Jay Goldberg
speaker
117 appearances
6 recordings
2 series
first heard Jan 2026
last heard 20 May
Jay Goldberg’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 6 in all, peaking in Feb 2026 with 3.
Appearances
Now an AI factory is what they define any data center that's running AI workloads in.
So they have chopped and changed between that story, but its origin was that they wanted to go beyond the hyperscalers to have people own their own infrastructure for AI workloads if they were, say, a software company or an enterprise of a slightly different size.
Sadly, my history with NVIDIA before coming to Silicon Valley and covering the data center thing was gaming.
I know.
It's now such a small piece of revenue.
I mean, Jay can jump in, but really it's not where anyone's looking right now.
There is some discussion later on as it relates to the balance sheet and cash flow on things to do with supply, but you're right that what they're basically saying is that
That was the gaming was the segment that was impacted.
And I don't know that that is having any effect really on how we see the stock trading in after hours.
Well, I leave it to the analysts to model money left on the table if there were any, so to speak.
But given the outlook in terms of sales and where it came in relative to consensus, you know, NVIDIA...
it's an enviable position to be in, right, where demand exceeds your ability to supply anyway.
But going back to that five-year handbag analogy of a few moments ago, you know, the other argument that Gentleman and NVIDIA have made quite consistently is that if you tell your suppliers what you plan to do five years in advance...
then the rest of the supply chain can help prepare with you and keep up.
And, you know, recently that's been most difficult in memory, but memory shows up in pricing, you know, and if you're Jensen Wang, I'm sure you can make a phone call and make sure that the corresponding high bandwidth memory that you need is there or thereabouts.
Yeah, like ultimately, it just always comes down to a projection of confidence that all of the stories we've been over in more than the last year is still intact.
You know, that the great build out of AI infrastructure is early, that they have long horizon visibility into what's going on.
All NVIDIA can control, I suppose, is the roadmap to go from one generation of compute to the next.
And so they talk about in the CFO document margins being supported by the ramp up in Blackwell and mix of new products.
Well, when does that mix change?
Showing 61–80 of 117 · page 4 of 6
← Previous
Next →