Clicky

Oracle Earnings, IBM LinuxOne Emperor 4, Adobe-Figma, Cadence AI, Tesla FSD, Arm Neoverse V2 – The Six Five Webcast
by Daniel Newman | September 21, 2022

On this episode of The Six Five Webcast, leading global tech analysts Daniel Newman and Patrick Moorhead analyze the tech industry’s biggest news each and every week.

On this week’s show we will be discussing:

  1. Oracle Earnings
  2. IBM LinuxONE Emperor 4 Announcement
  3. Adobe Figma + Earnings
  4. Cadence Verisium and JedAI Platform Announcements
  5. Tesla Self Driving Under Fire
  6. Arm Neoverse V2 Announcement

For a deeper look into each topic, please click on the links above. Be sure to subscribe to The Six Five Webcast so you never miss an episode.

Watch the episode here:

Listen to the episode on your favorite streaming platform:

Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we do not ask that you treat us as such.

Transcript:

Patrick Moorhead: Hi, this is Pat Moorhead with more insights and strategy and we are here for another Six Five podcast. This is Pat Moorhead. I already said that. I must be on vacation in an undisclosed location in Colorado, but I am here with my bestie Daniel Newman of Futurum Research. How are you my friend?

Daniel Newman: Not on vacation, but it makes me really happy to pull you off of your time away and make sure I suck you back into the work vortex. But that’s only because as your bestie, I know you actually get a lot of joy out of work. So those tweets about, “Oh, woe, is me? I’m working on vacation.” I actually believe it’s sanity for you to put in a couple hours every day because you love what you do. Just like getting on this pod, it’s the best part of your week. Hopefully no one in your family’s listening, but that’s just what’s up.

Patrick Moorhead: No, but they all left this morning and they’re watching a balloon show. So hot air balloon show as dad.

Daniel Newman: Yeah, you’d much rather be at hot air balloon thing than do the pod with me, right?

Patrick Moorhead: To be honest with you, getting up at 6:00 AM and seeing some hot air balloons, I’d rather be here. But no, this is this, in all seriousness, this is my favorite time of the week, and if this is your first time here, we cover six topics for five minutes each, maybe 10 if we feel like, because it’s our show so we can do whatever we want. But in reality we do what our fans want, which sometimes is talking between 5 and 10. We do talk about publicly traded companies, like we’re going to talk about Oracle today and their earnings. Please don’t take any of this as investment advice.

But no, we do have a great show. I mean there was a lot of stuff happening Daniel, we had to figure out what do we put in, what do we prioritize, what do we kick to next week? But we’re going to be talking Oracle earnings, IBM LinuxONE announcement, Adobe earnings and acquisition maybe going, trying to go after Figma. We’re going to talk about a big Cadence announcement, auto AI basically. We’re going to talk about the latest round of Tesla self-driving issues. And then finally ARM had a big disclosure on its infrastructure. We’re going to dive in, but let’s kick off here with Oracle earnings. Could Oracle do it again in the cloud, Daniel?

Daniel Newman: Well, let’s just say Oracle had a pretty good quarter. It was a mixed result on top and bottom, but overall, given the global macro situation, Oracle continues to impress. I’ll spend less than one minute because this isn’t really news, this is analysis, but just a quick minute on, hey, the company did well, it did great on revenue, it did good on the bottom line. It did add creatively a component of the Cerner acquisition, which made the growth look a little bigger than it really was. But when you pull that out, it was still nearly 10% organic with constant currency, which given the macro climate right now, I thought was really good.

But I went on Yahoo Finance, I did a little tv spotty, spotty, and my comment was basically, of course you were away. You were away. So they had to bring in the B team. And so my comment was really simple, Pat. It was this is a cloud story. This quarter for Oracle is a cloud story. The Cerner acquisition is a cloud story, it’s a healthcare cloud and we’ve all talked a lot about the importance of industry and vertical clouds, but it’s really a good for you Oracle for starting to provide a level of transparency on your cloud business that sort of backs Larry Ellison’s longtime comments about this company making this big transformation.

You don’t know the background, Oracle, about 70% of its revenue is totally recurring in the sense that it’s all tied to its core and legacy business, which has always made it a very stable and solid business to invest in during periods like this with a high percentage of that revenue accounted for. And then of course consistent margins, dividend, buybacks, the company does all the things that make investors feel more secure in these really volatile times.

