CDI’s business was fully integrated into the core operations of the Company as of June 1, 2020.

On October 31, 2018, OSS GmbH acquired 100% of the outstanding stock of Bressner Technology GmbH, a limited liability company registered under the laws of Germany and located near Munich, Germany (“Bressner”). Bressner was an integrator and distributor of hardware systems and components.

On December 30, 2025, the Company signed and closed a definitive agreement to sell all Bressner through a sale of all shares of OSS GmbH. The consummation of this transaction represented a strategic shift and prioritization of the Company's core business developing and manufacturing deployable edge computing systems for mission critical applications.

Our principal executive offices are located at 2235 Enterprise Street, Suite 110, Escondido, California 92029, and our telephone number is (760) 745-9883. Our website address is www.onestopsystems.com. Information contained in, or accessible through, our website is for reference purposes only.

Business Overview

OSS designs, manufactures, and markets specialized enterprise class high-performance compute, high speed switch fabrics, and storage hardware and software, which are designed to target edge applications for AI/ML, sensor processing, sensor fusion, and autonomy. Edge computing is a form of computing that is done on platform or on site, connected with the data source or the user, rather than in the cloud, minimizing the need for data to be processed remotely. This growing trend increases computing performance and security, as the data does not have to travel to distant datacenter locations. Edge computing is most recognizable in applications such as sensor processing, sensor fusion, autonomy, and AI/ML. To meet the demands at the edge, we offer specialized products and system solutions that consist of computers, switch fabrics, and storage products that incorporate the latest state-of-the art components with embedded proprietary software. Such products and systems allow us to offer high-end solutions to be integrated into edge platforms in our target markets.

The fast-growing edge computing space consists of three major segments. The first segment is comprised of smaller datacenters located near the user - on the edge. These typically include compute and storage racks in environmentally controlled buildings, similar to large cloud datacenters. Suppliers in this space tend to be the same large server and storage manufacturers whose products are used at cloud datacenters. The second segment includes billions of Internet-of-Things (“IOT”) devices that may reside in everything from home appliances to the factory

4

production floor. These IOT devices and applications tend not to be challenged on performance and easily communicate up to the cloud or the datacenters on the edge. OSS does not focus on either of the foregoing segments. The third segment is focused on edge platforms generally on the move or located in challenging environmental conditions. These edge platforms are primarily on land, in the air, or at sea on vehicles that need enterprise class datacenter level performance for sensor processing, sensor fusion, autonomy, and AI/ML applications. This is where OSS’ vision and strategy is aligned, and where we believe that we offer the greatest unique value.

Examples of these applications range from industrial autonomous trucks, to mining equipment and smart agricultural equipment, to military land, sea, and airborne weapon system platforms. Less mobile applications that utilize High Performance Compute ("HPC") and edge computing include items such as medical applications, mobile command centers, and certain datacenters.

Sensor fusion, autonomy and AI capabilities require that these demanding applications be connected to a wide array of data sources and sensors, and to have the ability to quickly access and store large and ever-growing data sets. They must be able to maintain ultra-fast processing power to act on data or process sensor inputs in real time at the edge location, which is independent of whether a high-speed network, like 5G or a tactical data link, connects the edge application back to the distant central data or command centers. Standard servers and storage systems available in the market do not address, and typically will not survive, the edge environment requirements. Although the network or data link, if it exists, may transfer data, or be used for updates, the latency is not acceptable for many of these applications where time is of the essence. This increases the need for enterprise class datacenter level performance using the latest generation of products from companies like NVIDIA, AMD, and Intel. In most of these edge applications, available space is limited and the number of inputs from sensors and other data sources are significant thus requiring high speed switch fabrics like the latest in PCI Express, for which OSS is a recognized expert in the market. In all cases, at the edge, there is a demanding operational environment that requires ruggedized solutions to deliver assured and continued operation, which must be balanced with size weight and power (“SWAP”) requirements. Solutions in this space must not only be light and fit into small volumes, but must also survive drops, g-forces, and vibration. Additionally, they must continue to operate at extreme temperatures and in dusty or wet conditions.

Many companies that enter this space tend to offer solutions based on older and lower performance technology, whereas we aim to advance our proprietary state-of-the-art technologies and utilize the latest generation of enterprise class products to help ensure superior performance. We aim to leverage our proven track-record of delivering first-to-market advanced technologies and technical strength, working with the latest high-speed networks like PCI Express, Ethernet, and NVIDIA’s NVLink®. This is in addition to our expertise with rugged servers, compute acceleration, and high-performance flash array storage systems. By combining our knowledge and execution for deploying these systems in challenging environments, we believe we bring the latest commercially available datacenter-level technology and products to this market.

Business Strategy

We continue to execute our strategic commitment to the rugged edge HPC market, and we believe that our engagement in the markets and with customers, supported by our portfolio of products and solutions, validates this approach. In this portfolio, we are witnessing some of our highest margins, repeat business, and expanding levels of interest. To this end, we have built a five year pipeline of target opportunities in excess of one billion dollars.

Although we are seeing progress, it will continue to take time to pursue, secure, and turn these target opportunities into increased revenue and profits, especially as we secure platform and incumbent positions in the military market. As part of this strategy, we are consistently focusing on expanding the number of customers and platforms while also securing multi-year contract opportunities. We believe this focus provides us with a unique opportunity to drive growth and shareholder value. We will continue to execute current profitable business where we offer product and solution differentiation, but our primary focus will be on our rugged edge HPC strategy.

We continue to see and pursue opportunities in both the commercial/industrial and military/government segments. In our military business, we will leverage commerciality to maintain margins. Commerciality is the ability to use commercial market pricing in defense procurements. We believe a balanced portfolio of commercial and military customers can serve as a strategic benefit in execution of the business, strategy, and growth. We believe that both commercial and military markets are currently seeing an influx of autonomy and AI/ML applications that may be driving market growth and strengthening opportunities. We believe commercial adoption of AI/ML is dramatically moving to the edge across almost all business segments. We believe that military spend on autonomy and AI/ML

5

represents one of the fastest areas of growth as the US and allied nations look to augment existing capabilities and field new sensors, platforms and weapon systems to maintain advantage over adversaries.

During 2025, we expanded our PCIe expansion portfolio with the launch of Ponto, a PCIe Gen 5 datacenter-focused expansion platform, and announced next-generation PCIe Gen 6.0 cable adapters and related expansion products to address emerging high-wattage AI accelerator requirements.

A key element of our business strategy is a product strategy based on technology leadership. We believe a first-to-market strategy is key to our ability to continue to win significant opportunities. As a result, we continue to develop new state-of-the-art products across a range of HPC demand, aiming to provide a unique value proposition for our customers in the targeted spaces. Current engagements for our products in the military space cover various autonomy, sensor fusion and AI/ML applications, including for aircraft, drones, ships, helicopters, and land vehicles, as the Pentagon prioritizes incorporating advanced technologies into their equipment. We have pursued these engagements throughout 2025 by incorporating an AI compute and storage product strategy that applies OSS expertise in three product levels that cover the needs of land, sea and air deployments.

