00:00What were your initial thoughts when you saw these numbers?
00:03Thanks, Romain. I mean, very good, I guess, as expected.
00:07I think the most interesting thing we've learned so far is just this, these plans to resegment, basically, and reclassify
00:17some of the revenues going forward.
00:19Still haven't had a ton of time to think through that, but, you know, that's kind of the most unexpected
00:25thing so far.
00:25The numbers themselves, I think, definitely touch better than expectations in the quarter, right around where expectations were on the
00:34guide.
00:35Margins were solid. So I think all that seems to check out as they kind of always do.
00:40I know you haven't had a whole lot of time to think about it, but I do want to ask
00:43you, because I know the edge thing really jumped out at me here.
00:45Because to break out something that is such a tiny part of your business now seems to be a signal
00:50to the market that this is where the puck is going.
00:52And I think we've all kind of known that, right? We've kind of gotten past the phase of just using
00:56chatbots to, you know, plan our vacations and things like that.
00:58And that the broader use cases are going to start to come to the real market sometime soon.
01:04And I am curious that as somebody who is deeply invested in this space, not just in NVIDIA, but in
01:09the broader tech space here, I mean, how do you see this evolving longer term?
01:13Whether it's with whatever the new device will be that consumers will carry around, whether what we see in the
01:19industrial space, whether it's autonomous driving, where are you looking to?
01:22Well, I agree with you. I have the same kind of hunch as you do that they're obviously doing this
01:28for a reason.
01:29I do think, yeah, this is a small part of the business. Presumably this will be the faster growing part
01:35of the business.
01:36And I think there's kind of two buckets to think about this through.
01:38One, there's new applications for edge products and then there's new products.
01:44And so I think on the new applications, we already know they're in areas like autonomous driving, like earlier use
01:50cases, robotics for AI that are poised to grow really fast over the years to come.
01:55But on the new product side, I wonder if maybe some of the forthcoming standalone Vero racks will be reported
02:03here.
02:03And maybe we can start to see or get a sense for how fast NVIDIA's CPU business is about to
02:09start growing.
02:10And maybe that has implications for the rest of the space.
02:13So more so thinking out loud because it's still so early.
02:16I've just had two minutes to look at this, but that's kind of the first thing that I thought of
02:20when I saw this.
02:20Yeah, we tend to do that to guess, John, here.
02:22I mean, we only give you a couple of minutes to think about things.
02:24I am curious, though, too, we talk about a company that is also going to get a lot more competition
02:31going forward.
02:32I mean, it's kind of had this market for the high-end chips, at least, to itself.
02:35AMD is obviously trying to do what it can do.
02:37There is obviously this whatever is going on behind the doors over at Alphabet and Google here.
02:43With the potential introduction of higher-end chips that can maybe do or at least do close enough to what
02:50NVIDIA can do,
02:51do you see that potentially as a threat or is the addressable market going to be big enough for everybody?
02:57Yeah, you know, there's so many different things to talk about under this umbrella.
03:02But I think within the accelerator space specifically and specifically for training, let's break it down by workload type within
03:10AI.
03:10So I think for training, NVIDIA clearly dominates the market.
03:14I think that is very likely to continue.
03:18I don't see a ton of promise for competition in that area specifically.
03:22I think there's always going to be a place for custom, and then maybe there's going to be certain players
03:26that also want kind of a second merchant option, which is AMD.
03:30Now, for inference, I think that's where the conversation is getting a little bit more complicated.
03:34I think the inference market seems to be bifurcating between different types of inference workloads, some of which are really
03:41well-suited for NVIDIA's products, some of which are less well-suited for NVIDIA's products.
03:47One thing they did in that area to address some of those workloads is the Grok acquisition last year.
03:52We're going to hear them talk about that on the call today.
03:54I think that's going to be very important to monitor.
03:57Yeah, and I am curious, too, about, and that's a good point, I am curious, too, also about just the
04:02cash juggernaut that it is free cash flow in the most recent quarter, roughly about $49 billion.
04:08The company announcing a new buyback program, raising its dividends here.
04:12Are we going to see, do you think potentially we are going to see a little bit more of a
04:16return of cash directly to shareholders in this way?
04:18Well, I got pretty excited when I saw this new buyback authorization, an incremental $80 billion on top of another
04:26just about $40 billion that's already authorized.
04:29So by my map, that's over $100 billion that's now authorized.
04:33And then I realized that $100 billion on a $5 trillion market cap, unfortunately, doesn't really move the needle.
04:40But, no, I do think that was one of the big announcements coming out of GTC in March.
04:47They're generating so much cash, they're going to return half of free cash flow to shareholders, I believe, this year.
04:54If they continue that, there could be a trillion dollars of free cash returned to shareholders over the next four
05:01to five years.
05:02So I do think that is part of the story.
05:04And for a stock that's trading at, you know, sub-20PE or basically a market multiple, you know, I think
05:11that becomes important.
Comments