20-F: SuperX AI Reports Soaring Losses Amid AI Pivot, Capital Raises

SuperX AI Technology Limited reported a significant net loss of $21.2 million for fiscal year 2025, driven by substantial investments and operational costs in its new AI infrastructure solutions business, despite successful capital raises.

Summary

• SuperX AI Technology Limited (formerly Junee Limited) rebranded and shifted its principal business focus from interior design to full-stack AI infrastructure solutions in June 2025.
• The company reported a net loss of $21,214,460 for the fiscal year ended June 30, 2025, a substantial increase from a net loss of $854,927 in FY2024 and a net profit of $38,678 in FY2023.
• Total revenue increased by 23.9% to $3,596,575 in FY2025, primarily due to $1,027,983 from the new AI server and related IT equipment sales, partially offset by a 13.1% decrease in interior design revenue.
• Gross profit decreased by 55.3% to $365,350 in FY2025, with the gross profit margin falling to 10.2% from 28.2% in FY2024.
• Operating expenses surged by 1,072.0% to $21,749,674 in FY2025, largely due to a $9,359,949 goodwill impairment, increased legal and professional fees, operating lease expenses, share-based compensation, and business development travel.
• The company successfully raised approximately $22.2 million in March 2025, $7.8 million in April 2025, $29.0 million in July/August 2025, and $4.8 million in August/September 2025 through private placements of ordinary shares and warrants.
• SuperX AI acquired 51% and then the remaining 49% equity interest in SuperX Industries (f/k/a MindEnergy AI Technology Pte. Ltd.), a Singapore-based AI software and applications developer, in April and May 2025, respectively.
• Established a regional supply center in Japan in July 2025 for in-house assembly and quality control of AI servers, and a wholly-owned U.S. subsidiary, SuperX AI Technology USA, in Nevada in September 2025.
• Entered into joint venture agreements in September and October 2025 with Enervell Power Pte. Ltd. (a Zhonhen Electric subsidiary) for SuperX Digital Power Pte. Ltd. (40% equity) and with Hongkong Chengtian Weiye Technology Company Limited for SuperX Cooltech Pte. Ltd. (35% equity) for AI-driven energy infrastructure and liquid cooling solutions.
• Acquired a 51% equity interest in MicroInference Pte. Ltd., a Singapore-incorporated company, for $3 million in October 2025, with potential to acquire up to 100% based on performance milestones.
• Identified material weaknesses in internal control over financial reporting related to accounting personnel, internal audit function, segregation of duties, and IT security/change management.

Sentiment

Score: 3

Explanation: The company's immediate financial performance is very poor, marked by a significant net loss and goodwill impairment. While the strategic pivot to AI and successful capital raises show potential for future growth, the high execution risks, unproven AI business, and identified internal control weaknesses create substantial uncertainty and risk for investors.

Positives

• Successfully pivoted to the high-growth AI infrastructure solutions market, introducing new products like the XN9160-B200, XN9160-B300 AI Servers, and All-in-One Multi-Model Server Series.
• Secured significant capital through multiple private placements in 2025, raising over $63.8 million in gross proceeds to fund AI business expansion.
• Strategic global expansion with the establishment of a regional supply center in Japan for in-house assembly and a U.S. subsidiary in Nevada to target North American clients and tech partnerships.
• Formed key joint ventures with established industry players like Zhonhen Electric and Chengtian Weiye for AI-driven energy infrastructure and liquid cooling solutions, enhancing ecosystem partnerships.
• Acquired majority and then full interest in SuperX Industries (f/k/a MindEnergy AI Technology Pte. Ltd.) and a 51% stake in MicroInference Pte. Ltd., strengthening AI software and application development capabilities.
• Experienced management team in the traditional interior design business with long-standing client and subcontractor relationships.
• Maintained adequate insurance coverage consistent with industry norms in Hong Kong and Singapore.