At the high end, we introduced Torrey Break, which is an edge super-computing platform based on the latest PCIe 5.0 technology. We believe that this product is currently the highest performance, most compact supercomputer in the market that can survive the most demanding edge environments, especially those in military programs available today and in the future. It brings to market the latest in Graphics Processing Units (“GPU”), Central Processing Unit (“CPU”), high speed PCI express switch fabric, and memory products in a rugged compact form factor, to meet sensor processing, sensor fusion and AI/ML applications in this space. Torrey Break has lead customer interest from multiple large military prime contractors as well as the U.S. Department of Defense (“DOD”) directly for key, high visibility programs that we are pursuing.

At the mid-level, we have fielded our Short-depth Servers (“SDS”) in many commercial and government program wins due to its flexibility. The SDS provides integrated AI compute, data recording, and storage at a performance level approaching Rigel, suitable for wider market adoption. In 2025, the 3U SDS was adopted by two additional original equipment manufacturer ("OEM") customers in medical and government markets that accounted for two new program wins. These systems utilized our first-to-market advanced cold plate liquid cooling technologies that widen the environments in which the SDS can be deployed. Top commercial applications for the SDS include edge data analytics for applications in medical and autonomy. Top applications for government use include AI sensor fusion, simulation, capturing AI datasets through data recording, and serving high-speed data to crews with our Ion Accelerator™ data storage software in ships, submarines, and aircraft.

At the low end of performance, but with a higher level of ruggedization, we have shipped our Sensor Concentrator, Cernis™ and Donati™ to more government customers. These products create the most compact solution for autonomy, AI/ML, and visualization, such as 360-degree situational awareness. The Sensor Concentrator, Donati CPU/GPU compute and visualization systems are interconnected using the Cernis PCIe switch, creating an optimized data path that takes advantage of OSS low latency breakthroughs in PCIe interconnect, typically found in higher performance products like Torrey Break and SDS servers, while being highly rugged for use in wheeled and tracked land vehicles that require conduction cooling.

Our ability to drive the leading edge of technology is enhanced by our strong relationships with strategic component manufacturers, including NVIDIA (for GPUs, ARM processors and networking), NetList and Micron (for flash memory), Broadcom (for PCIe switch components), Microchip (for vehicle rated switch and functional safety processors), and Intel and AMD (for CPUs). In many cases, we have access to product roadmaps and other technical information relating to future technology. Access to this information allows us to begin our design process well before the future components we are designing even exist. We have also expanded our market relationships to embrace companies in the AI/ML software space. These efforts allow us to establish relationships that enable our ability to offer end-to-end solutions for customers while also creating opportunity for these software companies to standardize on our hardware platforms.

We anticipate that the steps we have taken in implementing our strategy and the fast-growing markets that we are pursuing will help to provide accelerated revenue and margin growth for the Company over the coming years.

6

Industry Background and Market Opportunity

The notion of network-based computing dates to the 1960s, but many believe the first use of “cloud computing” in its modern context occurred in 2006 when Google’s then-CEO, Eric Schmidt, introduced the term during an industry conference. Years later, the explosive growth of internet of things (“IoT”) connected devices, along with new applications that require real-time computing power, started to create the drive for edge-computing systems. As the demand for AI on the edge and autonomous vehicles grew, so did the need for high-performance solutions to operate in harsh environments and reduce risk of loss of connectivity to the cloud.

Edge computing is one of the fastest growing markets in the computing space, driven by the need to do more at the edge. The concept of edge computing is simply deploying compute systems closer to the actual user of the system, rather than communicating with a distant cloud computing facility.

Edge compute markets tend to implement sensor processing, sensor fusion, AI/ML, autonomous, and/or semi-autonomous solutions. We believe markets for these products are large and growing. Applications deploying these technologies today, or that we expect to do so in the future, include, without limitation:

•
Commercial/Industrial - Trucks, buses, trains, aviation, agriculture, mining, medical, oil & gas, etc.

•
Military/Government - Aircraft, rotary winged aircraft, ships, mobile command, mobile radar, submersibles, land vehicles, drones, etc.

We expect these applications to deploy increasingly faster computing systems at lower latencies to meet industry and competitive goals. Whereas the goal used to be for an edge compute platform to perform a single application, such as autonomous navigation, now this has been expanded to include in excess of ten or more AI applications running simultaneously. This expansion requires significantly more compute power and data storage, as well as lower latency, than traditional embedded computers can manage while operating in harsh, challenging, and space constrained environments.

Product Offerings

Our systems are built using the latest CPU, GPU, high speed switch fabrics, and flash storage technologies that draw upon years of expertise in designing and manufacturing custom, semi-custom, as well as standard systems for military and commercial customers. We have a history of being first-to-market with many solutions for emerging technologies. Our technological leadership includes linking different OSS product systems together.

We use leading edge, state-of-the art components from major technology providers to design purpose-built systems that solve customer problems in an efficient, cost-effective manner. We apply the component technology provided by Intel, AMD, NVIDIA, Micron, Broadcom, Microchip, and others to deliver solutions to provide true value to our customers.

Four technologies are fundamental to the edge computing space: edge servers, GPU compute accelerators, flash memory-based storage, and high-speed switch fabric for data acquisition I/O. These technologies enable systems to ingest, process and store data at significantly higher rates than traditional systems. By harnessing these components, customers can process data much more quickly and in a more secure manner and, as a result, turn raw data into advantage in action.

Rugged Edge Servers

While simple AI applications, such as facial recognition to open a door to a secure area, may run on a traditional low power embedded processor, the needs of edge computing applications require datacenter-class server performance brought to a mobile platform. The sheer amount and speed of sensor processing, sensor fusion and AI/ML processing for operating an autonomous vehicle or a mobile weapon system requires multiple high-speed digitizers, high-performance networking, the fastest flash storage devices, and server-class processors. We enable the power of the datacenter to be deployed at the edge without compromising performance by employing groundbreaking ruggedization and cooling technologies for edge servers operating from various vehicle AC and DC power sources in small spaces.

GPU Compute Acceleration

7

When GPU technology and solid-state flash were first introduced, we began designing systems that maximize the effectiveness of these technologies. We now produce compute-systems with large numbers of GPUs and flash memory that communicate over PCIe and allow faster processing, data storage, and data retrieval. The more GPUs and flash devices available to a server, the faster that system can process and store data.

The capabilities and speed of GPU accelerated computers are driving significant advances in AI and machine learning. Massive amounts of data are collected, stored, and analyzed by today’s sophisticated algorithms. We are enabling the growth of such AI capability by adding scale with rugged systems that complement the highest performance rugged servers.

High Density Solid-State Flash Storage

The proliferation of larger and larger data sets used in edge computing, including AI, is feeding the need for higher capacity and higher performance storage devices. Traditionally, companies have used hard disk drives for their primary storage. Hard disk drive-based systems are being replaced by flash memory-based systems, which offer higher capacity, performance, reliability, and ruggedness. Flash-based storage systems also consume significantly less power. Our solutions offer the highest capacity and performance with the addition of removable media which enables the quick transportation and delivery of massive amounts of data.

Switch Fabrics For High Speed Data

At the front-end of edge compute systems is high speed data acquisition technology through switch fabrics. Depending on the application, data can be generated from a wide array of sensors and inputs. In many cases, such as autonomous driving or a mobile weapon system, platform data is generated through arrays of video, Light Detection and Ranging (“LIDAR”), radar, Forward Looking Infrared (“FLIR”), Radio Frequency (“RF”), and other computationally intensive sensors. Our PCI express based switch fabrics provide the highest speed and lowest latency data acquisition and movement.

