Gorilla Technology Group at Water Tower Research Virtual Insights Conference: AI push

Gorilla Technology outlined next-fiscal-year revenue guidance of $450 million-$500 million, versus $101 million in 2025 and more than $78 million in 1H 2026, underpinned by active sovereign-AI and GPU infrastructure deployments. The company’s India Yotta program covers 25,856 advanced GPUs and represents roughly $3.0 billion-$3.1 billion of potential revenue, while its total pipeline has expanded beyond $10 billion. Management targets more than 50% gross margin and more than 20% EBITDA margin next year, versus a 17% trailing gross margin, making delivery schedules, FX normalization and project financing key execution risks.

Analysis

GRRR’s equity value is being asked to capitalize a business model that has shifted from asset-light software into capital-intensive infrastructure before the financing structure, customer payment terms, and asset ownership economics are independently visible. The key valuation variable is not the reported pipeline or even hardware delivery; it is the conversion of installed GPU capacity into accepted, billable, cash-collected service revenue without recourse to parent-level equity. A small-company multiple can expand sharply if this is demonstrated, but the same model creates a reflexive downside: any project-finance delay raises working-capital needs, which increases dilution risk and compresses the credibility premium.

The near-term catalyst path is unusually binary. Over the next 30-90 days, installation acceptance, invoicing and disclosed collections should determine whether revenue recognition catches up to management’s run-rate narrative; quarterly cash flow, receivables aging, contract liabilities, debt covenants and capex commitments matter more than top-line guidance. The claimed margin step-up requires the company to be acting principally as a recurring infrastructure operator rather than a low-margin equipment pass-through; failure to show a material gross-margin inflection in the next reported periods would undermine the central rerating thesis.

The non-obvious read-through is limited for NVDA: a large GPU deployment can validate regional sovereign-AI demand, but it is likely already embedded in hyperscaler/OEM demand forecasts and does not establish incremental NVDA revenue without confirmation that orders are additive and not merely routed through Supermicro. SMCI is the more direct operational counterparty, but GRRR concentration also makes it vulnerable to supply allocation, installation bottlenecks, power availability, and customer acceptance delays. Consensus enthusiasm appears to underprice financing and execution risk, while bears may underprice the equity convexity if cash conversion is verified.

Key Decisions for Investors

• Maintain GRRR as a high-volatility watchlist name rather than a core AI allocation until the next filing demonstrates project-level debt funding, positive operating cash conversion, and receivables consistent with recognized infrastructure revenue. Do not underwrite management’s longer-dated revenue ambition before these disclosures.
• For a tactical position, initiate only a small long GRRR after verified customer acceptance and billing milestones are disclosed; target a 1-3 month catalyst window. Size for total-loss-style microcap risk, and exit if reported gross margin fails to improve materially from the current low-teens profile or if corporate equity/convertible financing is required to fund deployments.
• Avoid using NVDA as a direct sympathy long on this development. Treat any confirmed incremental GPU order evidence as a modest demand datapoint only; NVDA’s risk/reward remains driven by broader cloud capex, export controls, and supply rather than a single project counterparty.
• Monitor SMCI for order, backlog, and working-capital commentary tied to regional GPU infrastructure deployments. A confirmed shipment-and-acceptance cadence would be incrementally supportive, but delays would be more informative for GRRR than materially thesis-changing for SMCI.
• Set an event alert for GRRR’s next earnings release: require reconciliation between revenue guidance, deployed assets, project financing, capex, receivables, and operating cash flow. A widening gap between revenue recognition and cash collections is the clearest falsifier of the bullish infrastructure-service thesis.