E2E Networks Q4 FY26 Concall Decoded: 186% Revenue Growth, 60% EBITDA Margins… and Management Still Wants More GPUs Than the Planet Has
1. Opening Hook
Just when everyone thought AI infra stocks had already inhaled enough optimism, E2E showed up saying, “Actually, we may need thousands more GPUs.” Casual.
While others debate whether AI is a bubble, E2E seems busy acting like the bubble needs more compute.
Revenue nearly tripled. EBITDA margins crossed 60%. Profitability turned positive. And management still sounds like they’ve barely started.
The fun part? They’re talking about speculative GPU buying as if hoarding Blackwells is now a competitive moat.
Even better, they refuse to be called an asset-heavy infra company. They want to be valued as a technology company monetizing “higher value tokens.” That phrase alone deserves its own IPO.
And just when you think this is another straightforward growth quarter — the analyst Q&A starts throwing curveballs.
Read on. The good stuff gets increasingly bizarre and interesting.
2. At a Glance
Revenue up 186% YoY – Apparently hypergrowth is now reported like normal quarterly business.
EBITDA up to ₹581 million – Margin expansion showed up looking overdressed at 60.7%.
PBT swung positive by ₹120 million sequentially – Losses quietly left through the back door.
FY26 revenue up 50% – “Decadal AI story” seems to be billing quarterly now.
80%+ utilization – GPUs clearly aren’t sitting around binge-watching Netflix.
2,048 Blackwell GPUs lined up – Because 3,900 GPUs simply felt emotionally inadequate.
Stock narrative upgraded to ‘AI factory of India’ – Subtle ambition, obviously.
3. Management’s Key Commentary
“The future is all AI. This is a decadal story.” (Translation: Please stop valuing us quarter-to-quarter, think civilization-scale 😏)
“We have speculative GPU capacity where capacity is running out everywhere.” (Translation: We bought chips before they became social status symbols.)
“We don’t want you looking at us as an asset monetization business.” (Translation: Please apply software multiples to hardware depreciation.)
“We met and exceeded the guidance we had given.” (Translation: Yes, we noticed too. Kindly update your models.)
“We continue to explore private credit asset-light models.” (Translation: We found new ways to buy GPUs without selling kidneys.)
“Tokens are becoming more valuable.” (Translation: Even compute now has luxury pricing.)
“India can become the AI factory of the world.” (Translation: Small dream. Just rewire global compute economics.)
“GPU contribution may move to 85%-90%.” (Translation: CPUs are slowly becoming the fax machines of infra.)
“We are not fixated on guidance.” (Translation: Numbers might undersell what we think is possible.)
“We continue to drown in opportunity.” (Translation: Nice problem to have, if true.)
Management’s tone was unmistakably aggressive. Less “conservative operator,” more “manifest destiny with GPUs.” Bold? Yes. Slightly theatrical? Also yes.