Negatives

• Reported a substantial net loss of $21,214,460 in FY2025, a significant deterioration from previous periods, indicating high costs associated with the business pivot.
• Experienced a 55.3% decrease in gross profit and an 18.0% decline in overall gross profit margin in FY2025, largely due to increased subcontracting and material costs in interior design and lower margins in the new AI business.
• Operating expenses surged by 1,072.0% in FY2025, primarily due to a $9,359,949 goodwill impairment, increased legal and professional fees, share-based compensation, and business development expenses.
• The interior design business experienced a 13.1% decrease in revenue in FY2025 due to an economic recession in Hong Kong, increased inflation, and a downturn in the real estate market.
• High customer concentration in the interior design business, with the top five customers contributing 84.9% of total revenue in FY2025, posing a significant risk if any major customer is lost.
• Identified several material weaknesses in internal control over financial reporting, including insufficient accounting personnel, lack of a functional internal audit department, inadequate segregation of duties, and deficiencies in IT security and change management.
• The AI server and infrastructure solutions business is at an early stage of development with an unproven track record and has not yet generated material revenue, despite significant investments.
• The company does not intend to pay further dividends in the foreseeable future, retaining earnings for business expansion.

Risks

• Reliance on dividends from operating subsidiaries, with potential future restrictions by the PRC government on cash transfers out of Hong Kong, which could materially decrease the value of ordinary shares.
• Exposure to evolving PRC laws and regulations, including M&A Rules and data security laws, which could impact Hong Kong-based operations and the ability to offer securities overseas.
• Potential for the Chinese government to extend oversight and control over Hong Kong-based issuers, limiting the ability to offer shares and causing value decline.
• High competition in the fragmented Hong Kong interior design industry, with potential for new entrants offering lower prices or higher quality services.
• Profitability of the interior design business is project-based and dependent on negotiated terms, with past performance not indicative of future results.
• Significant increases in costs of sales for interior design could decrease gross profit margins, as the company does not have long-term supply contracts.
• Dependence on a few major customers in the interior design business, with loss of any potentially having a significant negative impact.
• Reliance on core management personnel in the interior design business, with loss of key personnel being detrimental due to fierce competition for talent.
• Vulnerability of the interior design business to conditions of the Hong Kong economy and property market, including slowdowns in economic growth or pessimistic outlooks.
• Dependence on materials suppliers and subcontractors for interior design projects, with risks of unavailability, lower quality work, higher costs, or delays.
• Exposure to liquidity risk and credit risk of customers in the interior design business, with potential for delayed or refused payments.
• Risks related to high energy and cooling demands of AI data centers, with increases in utility costs or environmental restrictions adversely affecting operations.
• Dependence on specialized hardware like GPUs, with risks of shortages, vendor concentration, or rapid technological change rendering infrastructure obsolete.
• Increased targeting of AI data centers by sophisticated cyberattacks leveraging AI, potentially leading to significant liability or reputational harm.
• Risks of employees or contractors misusing AI tools, resulting in inadvertent disclosure of confidential or customer data.
• Substantial costs or restrictions from new and evolving government laws and regulations relating to AI, data usage, energy consumption, and environmental impact.
• Reputational harm if AI workloads supported produce biased, inaccurate, or harmful outputs, even if not directly responsible for model development.
• Failure to realize anticipated benefits of significant capital expenditures on AI-ready data centers if demand slows, key customers reduce reliance, or competition intensifies.
• Inability to efficiently enhance the platform, develop new solutions, and respond to rapidly changing technology, industry standards, and customer needs in the AI sector.
• Uncertainty in the broader adoption, use, and commercialization of AI technology, and the rapid pace of developments in the AI field, impacting demand for solutions.
• The AI server and infrastructure solutions business is at an early stage of development with no assurance of success or profitability.
• Exposure to export controls, trade restrictions, and tariffs on AI-related components and servers, which could adversely affect business and supply chain.
• Client actions (cancellations, quantity changes, delays) impacting production scheduling, capital expenditures, and capacity efficiency.
• Integration risks among complex components in AI servers leading to underperformance, failures, or delays, harming reputation and results.
