WEBVTT - Arm CEO Rene Haas Talks Earnings

0:00:02.360 --> 0:00:07.560
<v Speaker 1>Bloomberg Audio Studios, podcasts, radio news now to a major

0:00:07.600 --> 0:00:12.080
<v Speaker 1>company story, Armstock. In the headlines, it's posting record quarterly

0:00:12.200 --> 0:00:16.160
<v Speaker 1>revenue but issuing cautious guidance for the current quarter, citing

0:00:16.160 --> 0:00:18.919
<v Speaker 1>the timing of new licensing deals in particular. Meanwhile, the

0:00:18.920 --> 0:00:21.960
<v Speaker 1>Trump administration pans to roll back Biden Nearer ai Chip

0:00:22.040 --> 0:00:25.440
<v Speaker 1>curbs with a new export control framework in the works,

0:00:25.440 --> 0:00:27.640
<v Speaker 1>and that may impact things too. So joining us now

0:00:27.640 --> 0:00:30.600
<v Speaker 1>to discuss it all as ARMS CEO Renee has along

0:00:30.600 --> 0:00:33.000
<v Speaker 1>with Bloomberg Tech co host Carolyn Hyde.

0:00:33.000 --> 0:00:34.239
<v Speaker 2>Carolyn, thank you.

0:00:34.240 --> 0:00:37.120
<v Speaker 3>Bonnie and Renee. It's always wonderful to welcome you to

0:00:37.159 --> 0:00:40.160
<v Speaker 3>Bloomberg Television. And let's just dwell on your numbers, because

0:00:40.200 --> 0:00:42.920
<v Speaker 3>revenue strong, up thirty four percent for your fiscal fourth quarter.

0:00:42.960 --> 0:00:45.159
<v Speaker 3>We saw more than a billion dollars being brought in

0:00:45.200 --> 0:00:46.280
<v Speaker 3>and indeed a profit beat.

0:00:46.320 --> 0:00:48.839
<v Speaker 1>But investors taken with what seems to be.

0:00:48.800 --> 0:00:52.680
<v Speaker 3>This cautious outlook, how uncertain are things from a demand

0:00:52.720 --> 0:00:56.120
<v Speaker 3>perspective that also perhaps you're not giving that annual target

0:00:56.120 --> 0:00:56.880
<v Speaker 3>that someone had hoped for.

0:00:57.960 --> 0:01:00.160
<v Speaker 4>Yeah a hi, Carolin, and thank you for the the

0:01:00.240 --> 0:01:04.200
<v Speaker 4>kind words. Yes, we just came off an amazing end

0:01:04.200 --> 0:01:07.559
<v Speaker 4>of our fiscal year, the first billion dollar quarter, which

0:01:07.600 --> 0:01:11.360
<v Speaker 4>was a record record royalties over six hundred million dollars,

0:01:11.440 --> 0:01:14.840
<v Speaker 4>record licensing over six hundred million dollars, and then we

0:01:14.920 --> 0:01:17.600
<v Speaker 4>finished the year at four billion dollars, which was also

0:01:18.120 --> 0:01:19.280
<v Speaker 4>a record for the company.

0:01:20.160 --> 0:01:22.880
<v Speaker 2>For Q one, we're looking at very good royalty growth.

0:01:22.920 --> 0:01:26.840
<v Speaker 4>Actually, Caroline, we're projecting anywhere between twenty five to thirty percent.

0:01:27.480 --> 0:01:29.360
<v Speaker 4>On the licensing side, I think, as you know, the

0:01:29.400 --> 0:01:32.560
<v Speaker 4>deals can be a little bit lumpy, So as a

0:01:32.600 --> 0:01:35.800
<v Speaker 4>result of looking forward in terms of whether deals closed

0:01:35.840 --> 0:01:39.039
<v Speaker 4>in the June or July timeframe, we decided to be

0:01:39.080 --> 0:01:41.360
<v Speaker 4>cautious and give the outlook that we did just based

0:01:41.400 --> 0:01:44.600
<v Speaker 4>upon licensing. But the overall health of the business is

0:01:44.640 --> 0:01:48.800
<v Speaker 4>really really good. You know, we have obviously coming off

0:01:48.800 --> 0:01:51.640
<v Speaker 4>a great quarter and we have really good visibility into

0:01:51.720 --> 0:01:52.320
<v Speaker 4>the current quarter.

