Log in

Data Centers & Cloud

The buildout behind the AI bill

Hyperscaler capex, colocation supply, power, cooling and networking — where the money goes after the GPU order, and which companies book it.

What Big Funds Own42 fundsWhat Berkshire, Bridgewater and other big funds reported holding last quarter.What Analysts Expect16 stocksRatings and price targets from the big institutions, each with its public source.

2026-09-11

The Pentagon Just Underwrote Fluidstack — And CoreWeave's $35.6B Debt Stack Suddenly Looks Optional insight cover
Private Company
CRWV · NBIS•14 min read

The Pentagon Just Underwrote Fluidstack — And CoreWeave's $35.6B Debt Stack Suddenly Looks Optional

The Pentagon's Office of Strategic Capital is in talks to lend roughly $5 billion to Fluidstack — more than six times the program's largest prior commitment — turning a private UK neocloud into a sovereign-backed competitor and reframing CoreWeave, Nebius and peers as the higher-cost option for the next leg of U.S. AI infrastructure. The mechanism matters: this is not commercial credit. Federal dollars below Treasury rates, paired with Google's existing $1.3B lease backstop, reset the funding curve for the entire neocloud sector — and push TeraWulf, NVIDIA and Texas power plays like Vistra into the most leveraged positions on the supply chain.

Pentagon loan to Fluidstack (in talks): $5BOSC's typical direct-loan cap: $150M
Read article

2026-09-10

Massachusetts Just Made Clean Power a Permitting Asset—Hyperscalers Without Nuclear PPAs Now Have to Siting Around It insight cover
Industry News
ES · CEG•16 min read

Massachusetts Just Made Clean Power a Permitting Asset—Hyperscalers Without Nuclear PPAs Now Have to Siting Around It

Governor Healey's September 8 Executive Order No. 658 forces every Massachusetts data center above 25 MW to procure 100% new clean electricity or write a check into a Ratepayer Protection Fund by year-end. The rule re-prices power as a permitting asset inside a top-tier US tech hub, hands a structural advantage to Constellation Energy, Talen Energy, Vistra and NextEra that already hold clean PPAs with Microsoft and Amazon, and pushes new hyperscale demand to Pennsylvania and Texas where those deals are already in place. For local utilities like Eversource, the EO turns a potential captive load-growth tailwind into a stranded opportunity.

Clean Energy Standard (CES) annual ramp: +6 ppt/yrData center peak-demand threshold: 25 MW
Read article

2026-09-09

Google’s €13B Finland push tests whether hyperscaler capex buys more than compute insight cover
Industry News
7 min read

Google’s €13B Finland push tests whether hyperscaler capex buys more than compute

Google says it will invest at least €13B in AI infrastructure in Finland across 2027–2028—alongside grid upgrades, clean energy and battery projects. The investor question is whether this shifts power, cooling, and network bottlenecks into durable leverage (higher utilization and pricing power), or whether it mainly extends a capex cycle that only tightens cash-flow pressure.

Event Date: 2026-09-09Topic Type: Industry News
Read article

2026-09-08

Anthropic’s Claude Fable 5.1 turns agent economics into the product: the near-term moat is “prompt-cache gravity,” not just model quality insight cover
Private Company
9 min read

Anthropic’s Claude Fable 5.1 turns agent economics into the product: the near-term moat is “prompt-cache gravity,” not just model quality

Claude Fable 5.1 and its pricing changes reframe agent rollouts around unit economics: Anthropic cut cache-read costs to $0.25 per million tokens (and describes up to ~45% savings for “highly agentic tasks”). The investor question for a future IPO-style narrative becomes whether enterprises will buy the compute-efficiency and enterprise-governance package—then scale agent usage faster than rivals can match.

Base token pricing (Fable 5.1): $10Base token pricing (Fable 5.1): $50
Read article
OpenAI's Malaysia Bet With Firmus Just Made the US Hyperscaler Capex Trade a Two-Region Story insight cover
Industry News
NVDA · MSFT•15 min read

OpenAI's Malaysia Bet With Firmus Just Made the US Hyperscaler Capex Trade a Two-Region Story

Nvidia-backed NVIDIA partner Firmus has signed OpenAI as anchor tenant for two Malaysian AI factories, lifting its contracted capacity above 900 MW days before a planned A$15–30 billion ASX float. The deal is the first credible proof that the AI build-out has a second, cheaper geography: Malaysia's medium-voltage industrial power runs ~6.7¢/kWh against PJM's ~8.99¢/kWh in Northern Virginia, and PJM capacity prices have gone from $29/MW-day to $329/MW-day in two auctions. For Microsoft, Oracle, Amazon and Alphabet, the trade is no longer