• Erosion of competitive advantage if proprietary AI designs are reverse engineered or leaked, especially through outsourced manufacturing partners.
• Shortages of components or price increases in critical materials interrupting operations, reducing profit, and increasing inventory costs.
• Licensing constraints on third-party chips, firmware, and software limiting deployment, increasing costs, or disrupting product offerings.
• Product liability claims from defects, failures, or malfunctions in AI server and infrastructure solutions, leading to significant costs and reputational harm.
• Geopolitical tensions, trade restrictions, and supply chain disruptions adversely affecting component sourcing, manufacturing, and delivery of AI solutions.
• The AI server and infrastructure solutions business is highly capital intensive, with significant upfront investments potentially straining liquidity.
• Execution risks in establishing and scaling in-house assembly facilities, such as the Japan supply center, leading to delays, cost overruns, or operational issues.
• Challenges in introducing new business models or programs requiring new competencies, affecting operations and financial results.
• Increasing commoditization of AI servers and intensifying competition leading to pricing pressure, reduced margins, and loss of market share.
• Aggressive pricing by large incumbents and increases in component, energy, or labor costs eroding margins, with inability to pass costs to clients.
• Inability to effectively implement business plans to achieve future growth, particularly in the new AI sector.
• Rapid innovation and component obsolescence in the semiconductor industry rendering current designs outdated and requiring costly redesigns.
• Defects in products (from outsourced manufacturing, suppliers, or internal assembly) damaging reputation, increasing costs, and adversely affecting results.
• Competition with numerous other diversified service providers and design firms in the industry.
• Inability to maintain engineering, technological, and production expertise in dynamic markets.
• Disruptions to information systems, including security breaches, data losses, or outages, adversely affecting operations.
• Failure to recruit, train, and retain skilled personnel for assembly and manufacturing operations.
• Operations and products subject to environmental, energy, and emissions regulations, increasing costs and affecting competitiveness.
• Rapid technological shifts, including new computing paradigms (optical, quantum, neuromorphic), reducing demand for current AI solutions and requiring reinvestment.
• Use of third-party designs, architectures, or reference designs exposing the company to intellectual property infringement claims.
• Unsuccessful expansion and operation of business in Japan, affecting results of operations.
• Fluctuation of the Japanese yen against foreign currencies materially affecting results of operations.
• Uncertainty regarding classification as a Singapore tax resident, potentially leading to additional income tax.
• Difficulties for U.S. investors to enforce judgments obtained in the United States against the company or its non-U.S. directors/officers.
• Restrictions under Singapore laws on the ability of Singapore subsidiaries to distribute dividends.
• Adverse material changes to the Singapore market (economic recession, pandemic) affecting business, results, and financial condition.
• Subject to Singapore laws, which differ from U.S. laws, potentially imposing more restrictions.
• Risks associated with operating in the rapidly evolving Southeast Asia region, including economic, political, and social conditions.
• Adverse changes in government regulations in Singapore materially affecting operations and financial condition.
• Exemption from certain U.S. domestic public company provisions as a foreign private issuer, potentially affording less protection to shareholders.
• Potential delisting from Nasdaq under the HFCA Act if PCAOB cannot inspect auditors for two consecutive years.
• Additional and more stringent criteria applied to emerging market companies by SEC and NASDAQ, adding uncertainties to business and share price.
• Substantial future sales of ordinary shares or anticipation of such sales causing share price decline.
• Volatility in ordinary share price potentially leading to securities litigation.
• Existing shareholders selling shares under Rule 144 potentially impacting trading price.
• Risk of being deemed a passive foreign investment company (PFIC) for U.S. federal income tax purposes, resulting in adverse tax consequences for U.S. holders.

Future Outlook

The company expects revenue from full-stack AI infrastructure solutions services to increase in upcoming fiscal years. It plans to expand its interior design client base and service capabilities, including establishing in-house teams for specialized electrical fit-out work and repair/maintenance, and developing a customer relationship management (CRM) module. In AI, the strategy is to shift to a 'one-stop-shop' model for integrated AI infrastructure, expand globally with a U.S. hub and Japan supply center, deepen ecosystem partnerships, and evaluate opportunities to move up the value chain with GPU-as-a-Service (GPUaaS) offerings.

