Boost Run is building an investable case around a demanding but commercially attractive proposition: secure customer commitments, finance the infrastructure required to serve them, and turn deployments into recurring compute revenue. The bullish opportunity lies in how much larger the operating business could become as contracted capacity enters service.
The August results provide a meaningful starting point. Second-quarter revenue reached $31.1 million, up 270% year over year. Boost Run reported $1.9 billion of long-term total contract value and targeted approximately $400 million of annual recurring revenue at year-end 2026. That target remains management guidance; the contract value covers multiple years and includes infrastructure still awaiting deployment.
My bullish view rests on execution against committed demand. A growing order book can improve revenue visibility and support equipment financing. The investment becomes more compelling when those commitments produce reliable collections and cash after infrastructure costs. That is the standard by which Boost Run should be judged.
A major customer contract gives the thesis substance
The Thinking Machines Lab agreement is a concrete example. Boost Run’s May 28 SEC filing describes two orders signed on May 21 with a combined value of approximately $471.7 million over an initial 36 months. The deployment covers 5,000 NVIDIA B300 GPUs, with related storage and CPU services. The orders require payment for the contracted term regardless of actual usage, subject to contractual exceptions.
Those terms matter. If a customer uses less compute than expected, a committed rental agreement can offer the supplier more protection than selling the same capacity hour by hour. For a business purchasing expensive equipment upfront, that visibility can improve the match between customer payments and financing obligations.
The protection has limits. The filing provides termination rights for an uncured material breach. Customer creditworthiness, service delivery and contractual remedies remain relevant. A large agreement is therefore evidence of commercial demand, while successful delivery determines its economic value.
For shareholders, the most useful follow-up disclosure would show how much contracted capacity has been accepted, how much is billing, and how collections compare with the deployment schedule. An announcement establishes the opportunity. Recurring billings establish the business.
The platform can strengthen the customer relationship
Boost Run offers GPU compute alongside CPU resources, managed Kubernetes and shared storage through a console and API. That broader offering matters because an enterprise needs a working environment around its GPUs: deployment tools, secure separation between users, storage and dependable operations.
There is specific evidence behind the software layer. Partner vCluster describes a managed Kubernetes launch completed in under 45 days, using dedicated GPU resources, isolated customer control planes and Netris networking. Its case study reports that Boost Run accomplished the launch without adding platform engineering headcount. This is a partner account, rather than an independent performance audit.
The strategic opportunity is to make Boost Run useful throughout a customer’s workflow. A well-integrated environment can reduce migration friction and create reasons to renew beyond the rental price of a particular GPU. Whether that produces stronger retention or better margins still needs to emerge in operating results.
NVIDIA also lists Boost Run among its Exemplar Cloud providers. That offers technical validation relevant to procurement decisions. It does not establish exclusive access to GPUs or guarantee future orders. The competitive advantage will depend on the service customers receive and the economics Boost Run can sustain.
Deployment capacity determines the revenue timetable
Hardware sourcing has progressed: Boost Run said its $1.44 billion Dell agreement was fully allocated, while additional procurement discussions remained prospective. Allocation is an important input, but servers still need suitable power, cooling, networking and customer acceptance before a deployment can contribute as planned.
The August 10X Infrastructure Partners agreement illustrates that sequence. Its initial phase covers 20 MW of gross capacity at an existing energized site, expected to be delivered in the fourth quarter of 2026. Expansion to 111 MW across additional sites is planned through 2027. The full expansion should therefore be treated as future capacity.
Existing powered facilities can shorten the path to service compared with constructing a new campus. That is an attractive operating choice for a company whose customers want compute quickly. It also transfers part of the execution dependence to landlords and infrastructure partners, whose schedules must align with hardware deliveries.
The critical metric is revenue-producing capacity. Accessible megawatts, installed hardware and billable compute describe different stages. Investors should watch the progression between them. A delay can defer customer collections while equipment or facility obligations continue, making the timing of expansion as important as its ultimate size.
Customer prepayments can improve capital efficiency
Management’s investor presentation describes customer prepayments averaging 22% of total contract value. It also models equipment debt being repaid before a customer contract expires, allowing later contract payments to support stronger cash generation. The presentation’s financing example is explicitly hypothetical and should not be treated as the economics of an actual order.
The underlying logic is attractive. A customer deposit can reduce the capital that shareholders and lenders must provide before service begins. If financing amortizes within the revenue term, the business may retain a larger share of collections after the equipment obligation has been repaid.