Firmus contracted capacity: 900+ MWDeal scope: 2 Malaysian AI factory sites
Read article
Qualcomm lands AWS: the $60B ceiling that turns the data-center story from slideware to balance sheet insight cover
Industry News
AMZN · META•13 min read

Qualcomm lands AWS: the $60B ceiling that turns the data-center story from slideware to balance sheet

Qualcomm's Sept 8 multi-generational AI silicon deal with AWS caps a potential $60B in chip purchases and grants Amazon warrants for 25 million shares at $161.26 — turning the data-center pivot from a slideware promise into a balance-sheet event. The warrant structure, vesting only with actual chip purchases, makes Amazon both customer and equity stakeholder, and reframes the AWS-Trainium versus merchant-silicon fight as a multi-vendor affair. Investors should track three things: when the next warrant tranche vests, how Amazon's inference silicon capex ramps, and whether Meta and Microsoft follow on the inference side the way they already have on the CPU side.

Deal ceiling (chip purchases): $60BWarrant coverage: 25M shares @ $161.26
Read article

2026-09-07

Why Patagonia’s Cold + Wind Can Break the Next AI Data-Center Siting Trade-Off—and Who Benefits insight cover
Industry News
AMZN · GOOGM•8 min read

Why Patagonia’s Cold + Wind Can Break the Next AI Data-Center Siting Trade-Off—and Who Benefits

A Sept. 7 Reuters report says hyperscalers are probing Argentina’s Patagonia for mega data centers, attracted by wind-powered power economics and a cold climate that can lower cooling costs. The market signal isn’t just “cheaper kWh”—it’s that AI siting economics may shift toward places with both dispatchable-like renewable pricing dynamics and long runway for grid and generation buildout, reshaping who wins interconnect, power contracts, and capacity. Public investors should watch the listed power and data-center-adjacent names most exposed to these contract and buildout cycles.

Amazon market scale (context for capex risk): TTM valuation multiplesAlphabet market scale (context for AI capex risk: TTM valuation multiples
Read article
Tata Consultancy Services plants a 1GW hyperscale flag in Hyderabad—forcing a power-and-capex price fight across India’s AI data-center buildout insight cover
Industry News
TCS.BSE · NVDA•7 min read

Tata Consultancy Services plants a 1GW hyperscale flag in Hyderabad—forcing a power-and-capex price fight across India’s AI data-center buildout

Tata Consultancy Services’s HyperVault is committing up to ₹70,000 crore (about $7.41B) for a purpose-built up to 1GW AI data-center campus in Hyderabad, backed by TPG—an inflection for India’s hyperscale build map. The move matters less for near-term IT services revenue and more for who controls “AI availability” in India: power, cooling, and capacity scheduling—where NVIDIA NVDA racks and Vertiv VRT infrastructure pricing flow downstream to hyperscale and colocation competitors.

Event Date: 2026-09-07Topic Type: Industry News
Read article

2026-09-06

When ChatGPT, Claude, and Grok All Went Dark for 93 Minutes, the Single-Cloud AI Thesis Broke insight cover
Industry News
GOOGL · ORCL•12 min read

When ChatGPT, Claude, and Grok All Went Dark for 93 Minutes, the Single-Cloud AI Thesis Broke

On September 3, 2026, three of the four frontier AI providers — OpenAI, Anthropic, and xAI — suffered overlapping downtime for 93 minutes. Google Gemini stayed up. The episode didn't expose a shared external failure: each provider cited a different internal cause. What it exposed is that \"cloud AI\" is now a single-point-of-failure problem at the inference layer, and that resilience has become a buying criterion — which is why Alphabet's vertical stack and Oracle's OCI (+93% YoY) are suddenly more relevant than another Azure region.