When PCIe was introduced in 2005, we were the first company to produce PCIe over cable adapters allowing system-to-system communication at the same speed as internal Input/Output (“I/O”) expansion, significantly reducing latency. Today, we are one of the largest providers of PCIe adapters and expansion components used worldwide.

What Sets OSS Apart

We believe several factors differentiate OSS from other suppliers of high-performance edge computing solutions, including, without limitation, the following:

•
Ruggedized SwaP-C – We provide systems that can support datacenter performance at the rugged edge by implementing unique form factors optimizing space and weight, variable power inputs sources, and cost competitive solutions across a wide range of applications, including full mil-spec systems.

•
Lowest Latency – We have expertise in PCIe and NVLink® switch fabrics that deliver the lowest possible latency. This allows us to design reliable systems using this differentiating high-performance technology with a greater quantity of GPUs and flash storage devices than other suppliers.

•
Advanced Cooling – We provide systems that can withstand extreme environmental temperature ranges. We engineer cooling for extremely high power heat generation, including heat from GPUs and Field Programmable Gate Arrays (“FPGAs”), utilizing differentiated methods for conduction, convection, liquid, and immersion cooling.

•
Key Partnerships – We have long and established leading edge technology partnerships with early access to product roadmaps, which include a differentiated relationship with NVIDIA. We are an “Elite Tier 2 OEM” in the NPN program with a unique NVLink® license.

•
Speed To Market – Our short development cycle times allow us to bring the newest technology to market quickly.

•
Powerful Software – We provide software products required to operate high-capacity, low-latency storage systems used by defense systems and commercial applications.

8

We have consistently followed a strategy of being first-to-market in leading edge deployment technologies by designing and developing products that are delivered before our competitors. We currently have products spanning the spectrum of high-performance computing, including servers, flash storage, GPU acceleration, networking, and PCIe data acquisition I/O expansion. Within these product areas, the approach implies that we:

•
Anticipate trends in these markets and do not hesitate to share our vision with customers to create thought leadership and deeper engagement;

•
Swiftly deploy resources in engineering and sales to bring innovative products to market before our competitors;

•
Leverage strategic relationships to get early access to future products and technologies;

•
Hunt for early program wins with market leaders and leverage close relationships;

•
Continuously monitor and influence the market for next generation technologies for which new concepts and solutions may be forming; and

•
Establish leadership in fast-growing edge computing markets

Earnings Growth Strategy

We intend to implement different strategies to increase our revenue growth, while improving earnings. We believe that earnings growth can be accomplished by taking the following actions:

Revenue growth driven by existing OEM and new design wins:

•
Focusing on the fast-growing, higher margin edge compute market;

•
Demonstrating technology leadership with a clear, ever building value proposition;

•
Utilizing a higher percentage of leading edge standard products for scalability;

•
Pursuing market leading OEMs in the edge compute market;

•
Continuing to layer in recurring high margin business;

•
Constantly improving and expanding a highly skilled direct sales force complemented by well positioned third party manufacturer representatives;

•
Expanding worldwide sales efforts and marketing opportunities appropriately; and

•
Completing sizable accretive acquisitions.

Higher margins:

•
Challenging OSS-designed solutions that leverage our skills;

•
Increasing proprietary content, software, stickiness, barriers of entry, and differentiating features;

•
Prioritization of the highest return programs/markets;

•
Maximizing military and other high value applications and sectors;

•
Retaining commerciality on products for the defense market;

•
Leveraging economies of scale and lowering material costs;

•
Eliminating lower margin products and markets;

•
Driving operational efficiencies up through automation, discipline, and process improvements; and

•
Layering in additional high margin services.

9

Optimize expenses:

•
Promoting a culture built on lean principles that innovates to minimize spending and drive higher efficiency per employee;

•
Utilizing technology and talent on team to increase efficiency;

•
Leveraging efficiencies of scale; and

•
Welcoming organizational change as business and markets adjust.

Our Opportunity

We believe the worldwide edge computing market is expected to grow rapidly. Within this market, we are positioned and focused on the AI/autonomous portion at the very edge.The products we develop to address this market include high performance compute, storage servers, and PCIe acceleration systems that can perform in the most challenging environments. If an application needs AI and/or autonomous capabilities, and it moves, OSS strives to deliver the highest performance solutions that will survive and enable these challenging applications.

Our Technology

We design and manufacture ruggedized high performance edge computing systems, which are designed to increase compute performance while surviving in harsh environments. Our rugged servers bring the power of leading-edge datacenter-class technologies to the edge for applications in AI/ML, sensor fusion, autonomous navigation, data logging, and video rendering. Our high-density compute accelerators connect directly to a server’s PCIe bus, delivering substantial compute performance. Our flash storage arrays support hundreds of terabytes of high-speed storage that can also be accessed by multiple servers.

Technology Drivers for OSS High-Performance Computing Business

We have developed expertise and core competencies in the three fundamental technology drivers within today’s high-performance edge computing market: high-speed serial interconnect technology, compute acceleration utilizing GPUs, and low latency flash storage. We believe that in combination, these three fundamental technologies, underpinned by PCI Express, are changing the economics of computing, bringing high-performance computing within the grasp of a wide range of new industries and commercial applications on the edge. Simultaneously, the emergence of massive amounts of data being generated by advanced sensors is pushing the requirement for innovative state-of-the-art technology. We are enabling this technology to be deployed at the edge by merging these fundamental technologies with our expertise in providing system level optimization to meet requirements for ruggedization and SWAP constraints. Our strategy is to be the disruptive leader in platforms for rugged edge applications, based on our unique ability to design high-quality, high-performance AI workflow compute/storage engines that can be deployed in harsh dynamic environments, which require unique system-level features for vibration, cooling, and power. Because our target platforms operate on land, in the air, and at sea, our compute systems must be compact, withstand harsh conditions and power disruptions, and deliver datacenter-class performance. We believe this is what we do well.

We strive to not only provide competitive advantage for our customers, but also to address some of the most fundamental challenges in military and industrial applications. We believe that we are well situated to leverage these major industry forces. By exploiting our unique set of expertise in the underpinning technologies of high-performance computing, we strive to continue to deliver industry leading solutions, disruptive at times, and to take advantage of the opportunity to capture a growing market share in this rapidly expanding marketplace.

Switched Serial Interconnect

Switched serial interconnects are the data highways connecting many elements of today’s high-performance computing platforms. At ever increasing speeds, these pathways move data between a system’s processing units, storage, networking, and peripheral elements. For high-performance computing, the primary processing, storage, and peripheral interconnect is PCIe. PCIe Gen 6.0 has an ability to run up to 16 lanes in parallel, which allows up to 256 gigabytes (full duplex) per second bandwidth between system elements. In 2025, we launched next-generation PCIe Gen 6.0 products, including CopprLink™ cable adapters, designed to meet increasing performance demands in edge computing. We expect to further address target markets with PCIe Gen 6.0, Compute Express Link, and related technologies in 2026 and beyond.

