# data center revenue — X 热门讨论 (2026-09-09 13:15 UTC)
## @LLuciano_BTC (Lucky) · 09-09 11:04 · ♥359 ↻20 💬53 We are not in an AI boom anymore. We are in the middle of an infrastructure race that could reshape the global economy. The AI race is no longer just about who builds the smartest model. It’s becoming a race for compute, chips, power, data centers and energy.
➜ NVIDIA’s data-center revenue has reached extraordinary levels as demand for AI compute keeps accelerating.
➜ Global data-center electricity consumption is projected to more than double by 2030.
➜ AI semiconductor demand is creating a massive new market for companies supplying the infrastructure behind these models.
Companies are signing infrastructure deals worth tens of billions just to secure future compute capacity. This is much bigger than an “AI model race.”
We’re watching an entirely new technology stack being built in real time.
The biggest winners may not even be the companies with the best model. They could be the ones supplying the compute, power and infrastructure underneath the entire AI economy. > 引用 @KobeissiLetter: AI is much bigger than most people realize.
We currently have multiple $1+ trillion companies switching spots as the "best" AI model on a monthly basis.
OpenAI just went from "falling behind" in the AI race to reaching Artificial General Intelligence (AGI) in a matter of weeks.
Meanwhile, frontier models from OpenAI, Anthropic, Google, and xAI continue to leapfrog one another across major benchmarks and capability categories.
When you have multi-trillion dollar companies losing and regaining the lead on a monthly basis, that's not normal.
That's indicative of a historic technological revolution, with hundreds of billions of dollars in annual CapEx spending and trillions of dollars in AI infrastructure investment planned over the coming years.
We are in the midst of a generational shift in technology.
Take advantage of it. https://x.com/LLuciano_BTC/status/2097642170534797349
## @JasonL_Capital (Jason Luongo) · 09-09 10:53 · ♥47 ↻4 💬5 According to $NVDA's Jensen Huang, AGI has arrived.
$NBIS CEO says AI compute demand is now stretching into 2028, with customers asking for tens of thousands of GPUs.
We need more compute, and that's why I'm investing in these 10 companies:
1. $MRVL - Marvell
Revenue grew 37% to $2.74B with data center now 79% of the business and up 46%. Management raised the fiscal 2028 revenue target to $18B and said the custom silicon business will more than double next year. The Google TPU deal signed in August carries up to $120B of revenue potential over six years. The stock is 32% off its high because that payoff lands in 2029, which is exactly the kind of setup I want to own early. https://x.com/JasonL_Capital/status/2097639452206080479
## @Aman81407075 (Aman Godhasara) · 09-09 04:25 · ♥35 ↻4 💬1 Shadowfax vs Shiprocket: The Delivery Guy vs The Traffic Controller — Two Very Different Bets on the Same E-Commerce Boom
Both companies sit right at the center of India's D2C and quick-commerce explosion. Both just posted strong Q1 FY27 numbers.
But they're built on almost opposite philosophies one physically delivers the parcel, the other decides who should.
And here's the twist: Shadowfax is actually one of the 42 couriers that Shiprocket routes packages through. They're not just competitors, they're partners and counter-parties at the same time. Let me walk you through both.
The core identity split, in one line each
-->Shadowfax — "We will physically deliver your parcel or grocery in 10 minutes, same day, or next day." A tech-enabled, asset-right physical logistics operator, running its own sortation centers, dark stores, and rider network.
-->Shiprocket — "We will find you the cheapest/fastest courier, optimize your checkout, and reduce your return losses." A pure asset-light digital commerce operating system, owning zero trucks, planes, or delivery infrastructure.
Shadowfax builds and operates the physical machine. Shiprocket builds the software that decides which physical machine (theirs, Delhivery's, Blue Dart's) should carry each package.
The scoreboard, side by side
-->Revenue (Q1 FY27): Shadowfax ₹1,358.0 Cr (+65.0% YoY, +10.0% QoQ) versus Shiprocket ₹592.1 Cr (+33.8% YoY, +6.8% QoQ).