Overlapping downtime window: 93 minutesChatGPT Downdetector peak: ~37,000 reports
Read article

2026-09-05

Foxconn's Q3 call turns the AI-server read from 'strong demand' into '2027 capacity stays very strong' insight cover
Industry News
DELL · AVGO•15 min read

Foxconn's Q3 call turns the AI-server read from 'strong demand' into '2027 capacity stays very strong'

Foxconn's August 12 earnings did more than beat Q2 — management explicitly told investors that 2027 demand for AI production capacity will stay very strong as Vera Rubin racks enter mass production this quarter, with cloud & networking guided up 'high double-digit' both QoQ and YoY. That forward signal lands alongside a cross-validating cluster: Dell's record $95B AI backlog, Broadcom's $21.7B Q4 AI-semi guide (+236% YoY), NVIDIA's $89B data-center print (+117% YoY), Wistron's AI servers hitting 95% of revenue, and Lenovo's $54B AI-server pipeline (+157% QoQ). The structural story is sharper than consensus: AI servers just crossed 51% of Foxconn's revenue, displacing iPhone as the swing factor and reframing the stock as an AI-infrastructure bet.

Foxconn Q2 2026 revenue: NT$2.53TFoxconn Q2 2026 net profit: NT$59.97B
Read article

2026-09-04

Ciena proves AI optical demand is real—and the next bottleneck shifts upstream supply, not downstream build insight cover
Earnings
CIEN•7 min read

Ciena proves AI optical demand is real—and the next bottleneck shifts upstream supply, not downstream build

Ciena reported $1.67B revenue in Q3 FY2026, up 37% year over year, confirming its “AI-networking order-book” thesis for optical interconnect. The key investor implication is timing: when bookings accelerate faster than component throughput, near-term reads move from demand validation to supply availability across the coherent/optics test chain.

Revenue: $1.67BYear-over-year growth: 37.0%
Read article
Crusoe’s $30B mark reframes AI data centers as a power-ownership business—boosting “turbine-to-rack” spillovers that scale-colo players can’t easily replicate insight cover
Private Company
8 min read

Crusoe’s $30B mark reframes AI data centers as a power-ownership business—boosting “turbine-to-rack” spillovers that scale-colo players can’t easily replicate

Crusoe’s reported $3B fundraising at a $30B valuation signals that investors will pay for AI capacity only when power reliability and grid stability are built into the product. The key tell is Crusoe’s disclosed path to on-site generation (via gas turbines) and power-conditioning/ride-through (via AI UPS), which shifts leverage from colocation landlords to power-integrators and equipment suppliers.

Reported new raise: $3BReported valuation: $30B
Read article
Nscale just locked in $3.5B of compute for humanoids—turning embodied AI into a tenant-class the “neocloud” playbook can underwrite insight cover
Private Company
NSCALE · FIGURE•9 min read

Nscale just locked in $3.5B of compute for humanoids—turning embodied AI into a tenant-class the “neocloud” playbook can underwrite

Nscale and Figure have signed a multi-year compute partnership starting in the second half of 2027, with an initial $3.5B commitment and an intent to scale beyond $6B. The deal matters less as robotics news and more as infrastructure proof: it extends Nscale’s contracted-revenue model (already validated by Anthropic-style compute deals) into a new physical-AI demand profile—low-latency, high-volume data pipelines tied to real-world deployment. That raises the stakes for how investors should think about compute capacity, utilization risk, and the timing pressure on incumbent robotics and cloud tenants.

Event Date: 2026-09-04Topic Type: Private Company
Read article
Microsoft's Foundry channel turns OpenAI GPT-6 Astra into the enterprise monetization gate—before open access insight cover
Private Company
MSFT · GPT-6-ASTRA•7 min read

Microsoft's Foundry channel turns OpenAI GPT-6 Astra into the enterprise monetization gate—before open access

OpenAI is rolling GPT-6 Astra out first through its Daybreak enterprise program (with ChatGPT tiering and API access following), while Microsoft promotes how OpenAI models are sold through Microsoft Foundry under Azure’s enterprise distribution mechanics. The tradeoff is clear: scarcity is being used to secure higher-trust buyers and tighter deployment controls—setting up Microsoft to capture more “resale-style” AI revenue economics per seat and per inference rather than waiting for universal open APIs.

FY2026 revenue: $331.8BFY2026 operating cash flow: $182.9B
Read article

2026-09-03

Dell clears the AI-rack “margin test” by proving its mix can expand—guidance lifts while gross and operating margin jump insight cover
Earnings
DELL · AVGO•7 min read

Dell clears the AI-rack “margin test” by proving its mix can expand—guidance lifts while gross and operating margin jump

Dell raised FY27 guidance after posting Q2 results that expanded both gross margin (20.9%) and non-GAAP operating margin (12.6%), alongside record AI-Optimized Servers revenue ($16.4B) and backlog ($95B). The key investor takeaway is not just AI demand—it’s that AI-related product momentum translated into structurally higher margins, directly undercutting the “AI revenue with lower-profit payoffs” narrative that has pressured Broadcom and Credo.