10

Compute Acceleration with GPUs

GPUs have evolved from graphics display acceleration to general-purpose processing workhorses for high-performance computing systems. Today, most of the fastest supercomputers in the world utilize GPUs as their primary compute engines. GPUs are ideal for high-performance computing workloads, including AI training and inference, because of their ability to complete massively parallel processing. While traditional CPUs may have dozens of processing cores, GPUs have thousands of cores that are able to execute calculations simultaneously.

NVIDIA, a key supplier of GPUs to the market, lists more than 500 such applications across a broad set of market spaces, along with focused teams on specific industries of high growth potential such as their NVIDIA Drive team for autonomous navigation. We believe the main markets serviced by NVIDIA GPUs include, without limitation:

•
Autonomous navigation;

•
Computational finance;

•
Climate, weather and ocean modeling;

•
Computational chemistry and biology;

•
Data science and analytics;

•
Deep learning and machine learning;

•
Federal defense and intelligence;

•
Genomics;

•
Manufacturing;

•
Media and entertainment;

•
Medical imaging;

•
Robotics;

•
Oil and gas; and

•
Safety and security.

While we believe NVIDIA is focused on the deployment of their GPUs in hyperscale datacenters and for consumer gaming purposes, we are focused on taking the datacenter class capability to the edge compute market, expanding the overall market significantly. Many of these applications also scale performance, based on the number of GPU components utilized. We have designed multi-GPU systems, including up to 16 GPUs in a single system. Current state-of-the art GPUs (NVIDIA B300) provide over 30 petaflops (FP4) of performance, with future products such as the NVIDIA Rubin, AMD MI400 series and Intel Gaudi series GPUs expected to dramatically increase overall processing capabilities in the years to come.

GPUs also pose significant system design challenges due to their high-power requirements. High-end GPUs can require 1000 watts of power or more, which generates a tremendous amount of heat. Sophisticated power distribution and cooling designs are required, especially for large-scale systems with multiple GPUs per chassis. OSS has significant expertise in addressing these challenges.

PCI Express Flash Storage – NVMe protocol

The use of flash memory technology for system storage has gained traction over the past decade, which we believe to be a result of the continuous decline in the cost per gigabyte. Flash memory has become the ubiquitous storage technology in high-performance systems.

Combined with the move away from traditional rotating hard drive technology, there has been the trend toward eliminating traditional storage protocols in favor of low latency flash memory protocols. Newer flash memory modules utilize a protocol known as NVMe, which connects the flash memory directly to the system’s PCIe interconnect. This

11

direct connection allows for very high bandwidth between the storage and the other system elements, which eliminates the need for protocol translation as data moves from storage subsystems to and from the compute complex.

Today, flash memory modules with capacities up to 120 terabytes and PCIe Gen 6.0 interfaces are now available for engineering samples. PCIe Gen 6.0 devices are expected to be available for general availability purchase in 2026 with capacities up to 240 terabytes. Our flash storage arrays with up to two petabytes of capacity are available, enabling the scaling of high-speed storage to meet the full range of high-performance edge application requirements.

OSS aims to leverage the latest technology to build complete storage and data logging systems, including all the software to provide the highest density and performance in a compact form factor ideal for vehicles of all types. This, combined with our hot swappable canisters, has enabled autonomous truck and military aircraft applications.

Our Core Technical Capabilities

We have developed unique expertise and core competency across the fundamental technologies of today’s rapidly expanding specialized high-performance edge computing marketplace. These valuable assets are embedded in the leading-edge engineering capabilities of our engineers, the proprietary intellectual property residing in our vast library of designs, and our brand equity based on our reputation as a high-quality producer of state-of-the-art, custom and standard solutions across a broad array of markets.

High Speed System Interconnect Design

Our electrical engineers are experts in high-speed digital signaling design. They have continually designed at the leading edge of the state-of-the-art signaling speeds, even as semiconductor technology has driven up the clock rate of digital transmission. We have consistently been among a small handful of companies able to come to market first with the latest technology. In fact, we delivered the industry’s first PCIe over cable solutions for PCIe Gen 1.0, Gen 2.0, Gen 3.0, Gen 4.0, Gen 5.0, and Gen 6.0. The expertise required includes circuit design, PCB (printed circuit board) layout, and routing optimizations, all of which focus on achieving the highest levels of signal integrity. In our current systems, PCIe Gen 6.0 signals are propagated across multiple PCBs and connectors, as well as both copper and fiber optic cabling, while maintaining the ability to recognize digital signal transitions at 32 billion times per second.

In high-performance computing systems, especially those systems that operate at the edge, the trajectory and need for ever-increasing signaling speeds is continuing; however, the number of companies that have the capability to design robust, highly reliable systems at speeds that can tolerate harsh conditions at the edge is limited. We believe our core competency in large-scale, high-speed design and layout will allow us to remain at the forefront of this growing industry.

Complex System Design

In addition to signal integrity design expertise, we have amassed expertise and intellectual property in high-performance system architecture design and software. This expertise allows us to develop extremely sophisticated systems with massive scaling, while also meeting customer demands for reliability, cost, and flexibility. To do so, we have developed deep knowledge for high-capacity input/output systems, operating system adjustments, and required configuration tuning. Because of this, our engineers are often called upon to co-design with OEM designers to create the perfect solution to fit the needs of their customers.

For highly scalable systems, a deep understanding and experience with switching topologies, interconnect fabric design, and low latency acceleration software is required. We have worked with serial switching technology, starting with the first generation of PCIe, and have been an innovator in creating unique and flexible topologies to meet the specific needs of customers. Creating custom solutions for unique customer requirements is a core competency at OSS, and we rely on this deep knowledge of switch capabilities and limitations.

Our software expertise includes developing unique hardware drivers and configuration software that accelerate bandwidth and lower latency in PCIe GPU and NVMe system designs. We have developed expertise in system design to leverage peer-to-peer data flows between GPUs, as well as pioneering techniques for optimized data flows between flash storage and GPU compute-engines. Our systems optimize switch and GPU configuration topologies to optimize GPU-to-GPU communication without requiring latency-inducing data transfer between host dual processors. Our

12

platforms feature RDMA (remote direct memory access) across compute-nodes, which support data transfer without burdening the host CPU, as well as NVMe over Fabrics for efficient data transfer from remote storage to compute.

We have pioneered the ability to extend the PCIe bus beyond the confines of a single enclosure, opening the possibility of flexible system expansion options. We believe we are one of the leading designers and suppliers of PCIe host bus adapters that extend PCIe signals from the host motherboard across copper or optical cables to expansion enclosures, which provide application acceleration through scale. Our adapters provide both ends of the external cable connection. Our expertise in high-speed signal design in printed circuit boards, connectors, and cables is essential to successful expansion designs. We also hold expertise in incorporating clustering and rack scale expansion into our system designs, including 200/400/800 gigabit Ethernet, 200/400 gigabit InfiniBand, and emerging PCIe top-of-rack switch technology.

Expertise in power, cooling, and mechanical design are required to address the requirements of high-performance computing customers, especially while meeting the constrained time requirements of rugged edge deployments. We have developed leadership design capability in high-power design and distribution within large rack enclosures as well as edge optimized configurations. High-end GPUs today require 500-675 watts or above, and in our high-end systems, up to 32 GPUs can reside in a single chassis. Thousands of kilowatts of redundant power are required. Power stability and huge thermal loads are some of the critical design issues that must be addressed. Additionally, at the edge for AI applications, a wide range of input power sources need to be supported from standard 110-220 VAC and 48-270 VDC for terrestrial vehicles as well as three phase 400-800Hz AC for airborne applications.

