E2E Networks Ltd Q2FY26: Cloud GPU King Turns Loss-Making Again While Flexing 2,048 H200s and ₹1,000 Cr War Chest
1. At a Glance
India’s homegrown AI-cloud poster boy, E2E Networks Ltd, just dropped its Q2FY26 numbers — and boy, the numbers are as electrifying as their GPUs. The company clocked ₹43.8 crore in sales (down 7.9% QoQ), while net profit took a nosedive into the negative zone with a ₹13.5 crore loss (a 211% decline). This after proudly showing off its H100s and H200s like an Indian kid showing his parents a U.S. visa stamp.
At a market cap of ₹5,657 crore and trading at ₹2,812/share, E2E is valued at a mind-bending P/E of 636 — because, apparently, investors think “AI Cloud” is the new Bitcoin. The company’s ROE (5.7%) and ROCE (8%) whisper caution, but the GPU mania screams louder.
Oh, and did we mention? Promoters have pledged 60.4% of their holding — the financial equivalent of saying “trust me bro” while juggling ₹1,000 crore fundraising plans.
So what happens when India’s first GPU cloud darling expands from 4.2 MW to 10.2 MW, ties up with L&T, launches a sovereign AI cloud, and still loses money? Welcome to the E2E paradox.
2. Introduction
If Silicon Valley had a child with Delhi NCR’s jugaad culture, you’d get E2E Networks Ltd — India’s own AI-first cloud company that loves GPUs more than tech bros love LinkedIn.
Started off as a scrappy cloud player and now wearing the “AI infrastructure pioneer” tag, E2E’s journey reads like an underdog script: two data centers in Delhi, one in Mumbai, another rising in Chennai, and a mission to make India’s compute stack as sovereign as an RBI governor’s frown.
But this quarter, things got dramatic. After flexing 2,048 H200s across India, the company hit a technical snag — not in hardware, but in the P&L. Revenues slipped QoQ, losses widened, and depreciation ballooned thanks to the shiny new GPUs burning both electricity and patience.
Still, investors are hooked. Because when you’re in the business of AI, normal valuation logic doesn’t apply. Somewhere between “₹1,000 crore capital raise” and “₹177 crore GPU contract from IndiaAI,” E2E has convinced the market that being unprofitable is just a growth phase — or as startups call it, “strategic burn.”
3. Business Model – WTF Do They Even Do?
Think of E2E Networks as India’s version of AWS, but designed by engineers who think ChatGPT and GPUs are cooler than server uptime. The company runs a full-stack AI/ML and GenAI cloud ecosystem. Translation: they rent GPUs (like H100, H200, A100, L40S) so that others can train fancy large language models without burning down their own servers — or their wallets.
Their flagship product, TIR AI/ML Platform, lets data scientists play with Jupyter notebooks, vector databases, LoRA models, and even vLLM integrations. Basically, it’s a digital playground for AI nerds.
They also offer:
Sovereign Cloud Platform – launched in 2025, it helps enterprises and governments build private/public AI clouds in just 15 days (no, not an exaggeration; they really said that).
AI Lab-as-a-Service (AILaaS) – for universities that want to give students GPU access without taking an education loan.
Self-Service Cloud Platform – because every Indian startup loves to say “we scaled on our own cloud.”
Add marquee clients like Zomato, Nykaa, IndiaMART, Zoomcar, and HealthKart, and you get a portfolio that screams “Indian tech stack,” literally.
The company earns 91% of its revenue domestically — which makes sense, because Indian startups are now obsessed with training models that can answer in Hinglish.
4. Financials Overview
Metric
Latest Qtr (Sep’25)
YoY Qtr (Sep’24)
Prev Qtr (Jun’25)
YoY %
QoQ %
Revenue
₹43.8 Cr
₹48 Cr
₹36 Cr
-8.9%
+21.6%
EBITDA
₹18 Cr
₹31 Cr
₹11 Cr
-41.9%
+63.6%
PAT
₹-13.5 Cr
₹12 Cr
₹-3 Cr
-211%
-350%
EPS (₹)
-6.69
7.20
-1.42
-193%
-371%
Annualised EPS: Not meaningful — losses don’t annualize well, just like crash diets.
Commentary: Depreciation (₹43 Cr this quarter) ballooned faster than GPU orders, eroding profitability. Even with a strong 41% OPM, the company bled at the bottom line. Clearly, E2E’s earnings per share are currently GPU per share — all processing, no profits.
5. Valuation Discussion – Fair Value Range Only
Let’s get nerdy.
Method 1: P/E Method Industry P/E (IT Services): 34.7 But E2E’s earnings = ₹4.54/share (TTM). → Implied value = ₹4.54 × 34.7 = ₹157.2/share. Current CMP = ₹2,812 → that’s 17x higher. Conclusion: Valuation currently exists in the “parallel AI universe.”
Method 2: EV/EBITDA EV = ₹5,397 Cr; EBITDA (TTM) = ₹128 Cr (approx). → EV/EBITDA = 42.2× (from screener). If we normalize to industry average 15×, → Fair EV = ₹1,920 Cr → Equity Value ≈ ₹1,950–₹2,100 Cr → Fair price range: ₹970–₹1,050/share.
Method 3: Simplified DCF (Educational only) Assume 30% growth in cash flow for 5 years (optimistic, given the GPU craze). Discount rate: 12%. Terminal growth: 5%. → Fair value range: ₹1,000–₹1,400/share.
⚠️ Disclaimer: This fair value range is for educational purposes only and not investment advice. (Unless your investment thesis is “AI + India = Infinity.”)
6. What’s Cooking – News, Triggers, Drama
₹1,000 Cr Fundraising: Approved via QIP/Rights/FPO. They’re not just raising money; they’re raising eyebrows.
L&T’s 15% Stake: L&T injected ₹1,000+ Cr through preferential allotment — a big corporate dad move to keep the kids in check.
Massive GPU Expansion: Now boasting ~3,700 GPUs, including 2,300 H200s — India’s largest private AI compute cluster.
Contracts Galore: Bagged ₹177 Cr GPU allocation order from IndiaAI, plus ₹88 Cr from Gnani AI.
Jarvis Labs Acquisition: To integrate another AI service provider under its TIR platform umbrella.
Unutilized Funds: As per latest monitoring report, ₹256 Cr remains unutilized from the ₹1,079 Cr preferential issue.
Auditors: No modification in review — always a sigh of relief when your GPU burns but your auditor doesn’t.
Basically, E2E’s corporate kitchen smells like burnt circuits and venture money — but the aroma is still attracting guests.
