The market still treats $IREN like a miner that got lucky. The reality is completely different. This is one of the few companies in the US that can turn POWER into COMPUTE at hyperscaler scale and the pieces are already in place: ✅ 3 GW secured renewable power ✅ Full vertical integration (power, land, construction, data centers, ops) ✅ Liquid-cooled AI facilities built for high-density GB300 racks ✅ A $9.7B / 5-year deal with Microsoft, with 20% prepayment upfront ✅ Sites energizing in April 2026 (Sweetwater 1), already locked and scheduled At $1.94B annual revenue for 200MW capacity, that’s $9.7M ARR per MW. This isn’t miner economics. This is hyperscaler economics. The moat isn’t GPUs. It’s the power + land + build + operate stack. Fast execution, low-cost energy, efficient infrastructure (1.1 PUE), and a pipeline competitors can’t replicate overnight. The AI era is entering a phase where energy becomes the real bottleneck, not silicon. $IREN is already positioned where everyone else is trying to go. People can debate the stock, but the infrastructure and execution aren’t up for debate. They’re already built. $IREN is going to surprise a lot of people.

$NBIS reports Q3 2025 on Tue 11/11 pre-market. Last Q2 was insane: ✅ +625% YoY revenue to $105M, ✅ EPS beat (-0.38 vs -0.41), ✅ ARR hiked to $900M–$1.1B. Stock exploded +44% after the $17.4B / 5yr Microsoft deal. $IREN also locked a ✅ $9.7B / 5yr Microsoft AI cloud deal with 20% prepayment ✅ $1.94B/yr revenue visibility. Stock popped +22% on announcement but got sold off in last week’s bloodbath. If $NBIS prints strong Q3 Tuesday → $IREN rallies in sympathy. We’ve already seen this in Sept: NBIS-MSFT news lifted BOTH tickers. Monday = likely last chance to buy IREN in the $60s before catalyst hits.

$IREN - The TRUTH behind the Q1 flex: ✅ Net income $384.6M - but $665M of that is mark-to-market gains on prepaid forwards + capped calls → Adj EBITDA $91.7M is the clean operating benchmark ✅ Today’s mix: $232.9M BTC mining vs $7.3M AI Cloud → yes, still 97% BTC - which means AI revenue is basically untouched upside as the GPU campuses turn on ✅ The $3.4B ARR end-2026 target = $1.94B contracted Microsoft + $1.5B BC buildout runway (based on GPU utilization assumptions) ✅ 140k GPUs is only 16% of IREN’s 3GW power envelope - the hyperscaler infra is already being engineered, paid for & power-secured ✅ $1.8B cash - funded by zero-coupon converts + GPU financing - aggressive, but this is exactly how hyperscalers are born: build before demand hits ✅ Childress redesign to Tier-3-like liquid cooling (100MW superclusters, 130–200kW racks) = they’re optimizing for high-density training loads, not hobby inference ✅ Timeline: BC full GPU conversion Eo’26, Sweetwater 1 Apr’26, Sweetwater 2 late ’27 - this is multi-year infra ramp, not a quarter-to-quarter trade Bottom line: IREN is already in hyperscaler territory - the Microsoft deal + NVIDIA alignment prove product-market-fit but the P&L hasn’t caught up yet. That disconnect = the opportunity. 🟢 $IREN