⚠️⚠️⚠️ $NBIS
I have a great idea. I am going to spend tens of billions building massive AI GPU clusters and cloud infrastructure so I can chase the AI gold rush and MAYBE generate large profits, if all goes well.
That’s a razor-thin return on capital – IF it works out.
Who’s buying $NBIS at these valuations?! Retail is 🚨
No need worrying about the risk that it might not work out! Arkady is the GOAT 😂
After a day reflecting on Jensen’s keynote last night, I believe this was the most important slide from Jensen Huang’s Computex keynote for $IREN investors.
Everyone focused on the NVIDIA ❤️ (Company) slides.
But I was drawn to what Jensen was actually saying about AI factory economics.
The entire slide is about generating more revenue from the same megawatt of power.
Faster time-to-first-token.
More tokens per watt.
Higher reliability.
Longer useful life.
The green curve isn’t showing more GPUs.
It’s showing more revenue generated from the same underlying infrastructure.
That’s important because NVIDIA isn’t just trying to convince the market to build more data centers.
They’re trying to convince the market to build AI factories using NVIDIA’s reference architecture.
And for that to happen, the first deployments need to prove they can generate materially more economic output per MW than traditional designs.
Cue the $IREN and $NVDA partnership & cloud services deal.
A few weeks ago NVIDIA announced its partnership with IREN and identified Sweetwater as a flagship deployment for its DSX AI factory architecture.
If you’re NVIDIA, this first deployment matters.
A lot.
The entire value proposition behind DSX depends on proving that the green curve is achievable in the real world.
Not on a slide or in a lab, but in the real world, at scale.
Because if operators can earn more revenue from every MW deployed, the economics of AI infrastructure change dramatically.
Jensen mentioned that 100 GW’s of AI factories would come online by the end of the century, and referenced four companies (I think it’s bullish for all four): $DLR, $IREN, $NBIS and NScale.
NVIDIA sees Sweetwater as one of the first opportunities to demonstrate what an AI factory can do when the entire stack is designed together.
NVIDIA is betting that the winners aren’t just the companies with the most power, but the companies that generate the most revenue from it. And $IREN is now set up to lead both…
@Agrippa_Inv@jiahanjimliu@ilzmcfly@mikealfred
Think about your favorite companies you wish you had equity in. Public and private.
Think about the most innovative companies in history.
Now think about their brand.
People questioning $IREN’s “acquisition” of a brand & media organization simply don’t understand the criticality of brand awareness & storytelling in today’s world.
IREN’s brand is a big part of what has attracted this massive retail community. Their website, video content, out-of-home strategy, etc. Now that will continue to be accelerated. Love it.
@brianfry01 I’m heavily invested in IREN and follow your posts. Thank you.
I’d love your perspective on AI cloud revenues. IREN had this foresight and vision since 2020… and yet AI Cloud revenue is minuscule. Compared to the other leaders in the sector, they are lagging over the last 6-12 months.
I am 100% aligned to the future vision, but am struggling to reconcile that if there was this much foresight… why is not showing up in the top line yet? And will this lag continue?
What’s your take on this?