We have expertise in power distribution, redundant power, and complex chassis cooling design, including materials selection, airflow simulation, fan technology, and various liquid cooling options including direct to chip conductive cooling and single and dual phase immersion cooling. We have also developed extensive expertise to help ensure regulatory compliance of our complex high-performance computing system designs that span across emission, shock, vibration, thermal, humidity, and other environmental requirements that are required for highly reliable and highly available solutions. Our engineers are experts in design for regulatory testing for FCC (Federal Communications Commission), CE (European Conformity), UL (Underwriters Laboratories), and Mil-STD (Military Standard) standards. Additionally, we have expertise in rapid prototyping, design for manufacturability, and design for serviceability.

We also have expertise in system management software that enhances our competitive advantages and provides robust monitoring and management of the functions of complex computer systems. While the baseboard management controllers ("BMCs") of standard servers and workstations provide passable system management for an office or datacenter with a benign environment, the unique missions of our rugged edge systems require a more diverse level of monitoring, management. and control.

First, products like our 4UPro and EB line of PCIe high density expansion systems are not servers or workstations, but instead, they provide scale-out expansion of high-performance GPUs, FPGAs, NVMe drives and edge I/O devices to a server. To provide the widest compatibility and largest serviceable market for our expansion systems, our U-BMC, or “Unified Baseboard Management Controller,” allows our expansion systems to seamlessly integrate with existing customer servers or our own SDS and EOS server products, allowing a server and one or more of our expansion products to be managed, monitored and controlled, as if it were a single integrated system with a massive amount of PCIe resources interconnected by our highest bandwidth, lowest latency PCIe switched fabric.

Second, unlike the standard datacenter server, our servers are designed to operate in both defense and commercial harsh edge environments where “dirty” power from generators, engines and batteries with large spikes are common. Mix in the environmental conditions in which our servers may operate autonomously, such as temperature variances in places from Death Valley to 50,000 feet altitude, with moisture ranging from salt fog to rain and extreme vibration and shock from washboard dirt roadways to a hard-landing propeller aircraft, and one can see that a higher level of management, monitoring and control than a server snug in a power and cooling conditioned datacenter is required. The U-BMC adds unique value to server-level systems in these environments, especially in autonomous operations where there are no service technicians for miles, by allowing the system to adapt to changing conditions automatically or by remote control without failing. The U-BMC has features to handle these environments, such as controlling sensors connected to the server, turning on heaters in extreme cold, changing the PCIe fabric to reroute data around failed components and connecting to the Controller Area Network ("CAN") bus in cars and autonomous trucks to monitor the vehicle conditions to be able to take action on information provided by the vehicle, such as ignition on/off. Many more U-BMC edge features unique to OSS and valuable to our customers were in development in 2025 and will continue to be added in 2026, such as instant cryptographic erasure of sensitive information, access to add-in

13

card telemetry for modules supporting MCTP reporting standards, and throttling NVIDIA GPUs in real time based on power or temperature fluctuations to keep systems running in a reduced state rather than shutting down.

Third, due to operating in diverse industry standard and regulatory markets required by military standards, commercial aerospace FAA or EASA, and highway NHTSA or ETSC agencies, the U-BMC is designed to adapt to the unique requirements imposed by servers residing on, or controlling, vehicles. A standard datacenter server BMC needs to conform to the basic agency requirements of electrical interference and personal safety regulated by agencies such as the FCC or CE and administered by testing companies such as UL and TüV. The U-BMC is designed to cover all edge requirements of the datacenter as well as adding standards organization compliance, such as those required by military customers, including SOSA compliant functions for sensor management, system management, and task management not found in datacenter server BMCs.

In 2022, we introduced the U-BMC, which is included in our Rigel edge supercomputer and PCIe Gen 5 4UPro products. In 2023 and 2024, we expanded the platforms that include the U-BMC to the PCIe Gen5 SDS rugged server. In 2025 we added the PCI3 Gen5 Torrey Break 2U SDS for commercial and military edge applications with more platforms to come. With U-BMC, we provide “single pane of glass” management of complex systems, even if the server is in a separate enclosure, and an open-source Redfish API for easy integration with industry-standard management tools. We expect to continue to expand and enhance our licensable software to create additional value, barriers to entry, and stickiness with our program wins.

Storage Management Software

Given our hardware design and integration expertise, we believe that our robust software capability allows us to offer more optimized and customized systems. Our Ion Accelerator™ software design team provides the expertise to deliver full server and storage solutions that produce the highest performance from today’s leading-edge flash storage devices. The Ion Accelerator software allows flash-based modules to be put into a variety of storage and network configurations which can then be accessed by multiple edge servers. The Ion Accelerator software can do this cost-effectively, while preserving the low latency and security that is vital for many mission-critical applications, from secure network boot, database, and transaction processing to massive data collection programs.

Benefits of Technology and Core Capabilities to our Customers

Due to our core capabilities, we can provide our high-performance computing customers with platforms that are highly reliable and cost effective. Such performance allows our customers to solve larger problems faster, and save the cost and time of highly paid engineers, data scientists, and other human resources. We believe our technology enhances innovation by allowing more “what-if” analysis in a finite amount of time. Our price/performance leadership aims to enhance our customers’ competitiveness and lower capital expense and total cost of ownership. We work with our OEM customers to develop custom “perfect fit solutions” for their unique requirements when the anticipated return justifies the investment.

Our Products

Compute Servers

Within the server sector, we have secured a niche position building purpose-built specialty servers, which the major server suppliers do not supply, as they require custom tuning and special features that major OEMs cannot easily provide. Our compute servers are designed to provide the highest level of performance that can be deployed in harsh edge environments. Our expansion optimized short depth (“SDS”) and Torrey Break server product lines are optimized for supporting a high number of add-in cards such as GPUs, FPGAs, NVMe drives and sensor inputs. Servers in this product family have a number of slots that are compatible with the PCIe host bus interface cards that we have developed. These cards enable PCIe connection over cable between the host processor and downstream I/O devices. These servers have custom basic input/output systems (“BIOS”) to ensure they work seamlessly with expansion chassis and support a high number of downstream I/O devices. Our SDS and Torrey Break products also support rugged deployment in space constrained environments providing a maximum depth of 20 inches. We believe that our “Rigel Edge Super Computer” (“Rigel”) is the highest performing, most dense, AI-compute platform that is deployable in extreme environments, including on military aircraft. We expanded our PCIe 5.0 server configurations in 2025 and are positioned to upgrade those systems to PCIe 6.0 as the market makes PCI 6.0 processors, NVMe storage, and GPU accelerators available in the second half of 2026.

14

GPU computing uses hardware components that are optimized to perform mathematical calculations in a rapid fashion. NVIDIA is the market leader in the design and manufacturing of these components. We work closely with NVIDIA to design and build systems which use multiple GPUs to accelerate applications.

Emerging markets and applications such as AI, image rendering and processing, autonomous vehicles, deep learning, molecular modeling, genomics, advanced visualization, machine learning, and image processing all benefit from the ability to use GPUs to accelerate the application. We build specialized compute-servers and accelerators used in these emerging growth markets. We estimate these markets to be very large and growing.