Shadowfax is running at more than 2.3x Shiprocket's revenue scale, and growing at nearly double the rate.
-->Order volume: Shadowfax ~25 crore orders in the quarter alone (+83% YoY), versus Shiprocket's ~5.5-6.0 crore quarterly orders (21.6 crore TTM transactions).
-->Annualized run-rate: Shadowfax ~100 crore orders/year versus Shiprocket's ~22-24 crore orders/year.
-->EBITDA: Shadowfax ₹67.0 Cr at a 4.9% margin, versus Shiprocket's ₹8.9 Cr at just 1.51%.
-->Bottom line: Shadowfax is profitable — ₹65.0 Cr PAT, a record quarterly profit. Shiprocket is still loss-making — a net loss of -₹13.7 Cr.
-->FY27 guidance: Shadowfax upgraded its full-year revenue growth guidance to 38-40% (from 27-30%), while Shiprocket is guiding for a more modest 21.0-23.5%, with Emerging targeted at 30%+.
Who actually bears the operational risk — this is the crux of the whole comparison
-->Shadowfax absorbs fuel price hikes, labor wage revisions, truck repairs, and sortation hub capex directly on its own books. When diesel goes up, Shadowfax feels it immediately.
-->Shiprocket takes zero asset risk. If diesel rises, the 3PL carriers on their platform bear the operational cost or pass the rate increase through — Shiprocket's own margin structure stays largely insulated either way.
That single difference explains almost everything else in this comparison — why Shadowfax is more profitable today (it captures the full economics of physical delivery), and why Shiprocket is structurally more insulated from commodity and labor cost shocks (it never owns the truck in the first place).
How each company actually makes money
-->Shadowfax — a physical logistics service fee per parcel delivered, plus dark store warehousing rent and slot fees. Revenue tied directly to physical throughput.
-->Shiprocket — freight rate arbitrage spread (the difference between what it pays a courier and charges the merchant), value-added service fees (COD settlement, transit insurance), and checkout/adtech software subscriptions. Revenue tied to transaction volume and software attach rate, not physical capacity.
Their core engines
Shadowfax runs four pillars — Express Parcel (16,000+ PIN codes), Quick Commerce (over 50% share of the outsourced market), a Dark Store Network (47 live locations), and SF 360, a self-serve platform for smaller sellers.
Shiprocket runs four pillars of its own — Core Shipping Aggregation, Omnichannel Dark-Store Sync, Martech & Checkout (Quikpay), and Cross-Border Exports.
The profitability story — and why the gap exists
Shadowfax has genuinely crossed into sustainable operating profitability — this was its third consecutive record profit quarter.
Part of how they did it: cutting "lost shipment debit costs" from 7.9% down to 5.5% of revenue using Vision AI reverse-logistics tools — a real, measurable operational improvement, not just scale doing the work.
Shiprocket, meanwhile, is still deliberately in an investment phase. Its Core Shipping segment is genuinely profitable on its own — ₹52.7 Cr in Adjusted EBITDA at a 12.8% margin.
But its faster-growing Emerging business (Omnichannel, Martech, Quick Commerce delivery) is still running an Adjusted EBITDA loss of -₹43.8 Cr. Add in ₹28.5 Cr of non-cash ESOP charges, and the consolidated result comes out to that -₹13.7 Cr net loss.
Unit economics — a genuinely useful way to compare two very differently sized businesses
-->Shiprocket: Adjusted Cash EBITDA per transaction expanded from ₹0.22 in FY25 to ₹1.50 in Q1 FY27 — roughly a 7x jump, a real signal of improving efficiency even while the top-line loss persists.
-->Shadowfax: processing ~25 crore shipments a quarter, their ₹67 Cr in Adjusted EBITDA works out to roughly ₹2.68 per shipment in EBITDA, and about ₹2.60 per shipment in actual PAT.