But this is a growth story here. And I think that’s really what has made a lot of us as analysts excited. Now they asked me the question straight away, “What’s up with the Oracle cloud?” And I think you and I have said this 20 times, so I’m just going to say it really quickly. The evolution from the original Oracle cloud to now has been impressive. It’s been very good and that’s why the company is seeing more wins. And then the company obviously has split it’s cloud story into really two buckets. It’s an infrastructure story and a SaaS story. And now they’re basically able to disclose very valuable data on where they’re at. So the cloud revenue is going up substantially, 45% this quarter, $3.6 billion. So as I said, I think one or two quarters ago, it’s more than a $10 billion a year business. That’s pretty good.

They also gave some data on the breakdown, so the infrastructure business, everybody always wants to know how much infrastructure is being sold. Well it looks like Oracle’s trajectory on the cloud infrastructure as IS business is about 4 billion a year right now. They had 0.9 this quarter and it was a 52% increase year over year, 58% constant currency. And then of course for a longer period of time we’ve been tracking the application revenue, the SaaS revenue, and that fills the rest of the gap. 2.7 billion, 43% up, 40% in constant currency with strong back numbers from Fusion at over 30% in NetSuite in the high 20 percentages of growth. So Oracle is a cloud story.

Now this quarter was robust, guidance was solid, the economy sucks. And so right now it really doesn’t matter. And if you’re good, if you’re bad, everything that was good a year ago is bad now. But what I would say is if you’re looking for a company that has a combination of stable revenue, good margins, is going to be robust and necessary in any economy, Oracle is telling that story and they’re kind of giving a little bit of a cherry on top by adding a higher growth cloud story that investors and technology enthusiasts should be excited about. Let’s just put it this way, this ain’t your mama’s Oracle anymore?

Patrick Moorhead: You pretty much sucked all the oxygen.

Daniel Newman: Come on. I was like two minutes. I only talked for two minutes.

Patrick Moorhead: No, no, you hit exactly what I want to hit on. But I guess what I would say is hell does freeze over, as well. I mean he’s cutting deals with AWS to put HeatWave in the AWS cloud and HeatWave, if you’re not familiar with it, is a no SQL database that can either be on-prem, could be the Oracle cloud, can be an AWS, and it puts a lot of the data in real memory and gets a hell of a lot of performance out of it. And here we have, if you remember, there are two companies that AWS makes fun of and picks on in all of their reinvent conferences. The biggest one is Oracle and the second one is IBM.

And if you saw OCI, which is their IS platform, OCI 1.0, it really was not good. It led in nothing. It was expensive. Larry brought in some folks from, quite frankly, he hired them from AWS and came in and built OCI 2.0, which is completely serverless and he moved his applications, completely rewrote his on-prem applications that then turned into Fusion. So, that 52% growth in IaaS is pretty much the highest growth rate in the top five clouds. And I know people will say, “Oh, it’s not that big”, but it’s at least 5 billion of IaaS, which is super impressive if you see that OCI 2.0 really only started two and a half, three years ago. They’ve literally only been in business for two and a half or three years with something that was even slightly competitive. So hats off to Oracle. I don’t fully understand why they’re stock skid more than most. Maybe investors were looking for a bigger bump from Cerner. Maybe not.

Daniel Newman: It’s the economy Dude, it sucks. The economy sucks.

Patrick Moorhead: They dropped lower than others.

Daniel Newman: Well you remember the CPI came out the next day, so we got the inflation data the next day and while you were on vacation. I’m not saying that cynically, I’m saying that took the whole market down. It’s such a huge magnitude. You want to talk about getting crushed, wait till we talk about Adobe. So I’m just saying I think, I do, I think it’s CPI. I think the inflation is way more persistent and we don’t have any strategy but a blunt force, freaking sledge hammer to crush demand and it’s going to take tech with it and I’m not happy about it in case you couldn’t tell.

Patrick Moorhead: No, no, I appreciate that. And you’re right, even though I’ve been working four or five hours a day-

Daniel Newman: I’m teasing dude.