Storage Servers

We also build standard and custom flash storage arrays utilizing our unique know-how in PCIe device fan-out, packaging, cooling, and PCIe-over-cable. We deliver dense, high-performance systems that provide customers with high value and utility in the most demanding, data-intensive operations. These OSS storage servers complement our compute servers to provide an end-to-end edge solution for AI workflows.

Through a strategic agreement with Western Digital, we acquired an exclusive software license for Ion Accelerator™ Storage Area Networking (“SAN”) source code and software development rights, and we hired their core engineering team in July 2017. Since acquiring this software asset, we have extended its capability to align with our edge computing strategy by adding Network Attached Storage (“NAS”), support for NVMe flash drives, NVMe over Fabric expansion capability, and several encryption methods required for government security applications. We have also implemented a proprietary Follow Me™ capability for removing a bulk pack of NVMe drives that can be easily transported to another system without rebuilding the data, operating much like a massive capacity USB stick. This provides our flash arrays with a high level of differentiation relating to storage management, latency, portability, and throughput. We provide standard flash array products and have the in-house hardware and software expertise to provide customized systems for demanding applications that are not suitable for standard offerings. For example, we provide products to a large military contractor for integration into military aircraft that requires us to design and manufacture a highly ruggedized mil-spec flash array. The resulting product provides high data density with low weight, a high degree of portability, and security for data protection. We believe our experience and capability in high speed, low-latency, digital signaling via PCIe gives us an edge in providing custom designs to OEMs, military programs, and other special purpose applications.

We believe that because our products are positively differentiated by speed, density, and management features for challenging edge applications, our offerings compete favorably in this market and provide a substantial growth opportunity.

PCIe Expansion and Adaptors

PCIe is the high-speed standard for communications within a computer. This standard defines the signals and connectors (i.e., slots) that are used for computer add-in cards (such as Ethernet or graphics). Traditionally, communication between computers in the network is completed via Ethernet. Although Ethernet is great for large networks, this introduces delays and latency challenges. To keep performance at the highest level, PCIe signaling can also be routed over a cable, allowing expansion input/output slots to be physically located in a separate chassis. This provides for high-performance and low latency, which are essential in this market.

Being able to separate the server from the I/O expansion, by using PCIe over a cable, facilitates disaggregation of server functionality. That is, with PCIe, server I/O functions no longer need to be contained in the physical server chassis, but instead, can be separated into a separate chassis and continue to operate at full speed. This offers many advantages over higher latency and power consuming traditional networking communications like Ethernet. From a practical perspective, servers can be connected directly to larger storage arrays or other peripheral devices, with the resulting group of chassis operating as if they were all in the same physical chassis.

We began developing our first PCIe-over-cable adaptor in 2006 and were one of the early providers of PCIe adaptors. We recognized this space as a prime opportunity to utilize our core strengths, such as:

•
High-speed board design and layout;

•
Signal integrity mastery;

15

•
Hardware tuning to improve signal integrity;

•
Design optimization for low cost;

•
Rapid design capability;

•
Rugged design to survive in harsh edge computing applications;

•
Custom BIOS to support a high number of connected PCIe I/O devices, well beyond what can be supported in off the shelf BIOS; and

•
Manufacturing and supply chain management.

This technology has now become a standard within the computer industry. Our customers have used our adaptors to connect their custom input/output chassis and achieve performance equivalence as if the I/O was integrated into the server box. This gives designers and integrators a degree of flexibility and utility in architecting computer systems that is unprecedented. We have expanded our PCIe adaptor market in breadth and depth, including making adaptors for many OEM customers.

With our expertise developed in designing adaptor cards, the logical extension of our capability led us to develop a method for expanding the PCIe bus into an external chassis containing one or many expansion slots and using this expansion to provide storage and AI application acceleration. This allows a customer to install multiple standard PCIe boards into a chassis and accelerate their system without having to add additional servers. These are typically GPUs, FPGAs, or NVMe drives to create large-scale Compute and Storage appliances. For example, we have developed a product for deployment in a mobile command center which aggregates large amounts of high frequency data from sensors and allows in the field AI algorithms to operate in real time. This is achieved through a cluster of our compute and storage products. A user can now connect a multiplicity of PCIe devices to a single server using a single memory domain, and achieve performance throughput and low latency, which was not possible prior to the introduction of PCIe.

We have been a leader in PCIe acceleration through generations 1.0, 2.0, 3.0, 4.0 and 5.0. We delivered our initial PCIe Gen 5.0 products in 2022, well ahead of any of our direct competitors, and we announced our PCIe Gen 6.0 cable adapters in 2025 and will start shipping these products to customers throughout 2026. As PCIe evolves through generations 7.0 and beyond, we believe that we are uniquely positioned to continue our leadership role in this market. We currently offer what we believe to be the largest PCIe acceleration product line, with chassis and backplanes that offer expansion from one to 64 slots. Due to its greater data throughput, lower latency, and flexibility of design, we believe this is a growing market, and we intend to maintain our leadership role.

Customers

We work to deliver the highest performance scalable products and solutions that enable AI/ML, sensor fusion, and sensor processing for global defense and commercial markets. We serve a global clientele consisting of multinational companies, governmental agencies, military contractors, military services, and leading technology providers.

Sales and Marketing

Our sales and marketing efforts are focused on the identification, engagement, and closure of significant targeted opportunities within the edge compute market.

Sales (OSS)

Our OSS sales efforts consist of five main channels:

•
General Sales – OSS maintains a website, a web store, and direct sales teams that sell directly to end-users, primarily in the U.S., Asia, and Europe, Middle East and Africa (“EMEA”) regions. This includes e-commerce sales via typical web store functionality, outbound calling and direct interaction with customers and potential customers to provide standard and unique solutions that fit their needs.

16

•
OEM Focused Sales – Our direct outside sales team, which consists of OSS employees as well as third-party manufacturer representatives, is organized to best identify, target, and develop the top potential commercial OEM and government program customers in the datacenter class, rugged AI, compute and storage space. These OEM and government programs form the largest and fastest growing parts of our business. The OSS direct sales teams interface directly with new potential customers at their facilities, live events, and virtual industry tradeshows, and present standard solutions and/or proposals for customized solutions to address such customers’ datacenter class, rugged AI needs at the edge.

•
Our Commercial Sales Team – Our commercial sales team focuses on OEM customers to whom we sell standard products and solutions or design and build customer specified systems based on OSS technology expertise that are branded with OSS or the OEM’s name and label. This includes target markets like autonomous semi-trucks, farming, medical and mining equipment deploying the latest technology. These companies, many of which are market leaders, then resell the products through their own sales channels. We actively seek this type of relationship, which is leveraged as a sales multiplier, allowing us to grow sales at a faster rate without adding as many dedicated sales resources. This tends to be a recurring sales model that lasts several years.

•
Our Government Sales Team – Our government sales team focuses on the large and growing portion of our business that provides products and systems to DOD programs, global Ministry of Defense (“MOD”) programs, government agencies, and national research laboratories. Our government sales team has the knowledge and expertise to identify major program opportunities in the emerging AI/ML, sensor processing, sensor fusion, and autonomy markets, and to provide the extensive technical and business development processes required to take these programs from concept to successful completion. This is a growing part of our business, one of our primary focuses, and provides a higher contribution of profit margin. Examples include compute and storage systems for aircraft, radar systems, command, control, intelligence, surveillance and reconnaissance platforms, and mobile command centers.

