Blaize Holdings, Inc.

Our Company

We are a provider of purpose-built, transformative artificial intelligence (“AI”)-enabled edge computing solutions that unite software and silicon to optimize AI from the edge of our customers' networks (the “edge”) to their data centers (the “core”). We provide a full-stack programmable processor architecture suite and low-code/no-code software platform that enables AI processing solutions for high-performance computing at the edge and the core. Our solutions deliver real-time insights and decision-making capabilities at low power consumption, high efficiency, minimal size and low cost. Our mission is to enable enterprises to harness the power of AI at the edge, delivering real-time insights and decision-making capabilities with compelling speed and efficiency. With our innovative hardware and software solutions, we believe we are at the forefront of transforming industries and unlocking new possibilities in an increasingly connected and data-driven world.

We provide AI accelerated computing solutions, optimized for vision analytics applications at the edge. This is particularly relevant in the mobility and automotive sectors, where we provide solutions for Level 2 (“L2”) and Level 2+ (“L2+”) applications and collaborate with Original Equipment Manufacturer (“OEMs”) to develop advanced Level 4+ (“L4+”) capabilities, as per the Society of Automotive Engineers’ “Levels of Driving Automation.” Our fully automotive-grade compliant chip roadmap supports our customers' continued evolution and reinforces our commitment to advancing AI-powered autonomy.

With over a decade of investment in research and development, we have developed groundbreaking technologies that enable applications, such as computer vision and machine learning, to deploy and run faster and more efficiently. Our engineering team has developed a proprietary parallel computing architecture that serves as the foundation for our products, enabling high-performance AI inference with industry-leading efficiency, delivering significant value to our customers.

We are actively developing our next-generation chip, which will introduce enhanced features and capabilities while leveraging the latest advancements in semiconductor processing. Additionally, we plan to offer an intermediate hardware solution for mobility and L2 automotive customers ahead of the launch of our full automotive-grade AI processor.

With a strong focus on innovation, performance, and strategic partnerships, we believe we are well positioned to enable organizations to process and analyze their data in real time from core to edge, unlocking new business opportunities and gaining a competitive advantage in the evolving AI landscape. Our integrated approach to AI accelerated computing ensures that enterprises can harness the full potential of AI.

The Merger

On January 13, 2025, subsequent to the end of fiscal year ended December 31, 2024, the fiscal year to which this Annual Report relates, BurTech completed the previously announced business combination pursuant to that certain Agreement and Plan of Merger, dated as of December 22, 2023 (as amended on April 22, 2024, October 24, 2024, and November 21, 2024, the “Merger Agreement”), by and among BurTech, BurTech Merger Sub Inc., a Delaware corporation and a wholly owned subsidiary of BurTech (“Merger Sub”), Blaize, Inc., a Delaware corporation (“Legacy Blaize”) and for the limited purposes set forth therein, Burkhan Capital LLC, a Delaware limited liability company (“Burkhan”). As contemplated by the Merger Agreement, on January 13, 2025, Merger Sub merged with and into Legacy Blaize, with Legacy Blaize being the surviving company and a wholly owned subsidiary of BurTech (the "Merger", together with the other transactions contemplated by the Merger Agreement, the “Business Combination”). We changed our name from “BurTech Acquisition Corporation” to “Blaize Holdings, Inc.,” in connection with the closing of the Business Combination. The Merger was accounted for as a reverse merger and recapitalization and BurTech was considered the acquired company for financial statement reporting purposes.

Our Markets

Historically, most processing solutions designed for the edge (“edge applications”) were built around cloud and data center-centric infrastructure. In such solutions, data is collected at the edge and is sent back to a centralized location where the real AI processing happens. This approach is power-demanding, adds latency, causes significant data traffic, makes real-time processing difficult, and is less secure. Prompted by the proliferation of data at the edge, industry participants have been looking for ways to enable the power of AI processing where the data is, allowing for effective, real-time decision-making at the edge with greater thermal efficiency, less latency, less power consumption, less costs (including communication costs), ability to keep data local and avoid the burden of cloud and centralized infrastructure.