Patrick Moorhead: No, but bro, we all take, do what we do. So it’s all good. Hey, let’s move to the next topic, which is IBM announced its latest Linux based mainframe called the Emperor 4. And you might say, “Well why talk about mainframes?” Well, mainframes are more popular today than they were 10 years ago and there are a lot of reasons for that. So I’m just kind of giving the backstory here. But LinuxONE obviously runs Linux. We’re in the fourth generation of this and it essentially uses the guts of the latest Z16 but uses a very different software ecosystem. It can run containers from Red Hat, it can run serverless applications, it can run your favorite Linux applications. It truly is incredible.

And the other big change that the mainframe platform has taken is that’s integrated AI on the chip has the highest levels of security, quite frankly. It’s the only hardware platform that supports the highest level of FIPs security. But two things stood out for me and that was even though the company wanted to go all in on sustainability, which I do think the boards of directors are figuring out how do we hit our goals? But what was really crazy is the consolidation story. And I looked at the data and it looked legit to me of what you might find in a current ecosystem. But essentially you can consolidate 55 Intel servers into one LinuxONE box sucking 62% less energy and taking up 86% less data center floor space. That fricking blows my mind.

So I think when clients are looking at, “Hey, do I bring in Sapphire Rapids or do I take a chance on this new LinuxONE box?” And only say take a chance as it relates to non clients or non-customers. And I wish IBM would put a little bit more effort into going into new business as opposed to upselling their current clients. That to me is a lot more interesting. And I don’t mean that this isn’t interesting, but I think that they would turn a lot more heads if they would hire a completely new sales force to go after new logos and a big one.

So I was really impressed. You can go on their website and plug in what your infrastructure looks like and it spits out a TCO calculator. It also helps you potentially go to the board or go to the executive staff if you’re trying to hit your company’s sustainability goals and say, “Hey, let’s do this.” And I don’t really thought about this companies like Salesforce and Mark Benioff are always talking about how green they want to be. They’re the leaders in green and this is what they’re doing. Shouldn’t companies take a chance if they’re currently using XA 6 servers today, either on-prem or in the cloud, take a chance on this LinuxONE box that by the way fits in a standard rack, fits in standard power. You don’t need water cooling, you can use air cooling just like all of your XA 6 stuff. If you’re running Linux apps, you can run them on this platform, freaking consolidate 55 Intel servers. That’s pretty cool.

Daniel Newman: Yeah, I mean I could think of a company that may not be a huge fan of that idea, but we do have to be somewhat as analysts have to, we have to set shape to the market. In this case, there’s this pretty significant argument to be made. Sustainability is at the top of every company’s mind. You go to an event right now, you will not hear a keynote where a company does not lean into how they’re helping Mother Earth, carbon net zero. And the bottom line is really the bottom line.

As we get into an era of measurement, we’re going to have to measure what companies are doing and what companies could be doing. So future analyst, Steven Dickens drove into this one. So I kind of went through his analysis. He provided some of the feedback during the development. Pat, you really hit the sustainability thing hard. I think that’s an area to lean in. I think you hit the scalability thing hard. The fact that again, you can consolidate a lot of investments and I like what you said about the sales side. I do think IBM has sort of been like, this is our little universe of customers and we’re going to keep selling them the next box every time the next box comes out and it creates the supercycle.

But what if you had that a bigger business development cycle that could drive more new adoption in between? And I guess that is the question of are new customers open to the idea of this type of technology versus what would be considered more traditional or Nuvo cloud architectures. Now IBM and its whole strategy about building this true hybrid cloud, that’s a little bit more full stack, means that IBM’s building LinuxONE to be compatible to not just the IBM cloud but to other hyperscale cloud and that company’s been very actively working to build partnerships that can create a fabric that enables companies that use a Google or an Amazon or Microsoft to be able to work strategically with their traditional mainframe hardware.

The only thing I’d add, Pat, is I thought there was a pretty impressive, you kind of touched on chip integrated accelerator. That seems to be a big focus following the Telem announcement that came out in the last generation of the Z box. So they’re continuing to add AI and inference features on chip in the LinuxONE so that doesn’t have to be done outside of the core infrastructure.

