---
type: "Topics"
locale: "en"
url: "https://longbridge.com/en/topics/39428171.md"
description: "🔥🚨 Why do I hold $IREN(IREN.US) in my growth portfolio? The key isn't &#34;mining,&#34; but the mispricing of &#34;power + computing power&#34;Many people still view $IREN(IREN.US) with the old label—A Bitcoin mining company.But I now tend to see it as something completely different:A &#34;power-driven computing power platform&#34; that is being repriced.My core logic for holding it is actually just four points.First, demand validation has already occurred.The nearly $100 billion collaboration related to $Microsoft(MSFT.US) is essentially not a single order, but a signal:Hyperscale customers have already started outsourcing their AI computing power needs to this type of platform.What does this mean?It's not &#34;whether there is demand,&#34; but &#34;demand is already looking for supply.&#34;Second, the power reserve far exceeds current demand.$IREN(IREN.US) has already secured about 4.5 gigawatts of power capacity,while achieving the year-end ARR target of about $37 billion requires only about 500 megawatts.In other words:The resources are already prepared, but revenue only needs to utilize a small portion.This structure is essentially &#34;supply first.&#34;Once demand is released, expansion doesn't require rebuilding infrastructure.Third, it's the leap in computing power scale.By 2028, the plan for 50,000 $NVIDIA(NVDA.US) B300 GPUscorresponds to an operating profit potential exceeding $10 billion.More importantly, this will push the total GPU scale to about 150,000 units.This isn't linear growth, but platform-level amplification.Once computing power reaches scale, marginal returns will significantly improve.Fourth, the cost structure advantage is beginning to materialize.Sweetwater 1 is expected to go online in Q2 2026,the key isn't the project itself, but the validation of the &#34;10:1 power advantage.&#34;In AI computing power competition, power is the cost foundation.Whoever can obtain cheaper, more stable powerwill have long-term pricing power.Putting these four points together, it's actually a very clear structure:Demand (AI computing power) is confirmedSupply (power + infrastructure) is already pre-positionedCost advantage (energy) is formingYet the market is still pricing it using the old framework of a &#34;mining company.&#34;This is why I'm willing to hold it.Of course, the premise for this logic to hold is also clear:AI computing power demand must persist,GPU deployment must proceed as planned,The power advantage must truly translate into profits.If any link fails, the valuation will be compressed again.So this isn't &#34;certainty,&#34; but a &#34;structural opportunity.&#34;Do you tend to see $IREN(IREN.US) as an extension of a traditional mining company,or as a platform that is transforming into AI infrastructure?"
datetime: "2026-03-22T08:37:57.000Z"
locales:
  - [en](https://longbridge.com/en/topics/39428171.md)
  - [zh-CN](https://longbridge.com/zh-CN/topics/39428171.md)
  - [zh-HK](https://longbridge.com/zh-HK/topics/39428171.md)
author: "[辰逸](https://longbridge.com/en/profiles/16318663.md)"
---

# 🔥🚨 Why do I hold $IREN(IREN.US) in my growth por…


### Related Stocks

- [IREN.US](https://longbridge.com/en/quote/IREN.US.md)
- [NVDA.US](https://longbridge.com/en/quote/NVDA.US.md)
- [MSFT.US](https://longbridge.com/en/quote/MSFT.US.md)