1. At a Glance – The Quarter That Made GPUs Sweat

If Indian equities had a “main character energy” quota per quarter, Netweb Technologies India Ltd would have consumed a disproportionate chunk of it in Q3 FY26. ₹805 Cr in quarterly revenue (up 141% YoY), ₹73.3 Cr in PAT (up 147% YoY), ROCE north of 32%, ROE flirting with 24%, and a stock price that still trades at a triple-digit P/E like it’s casually browsing NVIDIA’s keynote slides instead of a valuation textbook.

Market cap sits at ~₹18,964 Cr, current price ~₹3,347, and the stock has delivered a violent 76.5% return in six months while managing to disappoint impatient traders over the last three months. That’s classic “institutions accumulating, retailers arguing on Twitter” behaviour.

This quarter was not driven by hope, vision decks, or “AI optionality” buzzwords. It was driven by hard shipments—GPU systems, private cloud, HCI, and sovereign-scale AI infrastructure orders that actually showed up in the P&L. When a company does ₹805 Cr in one quarter against ₹1,149 Cr revenue for the entire FY25, you’re no longer dealing with projections—you’re dealing with scale shock.

So the real question is not “is growth real?”—that ship has sailed. The real question is: how much perfection is already priced in, and how long can Netweb keep outrunning gravity? Ready to peel the server rack? Let’s go.

2. Introduction – From Quiet OEM to India’s AI Backbone

Netweb Technologies India Ltd was incorporated in 1999, long before “AI infra” became dinner-table vocabulary. For most of its life, it quietly did the unglamorous job of designing and manufacturing high-end computing systems—supercomputers, storage, private cloud stacks—largely for research labs, government institutions, and enterprises that cared more about FLOPS than brand recall.

Fast-forward to FY25–FY26, and suddenly Netweb is being discussed in the same breath as sovereign AI infrastructure, NVIDIA Blackwell platforms, GPU orchestration software, and defence-grade compute clusters. That’s not a pivot—that’s timing meeting preparation.

India’s National Supercomputing Mission, defence digitisation, space research, and now enterprise AI adoption have created a rare sweet spot: demand that requires local manufacturing, integration, and long-term support. This is where Netweb thrives. You don’t import patriotism in a cardboard box from overseas—you build it in Faridabad with SMT lines and firmware engineers.

But before we get carried away, let’s be clear: Netweb is not a SaaS fairy tale. This is a manufacturing-led, working-capital-intensive, order-book-driven business with real execution risk. Which makes its current growth even more impressive—and its valuation even more dangerous.

So ask yourself: are we looking at the birth of India’s NVIDIA-adjacent infrastructure champion, or a phenomenal run that now demands Olympic-level execution every single quarter? Let’s dissect the machine.

3. Business Model – WTF Do They Even Do?

Imagine explaining Netweb to a smart friend who skipped hardware lectures and only understands AI through memes. Here’s the cheat code:

Netweb designs, builds, integrates, and supports high-performance computing systems—the stuff that sits inside data centres and crunches terrifying amounts of data for AI training, simulations, weather models, defence systems, and research labs.

Its portfolio spans:

• Supercomputing Systems – national labs, research institutions, government missions
• AI Systems & Enterprise Workstations – GPU-based compute for AI workloads
• Private Cloud & HCI – on-prem cloud stacks for enterprises that don’t want AWS bills inducing panic attacks
• High Performance Storage (HPS) – throughput up to 450 GB/s
• Software & Services – small in revenue, big in stickiness

What makes Netweb different is full-stack control. It designs systems, assembles hardware locally, integrates global components (Intel, AMD, NVIDIA, Samsung, Seagate), and then supports them long term. This is not reselling. This is engineering plus manufacturing plus servicing.

Revenue mix evolution tells the story. Supercomputing dropped from 39% in FY23 to 26% in Q1 FY26, while AI systems jumped from 7% to 29%. That’s not decline—that’s migration from academic compute to commercial AI workloads.

Question for you: how many Indian companies can genuinely claim to be mission-critical in both defence labs and enterprise AI data centres? Exactly.

4. Financials Overview – Numbers That Make Excel Nervous

Quarterly Performance (Figures in ₹ Crore)

(Result type locked: Quarterly Results → EPS annualised ×4)

Annualised EPS (Q3 FY26) = ₹12.94 × 4 = ₹51.76

At a CMP of ~₹3,347, that implies a forward-looking P/E of ~65x if this run-rate sustains. If it doesn’t? Welcome to volatility hell.

Witty takeaway: Netweb