•
Channels – We have dedicated sales resources that manage our worldwide network of resellers and distributors. We sell a large breadth of standard products through these channels, which allow us to achieve global customer touch without requiring a physical presence in all geographies. The master distributors in several countries have dedicated sales expertise to capture additional OEM business, with both Fortune 500 and second tier OEM firms extending our international footprint.

Marketing Communications

Our marketing communications department is responsible for positioning OSS as an expert, thought leader, and visionary in the edge compute market. We generate expert content to support our market leading products, while also building cost effective brand/product awareness in several ways. We use traditional and non-traditional marketing communications, as well as partnerships and word of mouth, to convey the uniqueness and compelling value of our products and services. The edge compute market applications we target include AI inference applications in autonomous vehicles, medical equipment, commercial aerospace, defense/government, agriculture, and mining. Among the many channels utilized are:

•
Trade Shows – OSS participates in several live and virtual tradeshows and events during the year to generate new relationships and foster existing relationships with customers and partners. These engagements allow us to showcase our standard and custom product expertise to our target customers. These trade shows include AUSA (US Army), Sea, Air and Space (Navy), Advanced Driver Assistance Systems (autonomous vehicle), AFCEA West (military), DSEI (international military), Supercomputing, NVIDIA GTC, and Embedded World. We evaluate ROI and costs of each show on an annual basis; accordingly, participation may change from year to year.

•
Electronic Media – OSS uses various forms of electronic advertising media to market both the rugged edge products and capabilities of the Company. Electronic media includes internal direct email campaigns, such as monthly newsletters and various press releases for new products, technology developments, partnerships and significant application design wins. In addition, we use media companies relevant to our target markets to disseminate information about the Company to a larger set of potential customers. The format of the electronic advertising varies, but includes a common focus on content advertising demonstrating our market expertise with a secondary focus on brand awareness. The various electronic media formats that we utilize include, but are not limited to, search engine ads and keyword campaigns, digital ads, display ads, datasheet emails, customer use cases, e-newsletters, and text ads. Our web site is key at leveraging our

17

leadership content, positioning, and search engine optimization (“SEO”) capabilities. We periodically update our website to capture new products, as well as align with new applications and emerging markets.

•
Social Media – OSS regularly uses Facebook, LinkedIn, and X to instantly alert the Company’s followers to new events, products, services, and customer stories.

•
Publications – We periodically publish white papers, customer success stories, and other demand generation technology articles in printed and electronic periodicals and newsletters, including, but not limited to, Autonomous Vehicle International, Military Embedded Systems, Edge Industry Review, Aerospace and Defense, and Auto News.

Competition

Our core business is to provide specialized high-performance edge AI computing platforms to customers who incorporate these products into their complete solutions for rugged AI compute and storage markets. By contrast, competitors in the AI hardware market space typically provide solutions designed to meet either high performance non-rugged or low performance highly rugged applications. Although a fragmented market, there are a number of categories of potential competitors of our products.

Customer in-house design resources

Many of our larger target OEM customers, such as Raytheon and Lockheed Martin, have in-house engineering design resources, which could be used as an alternative to engaging with us. This potential competition is mitigated by the technical specialization that we have, especially in high-end and large-scale PCI Express switch fabrics and PCI Express acceleration capabilities where the option to “buy” provides a better return to “make” in an internal make/buy decision. OEMs can invest their in-house resources on value-add capabilities within their specific vertical market and outsource these horizontal technology capabilities to us. We have also developed a trusted partner relationship with many of these OEMs and have established a market reputation for technical expertise and a responsive and cost-effective engagement model. We win when our customers realize that together we can produce better products faster, and more cost-effectively than they can by themselves. This has proven to be particularly evident when customers require state-of-the-art products that are constructed of commercially available parts but need to be deployed in harsh mobile environments. This has resulted in several program wins that demonstrate our flexibility and how we can work closely with large OEM and government customers.

Major Tier 1 & 2 Mainstream Computer, GPU and Storage Vendors

These vendors offer mainstream high-performance computing platforms, including servers and storage systems that can address some applications at the edge in our target markets. Typically, they do not, however, offer the enhanced value platforms or customization capabilities that we specialize in to meet unique form factor, power, ruggedization or scale out requirements sought by OEM customers. Generally, these vendors focus on the large, air-conditioned datacenters and compete with such vendors based on price/volume, as differentiation is challenging. Our strategy is specifically designed to avoid head-to-head competition in this part of the market with this class of vendors. In some scenarios, we can provide a complementary specialized component or building block, which interfaces with one of these vendors’ mainstream products. Examples of companies in this space include NVIDIA, HP, Dell/EMC, IBM, SuperMicro, Pure Storage, and NetApp.

Vertical High Performance Compute Vendors – Military/Aerospace

In certain vertical markets, there are competitors who focus primarily on the HPC military and aerospace markets. These vendors often provide complete solutions, including both hardware and software, and some specialization in terms of form factor and ruggedization. In these markets, we provide unique capability in terms of scaling of PCI Express components over cable (copper and fiber) that can address unique requirements of specific military or government programs. Many of these competitors use older technologies or low power processors and components in these more challenging environments. We are able to differentiate ourselves from such competitors due to the fact that we deploy the latest high-performance technology, which enables us to provide superior products to potential customers in this space. We believe we have also established good relationships with prime contractors and governmental agencies (Raytheon, Sierra Nevada, Lockheed, Boeing, NASA, GVSC, L3Harris, and others), which can be important influencers or decision makers on technology selection. Competitors in this space include companies such as Mercury, Crystal, Curtiss Wright, Kontron, Trenton, Core Systems and Systel. In the past, we have been able to

18

offer the latest differentiated technology to the rugged edge, which is normally only deployed in commercial applications, well before our competitors.

Manufacturing and Operations

Currently, OSS is certified under AS9100, the international standard for aerospace, space, and defense manufacturing. This means we have demonstrated our ability to consistently provide products that meet both customer requirements and applicable government regulations or statutory requirements. AS9100 is the pinnacle of quality management systems recognized by government and aerospace companies world-wide. It also indicates that we have programs and processes in place to ensure a high level of customer satisfaction, as well as continuous improvement and risk mitigation programs that ensure we get better over time.

While OSS primarily utilizes lean principles to drive our manufacturing and assembly processes, we recognize the importance of smooth builds and strategic inventory in this current climate of sustained supply chain shortages. One of the key aspects of utilizing lean principles is our application of just-in-time principles to ensure effective ordering and utilization of inventory. This also helps optimize cash flow throughout the manufacturing cycle. Within the manufacturing process, our operations encompass three categories of “builds”:

•
Standard Builds – These are builds of standard products that are sold with little or no customization or non-standard features. These are products that are ready to be installed or integrated by the customer upon receipt.

•
Custom Builds – Custom builds involve a product built to a customer specification at our facilities. Upon receipt, the customer has a unique product that performs all the functions and has the physical dimensions that match their specifications.

•
Engineering Project Builds – We support the product development process by building models and prototypes of products. Developed by our engineering group, these prototypes can be of standard or custom products. We build these products with the intent of shipping in volume later.