Management Comments

• Management monitors the cash position of each entity within the organization regularly and prepares budgets on a monthly basis to ensure adequate liquidity.
• Management believes that current cash and cash flows, bank loans, and net proceeds from IPO and private placements will be sufficient to meet working capital needs for the next 12 months.
• Management believes that no significant credit risk exists with bank deposits as these banks have high credit quality and no losses have been incurred.
• Management believes that its contract acceptance, billing, and collection policies are adequate to minimize material credit risk to accounts receivable.
• Management believes that it is more likely than not that OPS HK will be able to fully utilize its deferred tax assets related to the net operating loss carry-forward in Hong Kong.
• Management is of the opinion that the exposure of change in functional currency is minimal as HK$ is pledged to US$.

Industry Context

The AI infrastructure solutions industry is experiencing rapid growth, driven by demand for AI training and inference, especially for generative AI and large language models. The market is shifting towards AI Factory architectures, emphasizing GPU-centric servers, energy-efficient 800VDC power, liquid cooling, and high-speed interconnects. This market remains unsaturated with surging demand from enterprises and cloud providers. In contrast, the Hong Kong interior design and fit-out market is highly fragmented and competitive, with demand affected by the local economy and property market conditions, which have seen a downturn due to increased inflation and monetary policy tightening.

Comparison to Industry Standards

• The company's XN9160-B200 AI Server, powered by NVIDIA's Blackwell GPUs, is positioned as a leader in cost-effective, scalable AI infrastructure, outperforming rivals in speed, density, and sustainability for enterprise and research applications.
• The XN9160-B300 AI Server leverages NVIDIA HGX B300 module with 8 NVIDIA Blackwell B300 GPUs, offering 2,304GB of unified HBM3E memory, which is essential for large models and high-concurrency generative AI, providing 50% more NVFP4 compute and HBM capacity per chip compared to previous generations (e.g., H100 series mentioned in the context of XN9160-B200).
• The All-in-One Multi-Model Server Series aims to streamline AI model deployment from weeks to hours, suggesting a competitive advantage in deployment efficiency for enterprise-grade solutions.
• The interior design market in Hong Kong is highly fragmented with approximately a thousand companies, indicating intense competition where the company's OPS HK subsidiary operates.
• The company's gross profit margin of 10.2% in FY2025 is significantly lower than its previous year's 28.2%, which could be below industry averages for either interior design or emerging AI hardware, especially given the high R&D and operational ramp-up costs.

Management Changes

Corporate Governance

Legal Proceedings

• Not involved in any material litigation, claim, administrative action, or arbitration during the fiscal years ended June 30, 2025, 2024, and 2023, and as of the date of this annual report.

Related Party Transactions

• Rental fees paid to Sannogi Holdings Limited (controlled by directors of OPS HK) were $9,705 in FY2025, $9,668 in FY2024, and $9,643 in FY2023.
• Rental fees paid to Sannogi Engineering Consultants Limited (controlled by directors of OPS HK) were $78,100 in FY2025, $77,801 in FY2024, and $77,602 in FY2023.
• Management fee income received from Kin On Engineering (International) Limited (controlled by Alpha Generator Limited, a shareholder of OPS Holdings Limited) was nil in FY2025, $138,107 in FY2024, and $183,673 in FY2023. Services ceased in April 2024.
• Amounts due from related parties (Kin On Engineering (International) Limited, Alpha Generator Limited, Rich Merchant Limited, OPS Holdings Limited) totaled $70,674 as of June 30, 2025, and $65,946 as of June 30, 2024. These are unsecured, interest-free, and repayable on demand.
• Prepayments to related parties (Sannogi Holdings Limited, Sannogi Engineering Consultants Limited) for office rent totaled $7,261 as of June 30, 2025, and $7,298 as of June 30, 2024.
• Rental deposits paid to related parties (Sannogi Holdings Limited, Sannogi Engineering Consultants Limited) totaled $21,783 as of June 30, 2025, and $21,895 as of June 30, 2024.
• Amounts due to related parties (Ms. Zhang Jiaying, Mr. Chan Yuk Ki, Francis, Mr. So Siu Lai, Mr. Yip Sai Kit, Dicky) totaled $173,885 as of June 30, 2025, and $46 as of June 30, 2024. These are unsecured, interest-free, and repayable on demand.