In recent years we have witnessed AI moving from exploration and innovation to deployment phase across most industries. Business leaders and organizations have started to understand the value, benefits and the return on investment of AI, leading to a significant growth in adoption of AI and demand for AI accelerator technologies.

We are already engaged in various industries with great potential for the widespread adoption of AI such as automotive, defense, enterprise and more. The adoption of AI in these industries is growing at a healthy rate and is poised to continue to do so in the long term, supported by strong fundamentals such as the adoption of autonomous solutions and generative AI, decentralization of data and AI for data privacy and data ownership and the drive for sustainability. We are well positioned to immediately access the burgeoning AI accelerator market today with our software and solutions, in addition to our rich roadmap, which we believe will allow us to continue to increase our Serviceable Addressable Market (“SAM”).

We expect the AI landscape to continue evolving at a rapid pace. Key trends that will impact the future of our industries include adoption of generative AI, large language models (“LLMs”), small language models (“SLMs”), vision transformers and multi-modality. The industry will continue to seek alternatives to incumbent architectures such as the graphic processing unit (“GPU”) and the central processing unit (“CPU”), prioritizing performance, total cost of ownership and ease of use. These characteristics are just as relevant and necessary for data center inference applications as they are for edge AI vision analytics. As the Blaize chip roadmap is enhanced, the inherent advantages of the Graph Streaming Processor (“GSP”) architecture and ease of programmability enable Blaize to remain relevant to market and customer evolution.

The AI market is subject to rapid technological changes, evolving industry standards, customer needs and preferences and developing regulations in the United States of America (“U.S.”) and abroad, which have focused on, among other things, safety and efficacy of AI systems, algorithmic discrimination and protection, data privacy and notice and explanation. Further, continued growth in adoption of AI-based technologies will depend, in part, on the reduction of operating and capital costs associated with AI computing, the increase in customer confidence in data protection and privacy related to centralized AI systems, software responsiveness to dynamic data, models and customer needs and transparency in application from data preparation to production.

We believe that we help facilitate such growth in adoption by providing a platform which (i) reduces the total cost of ownership of AI across a wide range of use-cases and verticals, (ii) positions the AI computing functions behind customer firewalls for added data privacy and security, (iii) automatically responds to and remediates accuracy and performance-related software issues in the system and (iv) enables and implements open standards and comprehensive auditing capabilities for customers. Blaize also continuously monitors regulatory developments related to the AI market to remain compliant with the latest requirements.

Our target markets are large, growing and sustained by strong fundamentals. The drivers for this growth are mainly attributed to an adoption drive across industries for autonomous solutions, traditional and generative AI applications, with the trend moving towards decentralization of data, data privacy and ownership, and the global appetite for sustainability.

We target two of the AI market’s highest-impact verticals that require AI accelerated computing at the edge: short cycle industrial and long cycle automotive.

Short Cycle Industrial

Our accelerator cards can be embedded in a variety of host platforms, such as AI boxes, industrial PCs, rugged computer systems, and AI server blades for deployment in racks and data centers. This versatility enables systems based on our products to be deployed across an extensive set of use-cases and market verticals, enabling applications such as computer vision, advanced video analytics and AI inference. Our products and applications can provide optimal solutions to the various needs of enterprise, commercial, industrial and defense customers, while many other current solutions in the market are fragmented, with multiple vendor platforms, leaving customers with the associated cost and complexity of making the pieces work together.

We believe we are uniquely positioned to provide value to the market due to the following:

•

Our programmable hardware and AI Studio software platform enables customers to fully and easily integrate their solutions and deploy to market faster and more efficiently.

•

The ease of use of our products allows customers to quickly deploy, manage, and extract contextual value from their analytics to meet their needs. We enable our customers to easily update their models when new data becomes available, without having to resort to scarce and expensive data science skilled resources.

•

Blaize’s dynamic hardware and software solutions enable proactive maintenance, ongoing optimization and enhanced security in tailored applications, such as in full-field asset tracking and monitoring functions related to security and monitoring.

