The hook
1/10
A 334% jump that changed the story
E2E Networks’ Q1 FY27 operational revenue surged 334% year over year to ₹156.8 crore, swinging the company from a FY26 net loss of ₹15.6 crore to a quarterly PAT of ₹43.9 crore.
E2E Networks’ Q1 FY27 operational revenue surged 334% year over year to ₹156.8 crore, swinging the company from a FY26 net loss of ₹15.6 crore to a quarterly PAT of ₹43.9 crore.
E2E Networks began in 2009 as a contract-less CPU cloud provider for Indian SMEs. It now operates a large independent NVIDIA GPU fleet—A100, H100, H200 and Blackwell B200—and runs its own TIR MLOps platform.
India’s DPDP Act and RBI localization mandates push regulated financial, health and government AI workloads onto domestic infrastructure. E2E’s data centers in Noida, Mumbai and Chennai, with INR billing and no cross-border telemetry, fit that compliance need.
E2E scaled from about 3,900 GPUs in late 2025 to over 5,100 operational GPUs by mid-2026. Flagship clusters use InfiniBand NDR and NVSwitch, reaching 3.2 Tbps aggregate bandwidth to support distributed training rather than isolated VMs.
Larsen & Toubro agreed to acquire 21% of E2E Networks: ₹1,079.27 crore primary infusion for 15%, plus ₹327.75 crore to buy 6% from promoters. The alliance brings two board seats and priority space in L&T’s Chennai Vyoma data center.
In August 2026 E2E signed a binding term sheet with an India-based sovereign AI entity, worth about ₹1,000 crore, to supply NVIDIA Blackwell cloud GPUs through June 2029. The announcement helped the stock hit a 5% upper circuit.
Operational revenue hit ₹156.8 crore. EBITDA reached ₹117.9 crore at a 75.2% margin, and PAT swung to ₹43.9 crore from a ₹2.8 crore loss a year earlier. Exit MRR rose from ₹37.4 crore in March 2026 to ₹71.8 crore in June 2026.
FY26 PAT showed a ₹15.6 crore loss despite ₹126.3 crore EBITDA because depreciation jumped 182% to ₹169.3 crore. Q1 FY27 proved the model: once fixed GPU depreciation is covered, incremental leasing revenue flows quickly to the bottom line.
Building a GPU fleet demands massive upfront capex. Competitors like Yotta plan over 16,000 GPUs, and global hyperscalers still offer broader enterprise ecosystems. E2E must maintain high utilization on its B200 clusters while absorbing heavy, front-loaded depreciation.
E2E Networks no longer looks like a niche cloud provider. Sovereign contracts, L&T capital, and data-localization rules make it a foundational layer for Indian AI. Read the full article to explore the hardware, platform and financial mechanics behind the pivot.
Read the full analysis →