We are dedicated to quality and customer satisfaction. Our continuous improvement efforts require us to review products, services, and processes with the idea that minor changes can lead to greater outcomes for our customers. Although we serve the high-end of the rugged edge computing space, we are constantly leveraging lean principles to become more efficient and drive down costs while driving up margins and quality.

Research and Development

Our ability to compete successfully in our industry is heavily dependent upon our ability to ensure a continuous and timely flow of competitive products, services, and technologies to the marketplace. We continue to develop new products and technologies and to enhance existing products that will further drive commercialization. We may also expand the range of our product offerings and intellectual property through licensing and/or acquisitions of third-party businesses and technologies.

Our intellectual property research and development is focused on the exploitation of key technologies as they evolve in the marketplace. Our product roadmap reflects new technologies for CPUs, GPUs, flash storage, and advanced PCIe switches. We design first-to-market, unique implementations targeted at the edge compute market utilizing market leading component technologies. Accordingly, our focus lies not in the capital-intensive development of silicon implementations of technologies (i.e., chips, processors, GPUs, or storage devices), but rather leverages leading-edge technologies to build first-to-market products that fully exploit those technologies to solve customer problems in challenging environments.

In addition, we will take on customer funded research and development programs where we can leverage our technology, capabilities and products to deliver a purpose designed solution that ultimately will transition into production. Strategically, this provides us with the opportunity to create an incumbent position on a platform or program that will provide recurring business opportunity as the effort moves from research and development, to initial fielding, formal production, and then sustainment and support.

Intellectual Property

19

Our primary intellectual property value emanates from the more than 600 individual design projects that we have undertaken over the decades since our founding, experience, and know how, in addition to trade secrets and copyrights. These designs are archived and cataloged; we rarely begin a new design from scratch, but rather, use our archived and cataloged designs as a starting point to efficiently provide products to our customers. In general, we maintain intellectual property rights with respect to the components of the products we design and sell so that we may continue to use them for future sales and development efforts. We also patent key innovations and solutions as and when required to protect our technical advantage.

Over the years, our team has developed and maintained expertise in high-speed signal design and analysis, electronic and mechanical packaging, PCIe-over-cable, fiber optics transmission, high-speed/density flash arrays, and integration and deployment of GPUs in compute accelerators and servers. This extensive expertise positions us to expand and rationalize our product line to meet the growing and ever-changing high-performance computing market. We believe that the expertise of our staff is a considerable asset closely related to intellectual property, and attracting and retaining highly qualified employees is essential to our business.

Markets, Seasonality, and Major Customers

Our products and services serve a global clientele consisting of multinational companies, governmental agencies, military contractors, and leading technology providers. We typically experience some level of seasonality, with lower deliveries in the first half of the year. We believe this is the result of varied government and commercial customer appropriation cycles and the timing of budgets.

In the year ended December 31, 2025, our top three customers comprised 22%, 22%, and 17% of revenue, respectively. No other customers exceeded 10% of revenue in 2025. In the year ended December 31, 2024, our top customer comprised 20% of revenue. No other customer exceeded 10% of revenue in 2024. Customer concentration figures represent continuing operations and exclude customer activity within discontinued operations.

We typically sell our products pursuant to contract supply agreements or purchase orders.

Materials and Suppliers

Although most components essential to our business are generally available from multiple sources, we believe that the loss or limited availability of certain component suppliers and manufacturing vendors could have a material adverse effect upon our business and financial condition.

Throughout 2024 and 2025, we experienced supply chain pressures, including protracted delivery dates for certain components, unavailability or limited supplies of certain products and components, increased product costs, and changes in minimum order quantities. Additionally, tariffs and changes in trade policy have impacted the costs of procuring certain components from non-US locations. We have worked with suppliers and customers to provide multiple options, including alternative sourcing of similar products.

In late 2025, a global shortage of certain memory products resulting from datacenter build-out demand led to significant increases in lead times, pricing volatility, and significant price increases. We have worked with our suppliers to secure availability of supply, including through the negotiation of long-term agreements. While we attempt to pass on component cost increases to our customers, our ability to do so is dependent upon many factors, including market conditions for the Company's products. Management is closely monitoring and managing impacts and potential impacts to our business from ongoing supply chain disruptions related to memory and other critical components. For more information, see the section titled, “Risk Factors” found in Part I, Item1A, of this Annual Report.

Human Capital Resources, Employees, and Personnel

We believe that our future success will depend, in part, on our ability to continue to attract, hire, and retain qualified personnel. To achieve this objective, we strive to provide competitive compensation, benefits, equity participation, and a success driven work environment.

As of December 31, 2025, we had approximately 57 employees, of which 56 were full-time employees and one was a part-time employee; all of our employees were domestic. Our employees include highly skilled engineers, technicians, assemblers, and support staff. We are proud of the low voluntary turnover rate of our personnel to date, as we endeavor to continue to challenge our team and encourage input and creative thinking by all. Our management team

20

strives to provide transparency to our employees through regular meetings designed to update employees on current metric driven results and future expectations. None of our employees are covered by a collective bargaining agreement or represented by a labor union. We consider our relationship with our employees to be strong.

Environmental Matters

We are ISO14001 certified. ISO14001 is the internationally recognized standard for environmental management systems.

No significant pollution or other types of hazardous emission result from our operations and it is not anticipated that our operations will be materially affected by federal, state, or local provisions concerning environmental controls. Our costs of complying with environmental, health and safety requirements have not been material.

Furthermore, we do not believe that compliance with existing or pending climate change legislation, regulation, or international treaties or accords are reasonably likely to have a material effect in the foreseeable future on our business or markets that we serve, nor on our results of operations, capital expenditures, earnings, competitive position, financial position, or any of our operations. However, we will continue to monitor emerging developments in this area.

Government Approval and Effect of Government Regulations

Because our core business is to provide specialized high-performance edge computing building blocks and platforms to OEMs who incorporate these products into their complete solutions, which they sell to end users in specific vertical markets, we do not believe that any government agency approval is required for the products and services that we provide to our customers.

We believe that our operations are in compliance with all material applicable laws and regulations and that we hold all necessary permits to operate our business in each jurisdiction in which our facilities are located. Our worldwide business activities are subject to various laws, rules, and regulations of the United States as well as of foreign governments.

Compliance with these laws, rules, and regulations has not historically had a material effect upon our capital expenditures, results of operations, or competitive position. However, governmental regulations, including but not limited to import and export law, privacy laws, customer, and trade regulations, are subject to change and interpretation and may affect our business in the future. For more information, see the section titled, “Risk Factors” found in Part I, Item1A, of this Annual Report.

Company Website

We maintain a corporate Internet website at: http://www.onestopsystems.com.

The contents of our website are not incorporated in or otherwise to be regarded as part of this Annual Report. We file reports with the SEC which are available on our website free of charge. These reports include annual reports on Form 10-K, quarterly reports on Form 10-Q, current reports on Form 8-K, “Section 16” filings on Form 3, Form 4, and Form 5, and other related filings, each of which is provided on our website as soon as reasonably practical after we electronically file such materials with or furnish them to the SEC. In addition, the SEC maintains a website (www.sec.gov) that contains reports, proxy and information statements, and other information regarding issuers that file electronically with the SEC, including the Company.

21