0:00
What does an Electrolux oven and Anthropics Claude Code have in common?
Search this video
0:00
This is not a trick question, because believe it or not, if you own an Electrolux or Frigidaire branded oven and you are using Claude Code, there’s a chance both originate from the very same factory.
0:18
That’s because the “Colossus 1” datacenter in Memphis, Tennessee was built inside a former home appliance plant.
0:23
At its peak, the factory was producing hundreds of thousands of ovens, employing over 1,200 people before Electrolux closed the plant in 2019 and shifted production to Springfield.
0:34
A few years later, Elon Musk bought the now empty home appliance factory and retrofitted it into one of the largest AI datacenters.
0:45
But as fun as this story is, we are not here to talk about ovens.
0:45
We are here to talk about Anthropic buying the entire Colossus 1 compute capacity and what this deal means for Anthropic and xAI alike.
0:55
There’s a reason every stock somewhat related to semiconductors still only know one way: up.
1:05
It’s because even after a couple of years of high intensity compute build-out, compute for AI is still a limited resource.
1:16
That’s why AI models are more and more restrictive with their token use and some even started to downgrade their models to save precious compute.
1:22
This especially affects the latest frontier models that achieve the best results, because they often use the greatest number of tokens to do so.
1:33
So, it comes as no surprise that Anthropic is looking for more compute to power their latest AI models, such as Claude Code and Claude Opus.
1:38
And with the Colossus 1 datacenter, they have just found that.
1:44
The deal with SpaceX, since February the parent company of xAI, allows Anthropic to use all of the compute capacity that Colossus 1 offers. But how much is that?
1:53
The official blog post talks about “over 220,000 Nvidia GPUs”, but you wouldn’t watch a SemiAnalysis video if you didn’t want to know a bit more than that.
2:05
According to our information, Colossus 1 mostly contains Nvidia last-gen Hopper GPUs, approximately 200,000 of them, but it was upgraded with an additional 30,000 Nvidia Blackwell GB200.
2:17
We can also confirm the round-about 300 megawatts power draw.
2:24
And while 200,000 Hopper and 30,000 Blackwell GPUs sounds like much, Colossus 1 barely made it into our Top 10 datacenter list in January of this year.
2:35
And because it hasn’t been upgraded since then, but plenty of other sites have, it’s very likely not a top 10 site anymore.
2:40
Which doesn’t mean it’s not offering a ton of AI compute, because it still is.
2:46
Just not top 10 material any more.
2:46
By the way, let me know in the comments if you want an update to the Top 10 AI datacenter video.
2:52
For Anthropic the Colossus 1 deal with xAI has an obvious and immediate benefit.
2:58
It allows Anthropic to gain access to a lot of additional AI compute in a compute constrained environment.
3:03
And we can see instant results.
3:09
In the same blog post, that announced the compute deal with SpaceX, Anthropic detailed how the additional compute supplied by Colossus 1 directly allows them to increase Claude Code and Opus limits.
3:21
And by increasing token limits, Anthropic improves the accessibility and competitiveness of their frontier models.
3:28
Because in a compute constrained environment, model availability can matter almost as much as model quality.
3:33
If you can’t supply your model to customers, it doesn’t matter how good it is, because no one can use it.
3:39
So Anthropic gets AI compute to run their latest frontier models. But what about xAI?
3:45
Aside from the obvious monetary value, this deal offers more insight into the focus of xAI.
3:52
First, it indicates that the Colossus 2 datacenter, which is literally just a few miles down the road, has clearly become the number one focus for their in-house compute built-out.
4:04
At the end of last year, Colossus 2 already contained over 110,000 Nvidia Blackwell GB200 at around 350 to 400 megawatts.
4:09
Today it’s much closer to 1 gigawatt.
4:17
It might just be that Colossus 1 doesn’t really matter anymore for xAI.
4:22
But it also shows that xAI doesn’t have a competitive frontier model at the moment.
4:27
Because even though Colossus 2 is much bigger than its little brother, it’s still a compute starved environment.
4:32
And if xAI had a competitive frontier model, they could surely make use of an extra 300 megawatts of AI compute. But they don’t.
4:39
The contract with Anthropic also indicates that xAI probably doesn’t expect to have a competitive model in the foreseeable future, because again, if they did, they probably wouldn’t rent out precious AI compute resources.
4:55
Anthropic gaining access to Colossus 1 is a very interesting deal.
4:55
On the surface both companies benefit.
5:01
Anthropic gains access to existing infrastructure which immediately increases their compute capacity and directly impacts their Claude delivery.
5:06
And xAI slash SpaceX surely does get a decent amount of money for a datacenter they apparently have no internal use for.
5:18
This means xAI can focus on Colossus 2.
5:18
But it also indicates that xAI doesn’t expect to compete with other frontier AI models in the near future.
5:24
In any case, AI severs in a former home appliance plant are now running large language models.
5:31
And I think that’s amazing, because no matter if ovens or Claude Code, both can be used for cooking up something beautiful.