Long Cycle Automotive

Our GSP provides an efficient way to run standard and customized neural networks in automobiles and other transportation vehicles. We expect to use our GSP-based platforms to replace power-consuming GPU and Field Programmable Gate Arrays chips, improving system efficiency advantage and a single scalable architecture across various use cases. In addition, we provide full programmability for a range of platforms, dynamic workloads, and updates. By combining low latency with low power consumption, our GSP is designed to efficiently process AI workloads at the automotive edge, either as an embedded or accelerator solution.

We consider our technology to be uniquely positioned to win in this market. The automotive value chain is quickly evolving with profits coming from the ability of automotive OEMs and Tier-1 suppliers to offer software upgrades and new features in the field. Our technology can bring considerable advantages in these applications due to the programmability of the hardware, the support for industry standards and the overall advantage in total cost of ownership.

Our technology roadmap is robust and expected to meet the ever-growing complexity and size of models for current and future demands of autonomous driving.

Our Products

Our product portfolio includes highly efficient, programmable AI processors designed for deployment across multiple verticals, including enterprise, industrial, commercial, defense, and automotive markets. Our AI computing platforms power applications such as computer vision, advanced video analytics, and AI inference, all while ensuring compatibility with industry standards for rapid integration into our customers' existing systems and workflows. Additionally, our software tools empower non-expert practitioners to deploy AI models quickly and effectively without the need for extensive coding expertise.

We have created an innovative type of AI computing accelerator, the GSP. This processor has been designed from the ground up for computer vision and machine learning applications. It is fully programmable, making it flexible and adaptable for the foreseeable future. The GSP leverages parallelism at data, instruction, task, chip and system level — streamlining execution processing and enabling low energy consumption, high performance and scalability. These advantages are demonstrated time and time again through the numerous technical evaluations that customers undertake on their own data. As the demands of edge AI inference continue to increase at a rapid pace, and customer data is evolving and changing frequently, our GSP has key advantages in enabling customer confidence.

Our hardware products are compute cards that incorporate our GSP and come in a comprehensive set of commonly used form factors. We work closely with OEM vendors to qualify these cards to add AI compute acceleration to their systems, delivering solutions across a broad spectrum of performance points and deployment verticals, from edge to data center inference. We continue to expand the range and variety of servers and edge AI boxes powered by us that meet the demands of a wide range of customer use cases.

We support industry standard Application Programming Interfaces (“APIs”) and frameworks, enabling customers to execute current models and develop new applications on our hardware. We also provide a comprehensive software suite, Blaize AI Studio (“AI Studio”). This is an intuitive and visual non-code environment intended to simplify the creation and deployment of AI models, with the goal of empowering developers and data scientists to build and optimize AI applications more efficiently. Our AI Studio features a robust set of tools and libraries, and is designed to streamline the development lifecycle, allowing organizations to bring AI-powered solutions to market more quickly.

Collaborating with system integrators, software vendors, and hardware manufacturers, we are continuously expanding our technology ecosystem to ensure seamless integration and interoperability with existing infrastructure and AI frameworks. This approach enables organizations to leverage their existing investments while adopting our innovative AI capabilities.

The current generation of our products are primarily focused on video analytics applications and AI inference at the edge. Our current technology can address the autonomous systems market for drones, robotics and automotives, including AI accelerators for cockpit, in-cabin and infotainment applications.

Our current hardware products place us in a strong position into the rapidly expanding AI accelerator market today. We have worked with many customers in testing and technical evaluations across industries, such as enterprise, defense and automotive. In addition to hardware sales, we believe we are well positioned for growth in our software sales and services through our comprehensive, pre-integrated software stack, vertically integrated solutions, AI Studio development platform, and analytics framework and toolkit.

Product Roadmap

Our product roadmap focuses on creating optimized solutions that can fulfill customer demands for AI applications at the edge. The main target markets include: enterprise, commercial, industrial, and automotive. These markets require low-power systems with highly optimized total cost of ownership, agility and flexibility on development and deployment, and a reliable supply chain. We believe our technology and targeted marketing strategy are designed to deliver products that meet and exceed those requirements.

As part of our hardware offering, we deliver solutions that include silicon chips, boards, and systems. These are complemented with full-stack software that includes a Software Development Kit (“SDK”), optimized libraries for visual analytics, AI, high-performance computing, and AI Studio, which is a complete application development and deployment platform that enables customers to create, optimize, deploy, and maintain their own solutions.

