# data center revenue — X 热门讨论 (2026-09-22 16:01 UTC)
## @DefiyantlyFree (Insurrection Barbie) · 09-21 23:44 · ♥39 ↻14 💬3 The Trans-Constitutionalists https://x.com/DefiyantlyFree/status/2102182201978671475
## @MikeBlock07 (Mike_Block) · 09-22 15:02 · ♥36 ↻1 💬1 I’ve been watching AI stocks closely, and one thing stands out to me: this rally is being supported by real spending, not just headlines.
Nvidia just reported $96.2B quarterly revenue, up 106% year over year, while Data Center revenue reached $89B, up 117%. Jensen Huang has also said Nvidia’s chip sales could roughly double in 2027, showing how aggressive the next phase of AI infrastructure could become.
The spending behind this is massive. Reuters reported that Microsoft, Alphabet, Amazon, Meta and Oracle could collectively spend more on AI-related infrastructure than they generate in free cash flow by 2027. Amazon has raised 2026 capital spending expectations to about $220B, while Alphabet lifted its forecast to $195B–$205B and Meta expects $125B–$145B.
At the same time, the market is heating up again. On September 21, the Nasdaq closed at a record high, AMD reached a $1T market cap, and Meta jumped more than 11% after its new AI assistant gained strong early adoption.
So my view is cautiously bullish on the AI theme, but not blindly bullish on every AI stock. Demand looks real. The bigger question is how long spending can keep growing faster than the returns.
And then there is policy. President Trump has announced an “AI Force” and has said AI could eventually represent up to 25% of U.S. GDP. That is a political claim and future projection, not a current economic measurement.
I’m watching NVDA, AMD and META closely. For me, the next phase is about revenue, margins, power capacity and actual AI monetization not hype alone.
#AIStocksWhatNext https://x.com/MikeBlock07/status/2102413109583323572
## @marksotoges (Mark) · 09-22 11:21 · ♥30 ↻0 💬1 $CIFR — why a 1 GW hyperscaler contract before December could be a double catalyst
Start with the AWS economics: 300 MW gross ~$5.5B contracted revenue / 15 yrs ≈ $1.22M/MW/year
Assume CIFR gets just 20% better economics on its next hyperscaler deal: $1.22M × 1.20 = $1.47M/MW/year At 1 GW: ~$1.47B average annual revenue ~$22B contracted revenue over 15 years
But here’s the part the market may be missing: Colchis has been conditionally designated for 1 GW of Base Load and still has to clear ERCOT’s verification process. ERCOT is specifically testing whether these enormous data-center power requests represent mature, committed projects rather than speculative load.
What could demonstrate commitment more clearly than a signed 15-year, 1 GW contract with a Tier-1 hyperscaler?
A deal before the final audit could therefore do two things simultaneously: Create ~$22B of contracted revenue. Materially strengthen the case for Colchis receiving its full 1 GW power allocation. CIFR already has the 1 GW AEP direct-connect agreement and is targeting Colchis for 2028-29. So the potential catalyst isn't simply: “CIFR signs another customer.” It's: 1 GW hyperscaler lease → ~$22B contract → stronger ERCOT commitment case → full 1 GW Colchis becomes substantially more valuable.
At ~414M shares, $5–6.5B of incremental equity value alone equals roughly $12–16/share.
From ~$19 today, that gives you a plausible $31–35 post-deal valuation range before considering any additional re-rating from ERCOT confirming the full Colchis capacity.
That is why the next 2–3 months could matter enormously for $CIFR.
The contract and the power allocation are not separate catalysts.
They would reinforce each other. https://x.com/marksotoges/status/2102357501937103338