Mandeep Singh

We've seen that how good the WeChat app is in terms of being that super app.

So that's what they're trying to get to with the agentic use cases is deploy a lot more of that using the LLM layer and do that at a very efficient price because at the end of the day, they have espoused cost and efficiency when it comes to running their infrastructure.

So they want that to be reflected at the agent layer as well.

Right, so for our workload specifically, for image and video, H200s are a little bit overkill for us, at least for the Hopper generation.

But when we're actually...

Moving away from the Hopper architecture to Blackwell architecture, we're actually just going directly from H100s to B200s or B300s because that actually gives us more flops per watt or per dollar.

We're actually moving over.

Yeah.

And look, I mean, based on consensus right now, we are talking about at least a 50% increase in capex or meta next year.

And no one, I think, would mind that given everyone's fees, there will be lift in revenue from AI down the line in the near term.

And

I think they needed some offset because the stock was very expensive on a free cash flow basis.

So this is a nice offset and I think the TPU news was also quite productive in that direction simply because they will be spending close to $50 billion on accelerator chips next year if $110 billion is their CapEx number.

Some portion allocated to TPUs, which cost maybe 25%, 30% cheaper, I think bring down that free cash, like help offset that free cash flow, which is going to be negative next year.

I mean, I like to think of, you know, metaverse ambitions as a moonshot for meta.

And look at, you know, Google's moonshot, for example, like self-driving, Venmo.

That is a lot real then in terms of driving, you know, tangible revenue, top line growth compared to what meta has accomplished with that aggregate, you know, 17 plus billion dollars in spend over the last three, four years.

And to me, it's still a moonshot.

And that's where, you know, paring back and allocating that capital towards, you know,

Probably a more productive use case in AI where they will see some top line lift is the right thing to do.