We believe the scalable and programmable nature of our product architecture positions us well to provide advantageous AI solutions for the rapidly changing demands of AI applications and allows us to differentiate our products from established and emerging competitive alternatives that focus on raw performance for specific and narrow workloads.

The existing products will enable us to capture market share for video analytics applications in verticals such as retail, defense and healthcare by embedding in devices such as surveillance and security cameras, edge servers (including edge data centers), drones, machine vision (including industrial automation) and robotics. As these markets are established, we believe we will be able to expand into adjacent verticals through updated hardware and software products.

Our Strategy

The key elements of our strategy for growth include focusing on the following areas:

•

Scaling through expansion of our current products to reach deeper into our own industry’s verticals and expanding into adjacent verticals.

•

Growing our ecosystem of hardware manufacturers and independent software vendors specializing across a variety of industry verticals and form new partnerships with them to bring innovative turn- key solutions to market.

•

Continuing to enhance the features available in AI Studio in order to target a wider community of developers, including universities and other institutes of higher education.

•

Incorporating market AI trends into the design of our next-generation System-on-a-Chip (“SoC”) to further expand our total addressable market.

Sales and Marketing

Our commercial organization utilizes research to identify key markets and sales opportunities across our target markets. Our marketing activities focus on elevating brand, product and value proposition awareness across the broader industry, with prospects in selected target verticals. This includes participation at trade events, demonstrations of Blaize technology, collaborations and co-promotions with ecosystem partners, social campaigns, competitive positioning and other marketing activities.

Our go-to-market (“GTM”) strategy focuses on creating a partnership ecosystem to develop and define a portfolio of end-to-end, market ready solutions powered by our products that addresses enterprise customer requirements in the enterprise, commercial, industrial and defense verticals.

We focus our marketing and sales efforts on the following areas:

•

Hardware powered by Blaize — we build, qualify, and quantify market-ready hardware solutions with a large number of OEMs;

•

Our unique full stack approach — we are well positioned to deliver fully integrated software and hardware solutions which include innovative developer tools. Unlike alternative hardware vendors, we believe our full-stack approach helps us to engage directly with our customers to deliver solutions that fix their problems and add value to their endeavors. The resulting intimate relationship enables Blaize to gain important insight into customers’ data and needs, which in turn are incorporated into our innovation and product roadmap;

•

Applications powered by Blaize — our team concentrates on qualifying with application partners that deliver customer-facing software and services in our focused verticals;

•

Ecosystem channel delivery — we contract with cloud, distribution, integration, and value-add resellers already supporting our target markets; and

•

Enterprise adoption — we promote market ready solutions powered by Blaize and present use cases that meet our customers’ varying budget and deployment requirements.

Government Regulations

We are subject to various laws and regulations in the U.S. and internationally, which may expose us to liability, increase costs or have other adverse effects that could harm our business. These laws and regulations include but are not limited to data privacy and data localization, copyright or similar laws, anti-spam, consumer protection, employment, and taxation. Compliance with such laws can require changes to our business practices and significant management time and effort. Additionally, as we continue to develop and improve consumer-facing products and services, and as those offerings grow in popularity, the risk that additional laws and regulations will impact our business will continue to increase.

Data Protection and Privacy

All states have adopted laws requiring notice to consumers of certain security breaches involving their personal information. In the event of a security breach, these laws may subject us to incident response, notice and remediation costs. Failure to safeguard data adequately or to destroy data (including personal information) securely could subject us to regulatory investigations, enforcement actions, and litigation under federal or state data privacy, data security, unfair practices, consumer protection, or other laws. The scope and interpretation of these laws could change, and the associated burdens and compliance costs could increase in the future.

Privacy and security laws and regulations, cross-border data transfer restrictions, data localization requirements, and other domestic or foreign laws or regulations may expose us to liability, or otherwise adversely affect our business. Laws and regulations related to data privacy and the collection, processing, and disclosure of personal information are constantly evolving. Such laws and regulations could restrict our ability to store and process personal information (in particular, our ability to use personal information for purposes such as risk or fraud avoidance, marketing, or advertising), to control our costs by using certain vendors or service providers in certain jurisdictions and could limit our ability to effectively market or advertise to interested buyers and, in general, increase the resources required to operate our business. Additionally, such laws and regulations are often inconsistent and may be subject to amendment or reinterpretation, which may cause us to incur significant costs and expend significant effort to ensure compliance.

