• Video
  • Nvidia Earnings Show AI Is a Long Game

Nvidia Earnings Show AI Is a Long Game

Show Transcript
Bloomberg Aug 30, 2024 08:55 · 30.6k Views Trending

Investors were disappointed at Nvidia's earnings that showed an underwhelming forecast and news of production snags with Blackwell chips. While it was a change from the blowout earnings results investors have become accustomed to, Ayako Yoshioka, portfolio manager at Wealth Enhancement Group, says the results do show that we are in the early innings of AI deployment.

Disclaimer: This content is for informational and educational purposes only and does not constitute a recommendation or endorsement of any specific investment or investment strategy. Read more

Transcript

  • 00:00 The market seemingly thinking it's a disappointment.
  • 00:03 Did you think it was a disappointment?
  • 00:06 Hi, Caroline.
  • 00:06 You know,
  • 00:07 I was not disappointed with the actual numbers here.
  • 00:10 I think, you know, expectations were very high going into to this particular print.
  • 00:17 And, you know, especially when you hear of things like
  • 00:20 events in New York in which, you know, bars were being rented out to
  • 00:24 see what the earnings report was going to be, you know, it gets to a different level.
  • 00:30 I was not in one.
  • 00:31 I was at my desk
  • 00:33 looking at the numbers.
  • 00:34 Joe Wiesenthal, I saw a Bloomberg
  • 00:37 reporter was indeed out there in the bars.
  • 00:39 But
  • 00:40 I'm interested
  • 00:41 Iko in.
  • 00:42 Even though you
  • 00:43 didn't see this overall
  • 00:45 disappointment and very high expectations, was there any cause for concern?
  • 00:49 Do you still think, as Ed put to our previous guest Anthony, that the AI trade is intact,
  • 00:55 the AI trade is intact.
  • 00:57 I think we have to put this in the longer term context.
  • 01:00 You know, NVIDIA was generating revenues of about 27 billion about two years ago.
  • 01:06 They're
  • 01:06 on pace to generate revenues of over 125 billion this year.
  • 01:11 That's just enormous.
  • 01:13 Is that growth rate sustainable?
  • 01:15 You know
  • 01:16 it, it clearly is not,
  • 01:19 but they are continuing to grow and the expectations are for growth of over 40% in revenues for next year
  • 01:26 and Blackwell is just going to be another step up in allowing that to happen.
  • 01:32 Ioco I Wealth Enhancement Group has like what 7.8 million NVIDIA shares, like a billion dollars worth of stock.
  • 01:38 I know
  • 01:39 that you were part of our kind of watch party last night, watching the interview.
  • 01:44 You lead research, right?
  • 01:45 You use your expertise to try and work out what's going on.
  • 01:49 What did you work out during the course of that conversation?
  • 01:54 You know, I think this is a long game.
  • 01:56 I think we're still in the early innings of AI deployment.
  • 02:00 The hyper sailors meta.
  • 02:02 There's so much use from.
  • 02:05 From all of these, it is table stakes
  • 02:07 for these large cloud providers
  • 02:10 and they're seeing productivity gains
  • 02:13 from the use of AI.
  • 02:15 So I think that's going to continue.
  • 02:16 It's going to proliferate.
  • 02:17 It is going to take time,
  • 02:19 but.
  • 02:20 You know, the Internet took time as well,
  • 02:23 and 20 years later, it's still.
  • 02:25 Evolving.
  • 02:26 Were you convinced
  • 02:28 by what Jensen Wong just said?
  • 02:29 We played in that clip just then that their business has broadened out beyond the hyperscalers.
  • 02:37 Yes, I mean they mentioned about, you know sovereign entities
  • 02:41 purchasing and being in that low double digit billions
  • 02:45 of dollars in terms of revenues.
  • 02:47 You know that was exciting to see
  • 02:49 and you know hyperscalers were 45%, which means 55%
  • 02:53 is other revenues.
  • 02:55 So
  • 02:55 that is
  • 02:56 a nice diversification as well.
  • 02:59 But that consistency from the hyperscalers is likely to continue.
  • 03:03 It's just going to be choppy year to year.
  • 03:05 You know many of these companies, Apple, Microsoft,
  • 03:09 Amazon.
  • 03:10 Google, they're all going to be evaluating their Cap X budgets for next year.
  • 03:15 And each year it's not going to be the same.
  • 03:17 And so we have to be diligent about.
  • 03:20 About that, but for the most part they will be spending in order to get that product productive.
  • 03:25 The enhancement going forward,
  • 03:27 what about competition?
  • 03:29 A lot of people feeling that if there is just a slight blip in this annual cadence of innovation,
  • 03:34 maybe AMD can start to work its way in a little bit more.
  • 03:38 Do you agree
  • 03:40 or you know, I think from a hyperscaler perspective or from the cloud providers, the Internet service providers, you always want to have more than just one provider of your chips.
  • 03:51 So, you know, there is room for competition.
  • 03:53 I think,
  • 03:54 you know, that's always something that we have to keep in mind.
  • 03:58 You know,
  • 03:58 let's not forget that Intel had the majority of, you know, data center server share for for many, many years.
  • 04:06 And here we are, with NVIDIA
  • 04:08 really coming out and taking.
  • 04:10 Can share AMD as well, so competition is always something that we look for.
  • 04:15 And watch out for them cautionary tale.
  • 04:17 What's interesting is you have holds on Texas Instruments on Broadcom.
  • 04:21 What many of thought perhaps hasn't
  • 04:23 moved at the pace that we're expecting was that these hyper scalers would make their own chips to compete in quite the same way using a Texas Instruments using
  • 04:31 CSMC
  • 04:32 for example.
  • 04:34 I'm interested in
  • 04:35 if, if ever you see that really coming in as a competition,
  • 04:39 you know,
  • 04:40 it's, it's tough.
  • 04:41 You know, the, the design part of those chips is, you know, where NVIDIA has the advantage
  • 04:47 along with the software associated
  • 04:50 with AI.
  • 04:51 So you know, their CUDA,
  • 04:54 you know, all the chips are built off of CUDA and, and that really provides them with the benefit that many other chip providers and chip manufacturers just don't have.
  • 05:05 Ayoko, can you define
  • 05:07 what NVIDIA is, what it actually is?
  • 05:12 You know
  • 05:13 NVIDIA is
  • 05:14 a chip maker at the end of the day, right?
  • 05:17 But it is a chip maker and is a primary beneficiary for artificial intelligence and how computing will go forward in the coming decade.