It takes about one hour to read this article, but do not stop until you get to this part:
I am still having trouble finding one million of these things that are actually allocated anywhere, let alone in a data center, let alone one with sufficient power.
I do not know where these six million Blackwell GPUs have gone, but they certainly haven’t gone into data centers that are powered and turned on. In fact, power has become one of the biggest issues with building these things, in that it’s really difficult (and maybe impossible!) to get the amount of power these things need.
In really simple terms: there isn’t enough power or built data centers for those six million Blackwell GPUs, in part because the data centers aren’t built, and in part because there isn’t enough power for the ones that are.
If you do not know or understand what an SPV is or want to understand the financial implications of the AI bubble, you should listen (or read) this talk between Paul Krugman and Paul Kedrosky.
Once you read those two pieces you will understand why Microsoft et al are so desperately trying to sell you on AI. It comes down to this:
Big Tech Needs To Make $2 Trillion In AI-Specific Revenue by 2030 Or It’s Wasted Its Capex
2,000,000,000,000 additional revenue.