Cipher Mining’s $5.5 billion AWS pact re-rates a miner that reported $72.0m revenue in Q3. In Q3 the company reported $72.0 million of revenue (vs. a $79.1m consensus), delivered adjusted EPS of $0.10 (street loss estimate ≈ -$0.02) and announced a ~15‑year hosting/lease relationship with Amazon Web Services valued at about $5.5 billion to develop a 1‑gigawatt AI data‑center campus in Texas. The market’s reaction was immediate: Cipher shares jumped roughly 14.0% intraday on the AWS disclosure, while Cantor Fitzgerald published a $26 price target that frames the stock’s optionality relative to the latest operating scale.

Put simply: $72.0m of trailing quarterly revenue now sits under a $5.5bn multi‑year commercial framework. That math — one quarter’s revenue equal to ~1.3% of the contract headline — amplifies both upside and execution risk. For institutional investors the core questions are quantifiable: can Cipher convert large, multi‑year capacity commitments into recurring hosting revenue that dwarfs spot Bitcoin mining revenue (612 BTC produced by some peers in October), and how will capital allocation (capex, grid interconnect, and build‑to‑suit costs) pressure reported margins in the next 4–12 quarters?

Riot Platforms reported scale and cash this quarter but shares traded off ~14.0% on strategic repositioning. Riot reported record Q3 revenue of $180.23 million and GAAP net income of $104.48 million; management also disclosed a 112 MW core‑and‑shell development build at its Corsicana campus and ongoing deployments that convert power capacity into either Bitcoin‑mining rigs or AI hosting footprint. Despite the $180.23m revenue print, the stock fell about 14.0% after the company signaled it will accelerate transitions from pure mining to hybrid hosting — a trade‑off investors price as higher capital intensity and longer payback horizons even with the immediate $104.48m net income headline.

Riot’s numbers give traders precise levers to model. Use two facts: Q3 revenue $180.23m and Corsicana 112 MW pipeline. If Riot can monetize 112 MW at a hosting price of $X/MW (market bids for AI colo in the U.S. vary widely; institutional term sheets now reference multi‑hundred‑million‑dollar commitments), those dollars compound against Riot’s existing cash flow base of ~$104m in quarterly net income. Conversely, if utilization or pricing falls 20%, the same 112 MW materially extends payback. That sensitivity — presentable as ±20% revenue per MW scenarios — is why active traders marked the stock down even with the headline profit.

CleanSpark anchors the operational side: 612 BTC mined in October, $64.9m of realized Bitcoin sales, and a 285 MW AI‑focused site in Texas. CleanSpark reported 612 BTC mined in October and disclosed $64.9 million of Bitcoin sales in the period; it also signed an agreement for a 285 MW site in southern Texas explicitly aimed at AI infrastructure build‑out. The market priced those facts with short‑term pain: CleanSpark shares traded around $17.42 (-2.13% in a cited session) even as the company expanded its power footprint from treasury and balance‑sheet proceeds. For allocators, 612 BTC and $64.9m of monetizations are hard revenue proxies one can roll into cash‑flow models versus the long‑dated hosted AI contracts announced by peers.

Compare and contrast the three business models with concrete metrics. Cipher: Q3 revenue $72.0m, adjusted EPS $0.10, AWS contract ~$5.5bn for 1 GW. Riot: Q3 revenue $180.23m, GAAP net income $104.48m, 112 MW Corsicana build. CleanSpark: October BTC production 612 BTC, $64.9m Bitcoin sales, 285 MW Texas site. These are not rhetorical differences — they convert directly into balance‑sheet stress tests. For example, a 1 GW (1,000 MW) build at $1m+ per MW of installed civil/ electrical infrastructure (conservative in high‑spec AI campuses) implies capex of $1.0bn+, putting the AWS commitment in context of multi‑year CapEx schedules.

Execution cadence and milestones become the catalysts to watch — and trade — over the next 6–18 months. Key, dateable items: Cipher’s integration timeline for the 1 GW AWS joint entity (starts 2026 per press commentary), Riot’s phased Corsicana construction (112 MW ramps publicly highlighted in filings), and CleanSpark’s 285 MW site permitting and grid interconnection milestones. Each milestone shifts headline risk: a delayed commercial operation date (COD) pushes out revenue recognition and extends payback; an on‑time COD converts large capacity into recurring hosting fees, which investors have valued aggressively (Cipher +14% day‑of‑announcement).

Valuation and forward multiples: how to size positions. Use two benchmarks: street forward revenue and announced commercial commitments. Cipher’s Q3 trailing revenue of $72.0m vs. a $5.5bn contract implies a potential multi‑year revenue stream that can re‑rate the multiple from single‑digit mining comparable multiples to higher hosting multiples (apply a 10–15x rule of thumb to predictable hosting EBITDA). Riot’s Q3 revenue of $180.23m and outsized net income ($104.48m) suggest a materially different profitability profile — yet the market trimmed the multiple because of execution uncertainty on asset conversion. CleanSpark’s $64.9m monetizations and 612 BTC production act as an operational hedge and provide a nearer‑term cash conversion multiple to estimate downside cushion (inventory + realized crypto sales as liquidity buffer).

Risk matrix with numbers you can plug into models. Scenario A (baseline): 75% utilization of new hosting capacity, pricing at $X/MW yields incremental annual revenue of $Y for each 100 MW; Scenario B (stress): 50% utilization and 20% lower $/MW. For concrete tuning: if Riot’s 112 MW is monetized at $200k per MW per month of contracted equivalent revenue (≈$2.688m per year per MW), full utilization would add ~$301m annualized revenue; at 50% utilization that inflates to ~$150m — a swing that dwarfs quarterly revenues and explains 14% share price moves for the same headline profits. Cipher’s AWS $5.5bn headline functions the same way — the delta between committed capacity revenues and build/capex cadence determines near‑term volatility.

Trading checklist for allocators and prop desks. 1) Monitor cipher/AWS milestone filings and any Cantor or sell‑side estimates that update the $26 target; 2) re‑model Riot’s Corsicana 112 MW using $/MW per year revenue bands of $100k–$400k and stress test leverage; 3) treat CleanSpark’s BTC production (612 BTC in October) and $64.9m realized sales as a liquidity trigger — if realized sales fall below a rolling 3‑month average, probability of near‑term equity dilution rises. Each item is binary and dateable; markbooks should show sensitivity tables that move NAV and EV/EBITDA by ±10–30% on execution slippage.

Bottom line for portfolio managers: the market is repricing power as a programmable asset. Cipher’s $72.0m quarterly revenue now exists against a $5.5bn, 15‑year AWS commitment to build 1,000 MW; Riot’s $180.23m revenue and $104.48m net income come with 112 MW of new Corsicana capacity; CleanSpark’s 612 BTC production and $64.9m monetizations underpin a 285 MW AI site plan. These are concrete, measurable inputs you can put into DCFs, break‑even capex schedules and scenario P&Ls — and they explain why the same sector can see a +14.0% move (Cipher) and a -14.0% reprice (Riot) in close succession. Position sizing should therefore be a function of milestone probability, capex funding path, and the proportion of revenue that transitions from volatile Bitcoin sales to contracted hosting fees.