The overall regulatory framework for AI and machine learning technologies (“AI Technologies”) is rapidly evolving as many federal, state and foreign government bodies and agencies have introduced, or are currently considering, AI laws and regulations. Additionally, existing laws and regulations may be interpreted in ways that would affect the operation of our AI Technologies. As a result, implementation standards and enforcement practices are likely to remain uncertain for the foreseeable future.

For example, in the U.S., the Biden administration issued a broad Executive Order on the Safe, Secure and Trustworthy Development and Use of Artificial Intelligence (the “2023 AI Order”) that sets out principles intended to guide AI design and deployment for the public and private sectors and signals the increase in governmental involvement and regulation over AI Technologies. The 2023 AI Order established certain new requirements for the training, testing and cybersecurity of sophisticated AI models and large-scale compute centers used to train AI models and instructed several other federal agencies to promulgate additional regulations. Already agencies such as the Department of Commerce and the Federal Trade Commission have issued proposed rules. On January 20, 2025, President Trump signed an Executive Order revoking the 2023 AI Order, and on January 23, 2025, issued an Executive Order entitled “Removing Barriers to American Leadership in Artificial Intelligence.” The new Executive Order seeks to maintain US leadership in AI innovation, “revokes certain existing AI policies and directives that act as barriers to American AI innovation,” and directs the Assistant to the President for Science and Technology, working with agency heads, to identify such policies and directives. This new Executive Order also requires the development of a new AI action plan within 180 days that must be “free from ideological bias or engineered social agendas.” While this action could indicate a less interventionist approach to AI than the prior administration, it is not clear how this new Executive Order will be implemented or whether it will affect rules issued by other agencies following the principles set forth in the 2023 AI Order.

Legislation related to AI Technologies has also been introduced at the federal level and is advancing at the state level. For example, Colorado recently passed an AI Act, which regulates the development, deployment, and use of AI systems and the California Privacy Protection Agency is currently in the process of finalizing regulations under the California Consumer Privacy Act regarding the use of automated decision-making.

The developing landscape, and the uncertain interpretation of such landscape, may affect our use of AI Technologies and our ability to provide, improve or commercialize our services, require additional compliance measures and changes to our operations and processes, result in increased compliance costs and potential increases in civil claims against us and could adversely affect our business, operations and financial condition. For more information, please refer to “Risk Factors – We may in the future, use and develop AI, machine learning and automated decision-making technologies throughout our business, which may expose us to certain regulatory and other risks that could adversely affect our results of operations and financial condition.”

Our failure to comply with these privacy, security and AI laws or regulations could expose us to significant fines and penalties imposed by regulators and has in the past and could in the future expose us to legal claims by buyers, or other relevant stakeholders. Some of these laws permit individual or class action claims for certain alleged violations, increasing the likelihood of such legal claims. Similarly, many of these laws require us to maintain an online privacy policy, terms of service, and other informational pages that disclose our practices regarding the collection, processing, and disclosure of personal information. If these disclosures contain any information that a court or regulator finds to be inaccurate, we could also be exposed to legal or regulatory liability. Any such proceedings or violations could force us to spend money in defense or settlement of these proceedings, harm our reputation, result in the imposition of monetary liability or demanding injunctive relief, divert management’s time and attention, increase our costs of doing business, and materially adversely affect our reputation.

Anti-Corruption, Sanctions and CFIUS

We are subject to the U.S. Foreign Corrupt Practices Act of 1977, as amended (“FCPA”). The FCPA prohibits corporations and individuals from engaging in improper activities to obtain or retain business or to influence a person working in an official capacity. It prohibits, among other things, providing, offering, or promising, directly or indirectly, anything of value to any foreign government official or employee, or any political party or official thereof, or candidate for political office to improperly influence, induce, obtain and retain business from, expedite or secure the performance of official acts of a routine nature, secure an improper advantage, or otherwise violate any of the Anti-Bribery laws, from such or any person. Similar laws exist in other countries, such as the UK, which restrict improper payments to persons in the public or private sector. Many countries have laws prohibiting these types of payments within the respective country. Historically, technology companies have been the target of FCPA and other anti-corruption investigations and penalties.

