8-K: Boost Run Secures $471.7M GPU Infrastructure Contract

Boost Run Inc. has entered into a 36-month service agreement with Thinking Machines Lab Inc. for the rental of 5,000 NVIDIA B300 GPU servers.

Summary

• Boost Run Inc. entered into a Master Service Agreement (MSA) and two Order Forms with Thinking Machines Lab Inc. on May 21, 2026.
• The contract involves the rental of 5,000 NVIDIA B300 GPU servers, along with related network storage and CPU node services.
• The agreement spans an initial term of 36 months.
• The total contract value is approximately $471.7 million.
• The customer is obligated to pay for the full duration of the term, regardless of actual usage, with fees being non-refundable.

Sentiment

Score: 8

Explanation: StockSavvy.ai views this as a highly positive development, as it secures a massive, long-term, non-cancelable revenue stream that validates the company's business model and infrastructure capabilities.

Positives

• Secured a significant long-term contract valued at approximately $471.7 million.
• Provides 36 months of guaranteed revenue from a single customer.
• The contract is non-cancelable, ensuring revenue stability for the specified term.
• Demonstrates strong demand for the company's high-performance managed GPU compute and cloud infrastructure services.

Negatives

• Concentration risk: A substantial portion of future revenue is tied to a single customer, Thinking Machines Lab Inc.
• The company is subject to strict service level agreements (SLA) with a 99.5% uptime requirement, with potential financial penalties for failure to meet these metrics.

Risks

• Potential for service level credits if uptime falls below 99.5%, which could impact margins.
• Operational risks associated with managing and maintaining 5,000 high-performance GPU servers.
• Dependency on third-party data center facilities and infrastructure.
• Risk of material breach if security requirements (Exhibit D) are not met, potentially leading to contract termination.
• The customer has the right to terminate for cause if the company fails to cure a material breach within 30 days.

Future Outlook

The company expects to deploy 5,000 NVIDIA B300 GPUs across its data center facilities to fulfill the 36-month service commitment, providing a stable revenue stream through May 2029.

Management Comments

• The agreement establishes the overarching terms and conditions for the customer relationship, with individual server rentals documented through non-cancelable Order Forms.

Industry Context

StockSavvy.ai notes that this deal highlights the intense demand for high-end GPU capacity (specifically NVIDIA B300s) driven by the AI and machine learning sector, positioning Boost Run as a critical infrastructure provider in the current compute-constrained market.

Comparison to Industry Standards

• The 36-month non-cancelable term is consistent with enterprise-grade cloud infrastructure contracts.
• The 99.5% SLA metric is standard for bare metal and high-performance computing providers.
• The contract structure mirrors typical hyperscale cloud provider agreements regarding liability limitations and security compliance requirements.

Stakeholder Impact

• Shareholders: Positive impact due to significant revenue visibility and growth.
• Customers: Enhanced access to high-performance compute resources.
• Employees: Increased operational requirements to support large-scale deployment.

Next Steps

• Deployment of 5,000 NVIDIA B300 GPUs.
• Ongoing provision of managed GPU compute and cloud infrastructure services.
• Monthly invoicing and revenue recognition over the 36-month term.

Key Dates

Recommendation

The securing of a $471.7 million contract provides significant financial stability and growth validation, making the stock attractive for investors seeking exposure to the AI infrastructure build-out.

Keywords

GPU compute, cloud infrastructure, NVIDIA B300, data center, managed services, bare metal, high-performance computing