0:01:52.360 --> 0:01:53.080
<v Speaker 2>On royalties.

0:01:53.680 --> 0:01:56.560
<v Speaker 4>The reason we chose not to give the yearly guidance thing,

0:01:56.640 --> 0:02:00.600
<v Speaker 4>as you know, now vast majority of our businesses increasinglyyalties,

0:02:00.680 --> 0:02:04.080
<v Speaker 4>and that really comes from people who make either the

0:02:04.280 --> 0:02:07.760
<v Speaker 4>end ship or the end equipment, and that in the

0:02:07.800 --> 0:02:10.440
<v Speaker 4>next two to three quarters is a little hard to predict,

0:02:10.440 --> 0:02:12.680
<v Speaker 4>and in fact, we're not seeing that guidance coming from

0:02:12.680 --> 0:02:13.320
<v Speaker 4>our partners.

0:02:13.720 --> 0:02:14.320
<v Speaker 2>So as a.

0:02:14.280 --> 0:02:18.120
<v Speaker 4>Result, given that their visibility isn't as strong as it

0:02:18.160 --> 0:02:21.440
<v Speaker 4>usually is around that we decided not to go forward

0:02:21.480 --> 0:02:24.160
<v Speaker 4>and give a guidance beyond justice first quarter, and.

0:02:25.040 --> 0:02:27.240
<v Speaker 3>Just for the people who always need to be reminded

0:02:27.240 --> 0:02:29.760
<v Speaker 3>of the business model you are designing. Ultimately, people that

0:02:29.840 --> 0:02:33.519
<v Speaker 3>are licensing using these designs, you are not directly impacted

0:02:33.919 --> 0:02:36.200
<v Speaker 3>by tariffs, but of course the people that make the

0:02:36.240 --> 0:02:39.640
<v Speaker 3>smartphones eventually might be. Are you hearing anything about a

0:02:39.680 --> 0:02:42.080
<v Speaker 3>worry on demand in the future. Are you thinking that

0:02:42.160 --> 0:02:44.080
<v Speaker 3>we'll will see a downward trajectory for some of the

0:02:44.120 --> 0:02:47.200
<v Speaker 3>electronics parts that you've diversified the business into.

0:02:48.080 --> 0:02:50.239
<v Speaker 2>Yeah, thank you for pointing that out.

0:02:50.280 --> 0:02:53.440
<v Speaker 4>Yet exactly, we are a provider of the designs that

0:02:53.520 --> 0:02:57.440
<v Speaker 4>go into chips that go into end products like your iPhone,

0:02:57.919 --> 0:03:01.000
<v Speaker 4>your Ford f one fifty, your earbudd it's the AI

0:03:01.120 --> 0:03:04.240
<v Speaker 4>data centers. So we're not directly impacted, but we are

0:03:04.280 --> 0:03:08.079
<v Speaker 4>certainly indirectly impacted, and that would come on the royalty

0:03:08.120 --> 0:03:10.359
<v Speaker 4>side to your question, are we.

0:03:10.360 --> 0:03:14.800
<v Speaker 2>Seeing anything relative to folks?

0:03:15.120 --> 0:03:17.919
<v Speaker 4>Not clear in terms of where the terrorifts are coming from.

0:03:18.240 --> 0:03:20.280
<v Speaker 4>What we're seeing is kind of a lack of clarity

0:03:20.680 --> 0:03:23.280
<v Speaker 4>as opposed to any one indicator or the other. So

0:03:23.280 --> 0:03:26.040
<v Speaker 4>that's just really driving and it's just the lack of

0:03:26.120 --> 0:03:28.080
<v Speaker 4>visibility that our end partners have.