In addition, we are subject to U.S. and foreign laws and regulations that restrict our activities in certain countries and with certain persons. These include the economic sanctions regulations administered by the U.S. Treasury Department’s Office of Foreign Assets Control and the export control laws administered by the U.S. Commerce Department’s Bureau of Industry and Security.

Investment in the Company by non-U.S. investors may be subject to the jurisdiction of and review by the Committee on Foreign Investment in the U.S. (“CFIUS”), which has the power to rescind or mitigate transactions, or impose monetary penalties for violations of CFIUS regulations. Because we develop technology that is deemed “critical” from a U.S. national security perspective, certain investments by non-U.S. persons or entities could trigger mandatory CFIUS filings before an investment is allowed to close.

Intellectual Property

We believe our success depends in part upon our ability to develop and protect our core technology and intellectual property rights. We have built a portfolio of intellectual property rights, including issued patents and registered trademarks, copyrights, confidential technical information, and expertise in the AI and ML at the edge sectors.

We own numerous issued patents and pending patent applications. As of December 31, 2024, we held 39 patents and had 17 pending patent applications. Our patents are expected to expire between June 16, 2031 and September 24, 2044. Our two issued and allowed foreign patents include patents in two separate countries: the United Kingdom and Korea. Our five pending foreign patent applications include patent applications in India. Our patents contain a broad range of claims related to the field of AI and parallel processing architecture.

In addition to actively seeking patent protection covering inventions originating from us, from time to time, we review opportunities to acquire or in-license patents to the extent we believe such patents may be useful or relevant to our business.

We have filed patent and trademark applications in order to further secure these rights and strengthen our ability to defend against third parties who may infringe on our rights. We also rely on trade secrets, design and manufacturing know-how, and continuing technological innovations to maintain and improve our competitive position. Additionally, we protect our proprietary rights through agreements with our commercial partners, supply-chain vendors, employees, and consultants, as well as close monitoring of the developments and products in the industry.

Human Capital

As of December 31, 2024, we employed a total of approximately 232 people globally, including 56 people on a full-time basis in the U.S. and 164 people on a full-time basis internationally, either directly through our international subsidiaries or through a professional employer organization. This included approximately 166 hardware and software engineers. As the Company scales to deliver on its product roadmap, we expect to significantly expand our engineering teams. We also engage consultants and contractors to supplement our permanent workforce where the skill is specialized and temporary. None of our employees are represented by a labor union or covered by collective bargaining agreements. We believe we have strong and positive relations with our employees.

Corporate Information

Blaize Holdings, Inc. was incorporated in the State of Delaware on March 2, 2021, under the name of BurTech Acquisition Corp. In connection with the consummation of the Merger, we changed our name to Blaize Holdings, Inc. and amended and restated our certificate of incorporation (as amended, the “A&R Charter”), and amended and restated our bylaws (as amended, the “A&R Bylaws”), both effective as of January 13, 2025.

Our principal executive office is located at 4659 Golden Foothill Parkway in El Dorado Hills, California, 95762, which is where our records are kept and the principal business address for our executive officers. Our corporate telephone number is (916) 347-0050.

Available Information

We file our annual, periodic and current reports, and other required information, electronically with the SEC. The SEC maintains a website at www.sec.gov that contains reports, proxy and information statements and other information that we file with the SEC electronically. We also make available on our website at https://ir.blaize.com, free of charge, copies of these reports and other information as soon as reasonably practicable after we electronically file such material with, or furnish it to, the SEC.

We use our website, press releases, public conference calls and public webcasts as means of disclosing material non-public information and for complying with our disclosure obligations under Regulation FD. The information disclosed by the foregoing channels could be deemed to be material information. As such, we encourage investors, the media, and others to follow the channels listed above and to review the information disclosed through such channels. The contents of the websites referred to above are not incorporated into this filing.