Where they're increasingly colliding — both companies are pushing into each other's territory
1. D2C brand aggregation — Shiprocket is the undisputed market leader here, powering 2.24 lakh active merchants and over ₹34,600 Cr in GMV, with high loyalty built on automated onboarding and Shopify plugins.
But Shadowfax isn't standing still — they've launched Shadowfax 360, a self-serve platform for SMEs already with 1,200+ sellers live, plus Prime Same-Day Delivery serving 400+ top enterprise D2C brands directly.
2. Quick commerce and dark stores — Shadowfax controls a physical dark store network (47 live out of 100 planned across 6 metros) and already controls over 50% of the outsourced quick-commerce delivery market (Blinkit, Zepto, Amazon Now).
Shiprocket counters with a pure software layer — Quick Commerce Slot Booking automates purchase orders and delivery time-slots into Zepto, Blinkit, and Instamart dark stores for D2C brands, without Shiprocket ever owning a dark store itself.
3. Large parcel/heavy goods — Shadowfax has launched Prime Large, a foray into bulky LCV deliveries across 10,000 PIN codes, already running at ~₹75 Cr ARR. Shiprocket answers with Shiprocket Cargo/Omuni, offering PTL/FTL cargo aggregation through their acquired Rocketbox platform.
The moats and vulnerabilities on each side
-->Shadowfax's moat: high asset utilization and physical network density. Being the only national 3PL running dedicated same-day delivery (Prime) while controlling over 50% of outsourced quick commerce is a genuinely hard barrier to replicate.
Their vulnerability: heavy exposure to physical operating costs — fuel, gig-worker wages, linehaul delays. If a major enterprise client like Meesho or Amazon renegotiates pricing, or decides to insource deliveries entirely, volume throughput takes an immediate hit.
-->Shiprocket's moat: high software switching costs and data ownership. Merchants don't just use Shiprocket to ship — they use it for checkout software, RTO fraud detection, marketing, and working capital loans, making the relationship genuinely sticky.
And owning zero physical fleet means total insulation from fuel inflation and vehicle capex. Their vulnerability: low core pricing power on pure shipping.
As D2C brands scale to serious volume, they often start negotiating directly with 3PL carriers, forcing Shiprocket to either compress rates or lean harder on software cross-selling to defend margins.
So who's actually winning? It depends on what you're measuring
On current scale, hyper-growth, and real profitability — Shadowfax is winning. Generating ₹1,358 Cr a quarter (+65% YoY) with ₹65 Cr in net profit proves their physical operating model has genuinely achieved sustainable operating leverage.
They're the dominant physical partner powering India's quick-commerce boom, and management just raised full-year guidance to 38-40% growth — leaving both traditional logistics peers and software aggregators behind on sheer execution.
On asset-light scalability, ecosystem control, and merchant lock-in — Shiprocket is winning. Shiprocket owns the merchant relationship itself — sitting at the top of the funnel, controlling the 1-tap checkout, a 15.5 crore consumer identity database, and the merchant's own storefront software.
It carries zero asset risk, generates a 72.3% contribution margin on its core technology, and simply cannot be disrupted by a diesel price spike or a warehouse labor strike. Once the fast-growing Emerging segment (+70% YoY) crosses breakeven, that operating leverage converts directly into high-margin software cash flow.
How I'd frame the final call
If your thesis favors immediate profitability, real revenue scale, and riding the hyper-growth wave of quick-commerce delivery directly — Shadowfax is the clear winner today, generating ₹65 Cr in quarterly PAT on a 100 crore+ annual order run-rate.
If your thesis favors asset-light platform monopolies, consumer data ownership, and high-margin software/fintech cross-selling without capital risk — Shiprocket is building the more defensible long-term tollbooth for Indian direct commerce, even though it hasn't turned the corner to profitability yet.
Not investment advice. Curious which model you'd back — the profitable physical operator or the asset-light platform still investing for scale? Let me know below. https://x.com/Aman81407075/status/2097541991190143288