0:03:28.480 --> 0:03:29.000
<v Speaker 2>Ren I have.

0:03:29.000 --> 0:03:33.239
<v Speaker 1>Any plans on changing how you actually charge your royalties,

0:03:33.280 --> 0:03:36.040
<v Speaker 1>maybe you'll move into a per device basis. And also

0:03:36.080 --> 0:03:39.160
<v Speaker 1>how much pricing power do you have in an inflationary environment?

0:03:40.720 --> 0:03:43.800
<v Speaker 4>Yeah, thanks for asking that question. What we have done

0:03:43.920 --> 0:03:47.680
<v Speaker 4>is transition to more comprehensive products what we call our

0:03:47.680 --> 0:03:51.480
<v Speaker 4>compute subsystems, and think of that as a design that

0:03:51.600 --> 0:03:54.080
<v Speaker 4>not only are we licensing the individual blocks of IP,

0:03:54.960 --> 0:03:57.240
<v Speaker 4>but all of the other blocks of IP required to

0:03:57.240 --> 0:03:59.480
<v Speaker 4>make a subsystem that would go on a chip.

0:04:00.280 --> 0:04:03.000
<v Speaker 2>Those command higher royalties.

0:04:03.360 --> 0:04:06.760
<v Speaker 4>Generally double from what we see on traditional individual blocks.

0:04:07.160 --> 0:04:10.600
<v Speaker 4>The benefit to the customers is it really accelerates time

0:04:10.640 --> 0:04:14.720
<v Speaker 4>to market. Chip development time goes down, huge value that

0:04:14.760 --> 0:04:16.640
<v Speaker 4>they see in terms of predictability of the design.

0:04:17.120 --> 0:04:18.000
<v Speaker 2>So that's what.

0:04:18.000 --> 0:04:20.160
<v Speaker 4>We've been shifting to much more in terms of our

0:04:20.560 --> 0:04:23.240
<v Speaker 4>royalty model, which is why the royalty growth has been

0:04:23.520 --> 0:04:24.520
<v Speaker 4>as good as it has been.

0:04:25.320 --> 0:04:28.560
<v Speaker 3>I'm really interested in the AI data center part because

0:04:28.560 --> 0:04:30.880
<v Speaker 3>this has been a real era of focus of growth,

0:04:30.920 --> 0:04:33.520
<v Speaker 3>and I'm interested in the momentum that you're seeing and

0:04:33.520 --> 0:04:35.919
<v Speaker 3>how resilient it is. You're talking about uncertainty when it

0:04:35.920 --> 0:04:38.560
<v Speaker 3>comes to smartphones and electronics in the future. Has there

0:04:38.560 --> 0:04:41.960
<v Speaker 3>been any signal of uncertainty on spending on AI infrastructure.

0:04:43.760 --> 0:04:45.800
<v Speaker 2>Yeah, not that we've seen, Caroline.

0:04:46.440 --> 0:04:49.719
<v Speaker 4>You've got Stargate obviously, a huge program that's being done

0:04:49.720 --> 0:04:52.160
<v Speaker 4>between soft Bank and open Ai, which is all around

0:04:52.520 --> 0:04:56.279
<v Speaker 4>capital and power for data centers. You're seeing large capex

0:04:56.320 --> 0:05:00.320
<v Speaker 4>spending by the major hyperscalers. But probably more important, certainly,

0:05:00.880 --> 0:05:04.159
<v Speaker 4>I think we're nowhere close to really good enough in

0:05:04.240 --> 0:05:07.320
<v Speaker 4>terms of the capability what AI can do. Obviously, the

0:05:08.040 --> 0:05:10.359
<v Speaker 4>advances have been amazing, and if you look at just

0:05:10.400 --> 0:05:12.880
<v Speaker 4>what OpenEye has done with chat Gypt, it's been remarkable,

0:05:13.800 --> 0:05:17.960
<v Speaker 4>but enterprise usage from a mass deployment standpoint is still

0:05:18.040 --> 0:05:21.040
<v Speaker 4>quite modest. A factoid that I like to refer to

0:05:21.160 --> 0:05:24.479
<v Speaker 4>is chat GPT used a petabyte.