And then the security, Pat, I mean that’s the only thing I think we didn’t touch on too much here. Companies really showed some key focus on security, quantum safe capabilities, key generation encryption, key encapsulation mechanisms, hybrid key exchange schemes. And this is all the stuff Steven pointed out, but a number of cryptographic enhancements, a lot of focus on security, which anyone that listens to this that knows about mainframes knows security’s been probably the number one reason that workloads have not moved off of mainframes yet is that they’re beyond being reliable and scalable and available, they’re secure. And so IBM continues to lean into that. Good on them. Good on Ross and the team. Congratulations on the launch.

Patrick Moorhead: Yeah, just to add to the security piece, it’s the only box out there that’s not only quantum safe, supports confidential computing. And by the way, quantum safe, I don’t use lightly, right? Not like cloudification or something like that. Essentially what’s built in is in the future you can upgrade your security to be literally safe from quantum based attacks. And that could be five years in the future, that could be 10, that could be 10 years in the future. And the company brought out a really interesting stat, which says you can execute up to 20 billion secure transactions per day based on, with a microservice based application running on Red Hat OpenShift and it just blows my mind. But anyways, great ads there. And Steve might know a little bit about this stuff given that he used to be in charge of mainframe sales, at IBM or at least, or at least an executive in there.

Daniel Newman: He led the offering on the LinuxONE in the Z box. But let’s just say that there’s no one more passionate about it than I’ve ever met maybe, but Ross. So it was definitely a very in depth, I’ll put it in the show notes, Pat, but you wrote a great piece on it too. You put a research report out and that’ll be in the notes as well.

Patrick Moorhead: Hey, I appreciate you helping me plug my own research report.

Daniel Newman: I always plug you dude, I always plug you.

Patrick Moorhead: Yeah, I’ll have Forbes article coming out sometime after I get back from vacation.

Daniel Newman: Read his Forbes article, it’ll be awesome.

Patrick Moorhead: Yeah, I appreciate that. So hey, let’s move to the next topic here. This is going to be a fun one. We really have two topics in one and that is that Adobe is looking at buying Figma for $20 billion and then Adobe also pre-announced by a couple hours the earnings. Daniel, why don’t you take this one?

Daniel Newman: Yeah, so I’m going to hit the earnings in about 15 seconds and let’s talk about Figma. The earnings were really good and no one cared because the Figma deal came out at the same time. And when you make a deal with some of the stats, and I’ll talk about them in a minute and I’m going to play the bull here because I think you’re going to help me play the bear a little bit and we can come around, but the earnings were good. The company beat top and bottom their guidance, given the kind of overall guidance adjustments we’ve seen this quarter from a lot of companies, their guide was still pretty good, still going to beat on the bottom line, maybe a slight pullback on the overall revenue expectation, called out the headwinds, but still a ton of demand. I mean this company is a company that enables digital. So it doesn’t have a big physical presence requirement for it to gain more demand in fact.

But as things like eCommerce slow, that could affect it. But at the same time as holidays roll in, that’s a tailwind for a company like Adobe. So the fact that it came down 10% to prepandemic lows, I want to point that out, almost as if the company is not grown and it has. Record revenue this quarter for Adobe.

Now let’s just talk about Figma though. For those not familiar, I’m going to really overly simplify this, but if you’re familiar with visualization software like DESO or 3DEXCITE that’s used for creating products in a collaborative space, this is the collaborative whiteboard for creating a digital experience. So when you’re designing your Netflix, this is a go in and have an area that everyone can work and codesign at the same time and move things around and create a visual experience and a commerce or a customer journey in a shared environment. That’s what Figma did.

I’m going to kind of do some economics history here. Figma is a 10 year old company or a little over 10, 2011 start, but it did not have its first dollar in revenue until 2017. So when you’re tracking SaaS companies, the typical approach for doing that is how fast do you grow from your first dollar in revenue? So in 2017, first dollar, in 2021, 200 million and in 2022 it’s projecting to hit 400 million in revenue. As a little bit of a backdrop on what that means, the average SaaS, cloud scale SaaS company takes, it takes four and a half years I believe it is, to reach 10 million from their first dollar of revenue. Only eight companies of the 72 cloud scale SaaS companies that are currently, that were measured in at least one study that I looked at got to a hundred million in less than five years from revenue.