FY27 revenue guide: $192.0BQ2 FY27 gross margin (GAAP): 20.9%
Read article
Can “neutral” colocation capture AI inference profit—now that Nvidia is building inside Equinix’s fabric? insight cover
Industry News
9 min read

Can “neutral” colocation capture AI inference profit—now that Nvidia is building inside Equinix’s fabric?

Equinix’s Sept. 2, 2026 launch of “Equinix Inference Exchange,” anchored on Nvidia’s enterprise reference architecture and Together AI’s inference platform, is a bid to make vendor-neutral inference easier to deploy across Equinix’s interconnection layer. The investor question isn’t whether AI demand grows—it’s whether this packaging changes where the margin lands: colocation economics, or the hyperscalers’ walled-garden inference/distribution advantage.

TTM revenue: $9.83BTTM net income: $1.53B
Read article
Nscale’s $103B “contracted revenue” claim raises the hardest audit question in neoclouds: is it demand—or financing math? insight cover
Private Company
CRWV•7 min read

Nscale’s $103B “contracted revenue” claim raises the hardest audit question in neoclouds: is it demand—or financing math?

Nscale’s Sep. 2 claim that it has ~$103B of contracted AI revenue ahead of a potential IPO is big enough to reframe expectations for the neocloud buildout. But without line-item, investor-auditable contract terms, that number is difficult to distinguish from circular financing dynamics—exactly the risk investors have been stress-testing in CoreWeave, where reported revenue backlog is massive even as losses persist.

Revenue backlog (CoreWeave): $104BGAAP net loss (CoreWeave): -$626M
Read article
The Software Recovery Map: AI “Attach” Now Shows Up in Guidance—Here’s the Winner/Laggard Split insight cover
Industry News
10 min read

The Software Recovery Map: AI “Attach” Now Shows Up in Guidance—Here’s the Winner/Laggard Split

Software’s rebound since the SaaSpocalypse bottom is real, but it isn’t uniform. The re-rating is separating layers that convert AI demand into measurable guidance (data platforms, CRM/enterprise suites, and process “execution” layers) from names still priced off a future miracle—visible in what companies actually forecast and how quickly they monetize AI-related usage.

Revenue (TTM): $5.43BOperating margin (TTM): -22.8%
Read article

2026-09-02

Brazil just made data-center capex cheaper—turning the US power bottleneck into a potential LatAm “second queue” for AI builds insight cover
Industry News
9 min read

Brazil just made data-center capex cheaper—turning the US power bottleneck into a potential LatAm “second queue” for AI builds

Brazil’s Congress approved a special tax regime (REDATA) that suspends key federal taxes on data-center IT capex starting in 2026, reducing the cost and friction of deploying new capacity in-country. That timing matters because several US states are tightening approvals and interconnection timelines for new data centers, leaving AI demand to find the next buildable power queue. The twist for investors: the same Brazil-for-AI push is occurring alongside a new 25% US Section 301 tariff on Brazilian goods, which can raise cross-border equipment costs and complicate the economics of any Brazil-to-US supply chain.

Digital Realty (FY2025 revenue): $6.11BEquinix (FY2025 revenue): $9.26B
Read article
Credo’s post-earnings drop tests whether copper interconnect cash flows can keep up with AI rack spend — even as Dell confirms demand is accelerating insight cover
Earnings
CRDO•7 min read

Credo’s post-earnings drop tests whether copper interconnect cash flows can keep up with AI rack spend — even as Dell confirms demand is accelerating

Credo CRDO reported Q1 FY2027 results above the optics-friendly narrative investors were pricing, but the stock sold off sharply the next day. The key question for AI rack investors is whether the market is rotating from “copper rack capture” to “optics/other bottlenecks,” and whether Credo’s guidance supports that transition.

Revenue: $479.0MGAAP gross margin: 64.5%
Read article

What to expect

Evidence-first notes with a visible point of view.

This section collects sharp takes on earnings, shareholder meetings, and market structure. Each new piece should make the thesis, the facts, and the implications obvious within the first few screens.

Expect direct analysis, not generic commentary.

Expect the data to be explicit and the argument to be easy to follow.