0:05:24.000 --> 0:05:25.840
<v Speaker 2>Of data to create the model.

0:05:26.240 --> 0:05:28.600
<v Speaker 4>JP Morgan has one hundred and fifty petabytes of data

0:05:28.600 --> 0:05:31.000
<v Speaker 4>sitting inside their enterprise. In one way, shape or form,

0:05:31.520 --> 0:05:34.880
<v Speaker 4>harnessing that data, turning that data into real valuable information,

0:05:35.320 --> 0:05:38.599
<v Speaker 4>and then helping productivity. There's a lot of white space there,

0:05:38.680 --> 0:05:41.840
<v Speaker 4>and I think still there's a long way to see

0:05:41.839 --> 0:05:42.400
<v Speaker 4>growth here.

0:05:42.640 --> 0:05:45.880
<v Speaker 3>There's been some uncertainty, of course, around ultimately how those

0:05:45.920 --> 0:05:48.960
<v Speaker 3>that use your designs can ultimately sell their chips. And

0:05:49.160 --> 0:05:51.800
<v Speaker 3>we've now got a pullback from the AI diffusion rule

0:05:51.800 --> 0:05:53.320
<v Speaker 3>as it's called here in the United States. It would

0:05:53.320 --> 0:05:55.520
<v Speaker 3>seem that's going to be announced anytime soon by the

0:05:55.520 --> 0:05:57.800
<v Speaker 3>Trump administration. Is that going to be a boost for

0:05:57.800 --> 0:06:00.480
<v Speaker 3>your business? Do you think that leads to most You're

0:06:00.480 --> 0:06:03.040
<v Speaker 3>actually lack of it, because now we have countries having

0:06:03.040 --> 0:06:05.040
<v Speaker 3>to negotiate on a case by case basis.

0:06:05.600 --> 0:06:07.920
<v Speaker 4>Yeah, I don't know that I could predict it. I mean,

0:06:08.040 --> 0:06:11.120
<v Speaker 4>actually seen what the what the outcome looks like. The

0:06:11.160 --> 0:06:14.680
<v Speaker 4>wonderful thing about our jobs these days, CEOs, that we're

0:06:14.760 --> 0:06:17.799
<v Speaker 4>reading the news and hearing that from you at real time.

0:06:18.279 --> 0:06:20.479
<v Speaker 4>I think we'll just have to see, is the short answer.

0:06:21.240 --> 0:06:25.880
<v Speaker 4>I don't think AI demand is a bubble in any way,

0:06:25.920 --> 0:06:27.560
<v Speaker 4>shape or form. And as I said, I don't think

0:06:27.600 --> 0:06:31.360
<v Speaker 4>we've gone anywhere close to reaching its true potential.

0:06:32.080 --> 0:06:33.240
<v Speaker 2>When you think about.

0:06:33.320 --> 0:06:37.160
<v Speaker 4>End devices that will run AI models, everything will run them.

0:06:37.800 --> 0:06:40.320
<v Speaker 4>They'll run in earbuds, they'll run in cars, they'll run

0:06:40.320 --> 0:06:43.000
<v Speaker 4>on mobile phones. They're certainly going to run in the cloud,

0:06:43.320 --> 0:06:46.760
<v Speaker 4>but they won't run exclusively in the cloud. And I

0:06:46.800 --> 0:06:49.760
<v Speaker 4>think we still have a long way to go in

0:06:49.880 --> 0:06:55.440
<v Speaker 4>this area. The progress has been remarkable, shockings in some

0:06:55.520 --> 0:06:58.800
<v Speaker 4>ways just how good the models have become, but there's

0:06:58.800 --> 0:07:01.159
<v Speaker 4>still a long ways to go in terms of I

0:07:01.200 --> 0:07:03.679
<v Speaker 4>think true AGI and then ultimately ASI.

