E2E Networks Q1 FY27: B200 goes live, revenue scales sharply, margins expand

E2E Networks opened FY27 with a step-change quarter. Revenue from operations for Q1 FY27 rose to INR 1,568 million, up 334.3% year-on-year and 64.0% quarter-on-quarter. The company also reported EBITDA of INR 1,179 million, translating into an EBITDA margin of 75.2%. Profit after tax came in at INR 439 million, with a PAT margin of 28.0%.

Management linked the quarter’s scale-up to two core drivers: additional capacity going live and higher utilisation across the GPU fleet. In the earnings call, the company stated that pricing had only a moderate impact in the period, and that the larger contribution came from capacity additions and strong utilisation.

The operational trigger: Blackwell B200 capacity enters revenue

The most important operating milestone in the quarter was the go-live of the NVIDIA B200 cluster. Management stated that 1,024 Blackwell GPUs went online and were put on revenue during the quarter. In the investor presentation, E2E also reported that it had approximately 5,100 cloud GPUs live at Q1 FY27 exit, including 1,024 B200.

The company framed this as part of a broader capacity trajectory.

E2E’s presentation also showed a multi-year ramp in monthly revenue run-rate, from INR 51 million in Jun-22 to INR 718 million in Jun-26. This four-year expansion is being positioned as evidence that the company has steadily built demand while increasing supply.

Profitability: operating leverage and utilisation drive margin expansion

Quarterly costs remained controlled relative to revenue growth. Total expenses in Q1 FY27 were INR 389 million versus INR 375 million in Q4 FY26, even as revenue increased materially. This operating leverage is visible in the margin line, with EBITDA margin expanding by 1,450 basis points quarter-on-quarter.

Below EBITDA, depreciation and finance cost were significant in the quarter, reflecting the asset-heavy nature of GPU infrastructure.

Depreciation in Q1 FY27 was INR 606 million, up from INR 513 million in Q4 FY26. Finance cost rose to INR 101 million from INR 37 million in Q4 FY26. On the call, the company attributed the rise in interest to loans taken to fund the first lot of B200.

This matters for investors because the income statement now reflects both the upside and the structural costs of scaling GPUs. While utilisation can lift margins sharply when demand is tight, depreciation and interest costs can weigh on profitability if future utilisation softens or pricing normalises.

Strategy positioning: sovereign AI and the in-house platform stack

E2E’s investor presentation and management commentary leaned heavily on the theme of sovereign AI. The company’s stated positioning is that AI infrastructure is being reclassified from a utility to a strategic national asset, and that sovereignty increasingly depends on who controls the models, data, and compute.

Operationally, E2E is building an integrated stack.

First is infrastructure: NVIDIA GPUs from Hopper to Blackwell deployed in India, supported by data centres in NCR and Chennai.

Second is the public platform layer: TIR and Jarvislabs.ai. The company describes TIR as covering the workflow from notebook experimentation to production deployments, with features spanning GPU notebooks, model fine-tuning, RAG pipelines, and model endpoints through common inference runtimes. Jarvislabs.ai is positioned as a developer-first GPU cloud with per-minute billing, pause and resume, and CLI and SDK workflows. The presentation states Jarvislabs.ai has been part of the platform since December 2025.

Third is go-to-market: a combination of self-serve portal, enterprise sales, partners, and government enablement.

The company also referenced a sovereign or private platform deployment model for regulated or data-residency-bound customers, where E2E’s cloud software can be deployed to meet specific compliance requirements.

In the earnings call, management connected this strategy to the progress of open-weight models. The company argued that open models are now close to frontier models, allowing enterprises to fine-tune and deploy within their own boundaries, reducing reliance on external API access and control.

Capital allocation and structure: capex history, funding mix, and new subsidiaries

The presentation provided a capex trend that highlights how sharply investments increased in recent years.

Capex (INR Mn) was presented as 249 in FY22, 350 in FY23, 1,853 in FY24, 8,700 in FY25, 6,962 in FY26, and 177 in Q1 FY27. The company noted that FY26 includes CWIP of INR 5,334 million for GPUs deployed in May 2026.

E2E also disclosed proceeds and utilisation from preferential equity issues.

On the call, management disclosed that total loan outstanding stood at broadly INR 450 crores as of now, and that it would increase in the near term with the next lot. The company did not quantify peak debt for the year.

From a corporate structuring standpoint, the company highlighted incorporation of a wholly owned subsidiary, described in the call as SovCloud, focused on infrastructure build-out and large-scale cluster contracting. It also referenced a Delaware entity intended to support international expansion, alliance management, and serving customers outside India.

What investors will likely track next

The Q1 FY27 performance reflects a classic infrastructure scaling moment. Demand is strong and utilisation is high, so incremental capacity can translate quickly into revenue and margin expansion. The sustainability of this outcome will depend on how consistently E2E can add capacity, maintain utilisation, and manage the capital intensity of expansion.

Key follow-through points discussed by management include additional B200 deliveries expected in the next couple of months, plans to expand Blackwell capacity further, and early planning for next-generation architectures such as B300 and Vera Rubin. Management did not provide quantified timelines beyond near-term delivery expectations.

The quarter also reinforced that E2E is still building its disclosure framework. Management declined to provide detailed splits such as training versus inference contribution, or customer segment mix, and reiterated that it does not provide MRR guidance.

Closing takeaway

Q1 FY27 was a milestone quarter for E2E Networks, driven by Blackwell B200 capacity entering revenue and strong utilisation across the GPU fleet. Financial performance showed substantial operating leverage, with EBITDA margins rising to 75.2% and PAT turning strongly positive.

The next phase will test execution in a more capital-intensive setting, with higher debt and depreciation now embedded in the model. If E2E continues to bring capacity online on schedule and sustains utilisation, the company’s sovereign AI positioning and platform-led approach could remain a central part of its growth narrative through FY27.