Three conditions determine whether that logic works: the equipment must enter service on schedule, the customer must pay, and the remaining collections must cover operating costs as well as financing. Hardware useful life also matters. A server that remains commercially productive after its initial financing period offers a different return profile from one that requires early replacement.
Prepayments are a funding advantage accompanied by a delivery obligation. They cannot be counted again as fresh cash when the associated revenue is recognized. The strongest version of the bullish thesis combines disciplined upfront financing with durable cash generation across the full contract term.
Read the cash flow statement carefully
The first-half cash flow requires that distinction. Operating cash inflow was $114.2 million, including a $112.9 million increase in customer deposits. The investor presentation reported $77.4 million of adjusted free cash flow after its stated capital expenditure deduction. Separately, the 10-Q records $99.9 million of payments on finance lease liabilities within financing activities. Those payments are outside the presentation’s adjusted free cash flow calculation.
The headline cash flow is not a measure of cash available to shareholders after all equipment financing payments. Deposits can finance expansion effectively, but their arrival ahead of service makes a simple annualization misleading. Investors need to examine the operating, investing and financing sections together.
The accounting earnings also need context. Second-quarter net loss was approximately $75.1 million, including a $55.7 million income-tax expense associated with the company’s change in tax status. A large transaction-related tax effect can obscure operating progress, while the capital intensity of the business remains economically real.
The right question is whether a deployment earns an adequate return after power, colocation, service costs, interest, equipment payments and eventual replacement. Adjusted performance measures can help explain the business, but that full investment cycle determines what equity owners ultimately receive.
The capital raise improves flexibility and expands the share count
Boost Run announced completion of warrant redemption in August, reporting approximately $130 million in aggregate gross exercise proceeds. It disclosed 54,312,123 Class A shares and 29,533,018 Class B shares outstanding as of August 25, with no warrants remaining. That is approximately 83.85 million common shares in total, calculated from the two classes.
The capital strengthens funding flexibility and removes that warrant overhang. The corresponding dilution also belongs in the investment analysis. A valuation based only on freely traded Class A shares would miss a substantial portion of the common equity.
The June 30 accounts already included $43.4 million of warrant exercise cash. Consequently, adding the full later aggregate proceeds to the June cash balance would double-count receipts. A current liquidity estimate also needs subsequent investment and financing movements.
For valuation, I would assess enterprise value against deployed revenue and the cash expected after financing obligations, using a current share price and a reconciled capital structure. Comparing a small revenue base with a much larger future run rate can highlight growth potential; it does not by itself establish that the shares are cheap. This article makes a bullish operating case without assigning an unsupported price target.
What would strengthen or weaken the thesis
The next reporting periods should make the investment case easier to test. Progress toward the stated revenue run rate, customer acceptance of new installations and timely delivery of the initial 10X capacity would strengthen confidence. So would evidence that mature deployments generate cash after their scheduled financing payments.
I would also look for broader customer participation. Large contracts can accelerate growth, but they make individual customer payment and deployment decisions more consequential. A wider customer base would reduce that dependence and provide a better test of the platform’s commercial appeal.
The principal downside scenario is a mismatch between fixed infrastructure obligations and delayed or weaker customer receipts. Competitive pricing, faster hardware obsolescence, higher financing costs or service failures could compress project returns even if the wider AI market continues expanding. Growth is valuable only when incremental commitments earn adequate returns.
Financial reporting maturity deserves attention as well. The amended first-quarter filing disclosed material weaknesses involving IT controls, segregation of duties and accounting resources, with remediation still in progress at the time of that filing. The amendment addressed controls disclosures and did not restate the financial statements. Evidence of remediation would improve confidence as the company scales.
Why the bullish case deserves attention
Boost Run has a credible route to a materially larger business: substantial contracted demand, a defined deployment pipeline and a financing approach that brings customers into the funding equation. The managed platform adds a potential source of customer retention beyond hardware availability.
My conviction would rise as successive deployments demonstrate that the model works in cash terms. Reliable delivery can convert commercial commitments into a recurring revenue base, while disciplined financing can allow more of that revenue to accrue to shareholders over time.
The opportunity is the value created between signing a contract and collecting the cash from a successfully operated deployment. Boost Run has enough commercial evidence to warrant serious attention. Delivering against those commitments, at attractive returns on capital, is what can turn today’s bullish thesis into a durable investment case.