Stakeholder Impact

• Shareholders: Significant dilution from multiple private placements and a substantial net loss in FY2025. The strategic pivot to AI offers potential long-term growth but carries high risks. Lack of dividends in the foreseeable future.
• Employees: Increased headcount in AI infrastructure solutions, with new equity incentive plans (2024 and 2025) to attract and retain talent. Resignations of some high-paid staff in the interior design segment.
• Customers (Interior Design): Decline in demand due to economic recession in Hong Kong and property market downturn. High customer concentration poses risk of business loss.
• Customers (AI Infrastructure): New product offerings (XN9160-B200, XN9160-B300, All-in-One Multi-Model Server Series) aim to provide cutting-edge solutions, but the business is unproven and subject to rapid technological change and competition.
• Suppliers/Subcontractors: Increased subcontracting and material costs in interior design. Reliance on a global network of suppliers for AI components, with risks of shortages and geopolitical disruptions.
• Creditors: Increased bank and other borrowings, but also significant cash from capital raises. The substantial net loss and goodwill impairment could raise concerns about financial health, though working capital remains positive.
• Regulatory Bodies: Identified material weaknesses in internal controls require remediation. Compliance with evolving AI, data, and environmental regulations will be crucial.

Next Steps

• Complete the remaining portion of the August 2025 private placement, subject to payment and closing conditions.
• Continue to implement measures to improve internal control over financial reporting, including hiring qualified accounting staff, establishing a financial and system control framework, and strengthening corporate governance.
• Focus on executing the sales and marketing strategy for interior design services, utilizing digital and traditional media.
• Expand interior design service capacity by establishing in-house teams for specialized electrical fit-out work and repair/maintenance.
• Develop a customer relationship management (CRM) module with a web interface and mobile app for interior design clients.
• Innovate, expand, and transform AI technology by adopting full-stack solution integration, global expansion and localization, deepening ecosystem partnerships, and moving up the value chain with cloud services (GPUaaS).
• Complete the formation and investment in SuperX Digital Power Pte. Ltd. and SuperX Cooltech Pte. Ltd. joint ventures.
• Complete the closing conditions and acquisition of 51% equity interest in MicroInference Pte. Ltd., with potential for full acquisition based on performance milestones.
• Monitor the situation regarding COVID-19 and other potential pandemics throughout 2025 and beyond, taking a conservative approach to cost budgeting.

Key Dates

Recommendation

The company is undergoing a significant and costly strategic pivot from a declining interior design business to the high-growth but highly competitive AI infrastructure sector. While the successful capital raises and strategic acquisitions/JVs provide necessary funding and capabilities for this transition, the immediate financial results are severely negative, marked by a substantial net loss and goodwill impairment. The identified material weaknesses in internal controls also present a significant operational risk. The long-term potential in AI is attractive, but the business is unproven, capital-intensive, and faces intense competition and regulatory uncertainties. Given the high risk associated with this transition and the current financial performance, a 'hold' recommendation is appropriate for existing investors to observe the execution of the AI strategy and remediation of internal control issues. New investors should exercise extreme caution due to the speculative nature of the pivot and current financial instability.

Keywords

AI infrastructure, AI servers, Data centers, GPU computing, Artificial intelligence, SEC filing, Financial results, Corporate governance, Capital raise, Private placement, Hong Kong interior design, Singapore operations, Japan supply center, Risk factors, Nasdaq Capital Market, SuperX AI Technology Limited, XN9160-B200, XN9160-B300, Multi-Model Server Series, NVIDIA Blackwell, Liquid cooling, HVDC solutions, Joint ventures, MicroInference, Internal controls, Goodwill impairment