0:07:04.200 --> 0:07:07.039
<v Speaker 1>Ronnie, your top five customers are responsible for sixty percent

0:07:07.040 --> 0:07:09.400
<v Speaker 1>of revenue and are in China twenty percent of that.

0:07:09.920 --> 0:07:13.120
<v Speaker 1>I'm curious as to how you're making an effort to diversify.

0:07:13.200 --> 0:07:16.800
<v Speaker 1>That's a phenomenally concentrated client base, particularly when you consider

0:07:16.840 --> 0:07:20.280
<v Speaker 1>the China. You know, you might have some concerns around China.

0:07:21.720 --> 0:07:25.280
<v Speaker 4>Yeah, the industry that we're in has been seeing, you know,

0:07:25.360 --> 0:07:28.000
<v Speaker 4>quite a bit of consolidation. You know, for quite some time.

0:07:28.280 --> 0:07:33.760
<v Speaker 4>The amount of capital required to make chips is quite

0:07:33.760 --> 0:07:35.160
<v Speaker 4>a bit. So you look at the number of companies

0:07:35.200 --> 0:07:39.160
<v Speaker 4>who are their end products are chips. Its shrink over

0:07:39.160 --> 0:07:41.280
<v Speaker 4>the last number of years. And at the same time,

0:07:42.000 --> 0:07:45.560
<v Speaker 4>you've got large hyperscalers in the Microsoft's, the Metas, the Googles,

0:07:45.560 --> 0:07:48.640
<v Speaker 4>to Amazons that you know, there's not many people have

0:07:48.720 --> 0:07:52.280
<v Speaker 4>that capital. That being said, the number of chips, the

0:07:52.320 --> 0:07:54.640
<v Speaker 4>demand for chips the compute has actually gone way up,

0:07:55.040 --> 0:07:57.880
<v Speaker 4>So I probably don't think so much about the number

0:07:57.920 --> 0:08:00.400
<v Speaker 4>of people who buy. I would get more worried if number,

0:08:00.560 --> 0:08:03.520
<v Speaker 4>if the quantity was going down, and we're not seeing that,

0:08:04.120 --> 0:08:05.360
<v Speaker 4>and why are we not seeing that.

0:08:06.000 --> 0:08:08.239
<v Speaker 2>The insatiable demand for more and more.

0:08:08.080 --> 0:08:13.120
<v Speaker 4>Compute driven by AI, I think again, is we're in

0:08:13.240 --> 0:08:14.320
<v Speaker 4>very early stages of it.

0:08:14.480 --> 0:08:15.120
<v Speaker 2>You also have to.

0:08:15.040 --> 0:08:18.640
<v Speaker 4>Remember that running these AI workloads, you still need to

0:08:18.680 --> 0:08:21.160
<v Speaker 4>run them on top of running a Windows operating system,

0:08:21.240 --> 0:08:24.240
<v Speaker 4>on top of running an iOS operating system, on top

0:08:24.280 --> 0:08:28.360
<v Speaker 4>of running an IBI instrumentation inside a car or autonomous

0:08:28.640 --> 0:08:32.520
<v Speaker 4>So it's not that AI compute replaces traditional compute. It's

0:08:32.600 --> 0:08:35.640
<v Speaker 4>on top of which is a strong demand signal for

0:08:35.840 --> 0:08:38.559
<v Speaker 4>personmic conductors, and I has the m CEO.

0:08:38.840 --> 0:08:40.640
<v Speaker 3>It's going to enjoy having you. Thank you very much.

0:08:40.720 --> 0:08:41.080
<v Speaker 2>Indeed,