Those are Slack, Square, Dropbox, Elastic, Hortonworks and Pivotal. So it’s actually six, sorry, not eight. Sorry, six, I can count but not this time. Most companies take 10 years to 15 years to hit a hundred million. And let’s just be clear here, this company hit 400 million in five years. Why am I sharing this? Well, cloud economics are interesting and a lot of times, even as these companies were hitting a hundred million dollars, they didn’t make any money. This company’s operating is net operating profitable. So they’re making money, they kept 90% margins, 150% net customer retention, one of the fastest, if not the fastest company to hit 200 million, if not 400 million. They’re going after a 16 and a half billion dollar TAM. And they’ve almost zero cost of customer acquisition. And the reason I point all this out is rare companies are still rare.

And so when you’re in Adobe and you’re looking at a market you want to go after, you have to consider, “Well what would it cost us to build this?” Well, XD was their product and it hadn’t succeeded, they were deinvesting in it. That’s the only competitive product they have and they don’t use it. They’re not selling it or promoting it. So this is an augmentation of their current revenue. It’s not going to be a competitor. The question is, there’s some antitrust, I’ll let you do that part, but in the end of the exercise, a company that’s creating that kind of margin, this kind of growth, does it warrant a 50 time forward earnings?

And a year ago, I think everyone would’ve just applauded and been like, “This is great.” But in this economy it’s really hard to get away with it. So I’ll leave it at this, in particular time in this market, it’s very hard to say that was a good decision for Adobe. It’s very hard to say it. However, in five years, if the company continues to grow at this rate, they will look back and we will realize that once again, the crazy hive mind market pundits will have gotten this wrong and said to you and the leadership at Adobe got this right. And I think that’s what we’ll end up finding. But again, in this current market, there’s no way this would be accepted or people are going to like it, hence the selloff. But I don’t necessarily think Adobe was wrong.

Patrick Moorhead: I’m not going to hit on, excuse me, I’m not going to hit on the earnings. I’m just going to hit on the deal here. It was interesting, I was thinking through this and this absolute ridiculous 50 x in this timeframe and Daniel, you kind of got me into Succession and I was wondering, “Hey, which Waystar Royco deal is this more alike? Is this like the Vaulter deal or is this the GoJo deal?” And if you remember the way that Vaulter ended up, they essentially just went in and shut down the company because they lied about all the, so it’s probably not going to end up like that, but maybe more like GoJo, where literally they came in and did that. But then again, when I look at Vaulter, it was this young company. People weren’t using the old media, which is exactly what we’re looking at here with Figma because I couldn’t help but to look at the comments about it. And you had these really passionate Figma users that were almost like, “I use Figma but I hate Adobe.” Kind of like-

Daniel Newman: Vaulter and Salesforce.

Patrick Moorhead: Or Waystar Royco or Instagram and Facebook. So that’s kind of what it brought up with me. But if I get real here, like a real analyst like I should be, top two concerns number one is antitrust. The current environment in the Department of Justice and the FTC with the crew that the Biden administration brought in, not only do they dislike huge companies buying smaller competitors, the agencies are actually actively investigating companies like Facebook for buying Instagram, what over five years ago? So-

Daniel Newman: Like 15, I think.

Patrick Moorhead: That’s not accurate, but I’ll go along with it.

Daniel Newman: How long ago? Okay, go ahead.

Patrick Moorhead: It definitely happened after I started my company, but heck, Facebook didn’t even go IPO 15 years ago. It went more like I think 11 years ago.

Daniel Newman: I’m just screwing around. It was a long time. Go ahead.

Patrick Moorhead: So coming into this environment, how on earth are they going to pull this off? And I kind of riffed on one of the first articles that came out from Tech Crunch. Literally the headline was Adobe is buying Figma for $20 billion, taking out one of its biggest rivals in digital design. So issue number one and issue number two, I think you talked about it, which was 50 x four ARR, $20 billion, one of the highest SaaS valuations I’ve seen at least since the market’s been down. And is this the best use of Adobe’s capital? I don’t know. We’ll see.

Daniel Newman: Well, so 2012, so we were right in the middle, you were five, I was 15, it was actually 10. So there we go. That’s the exact number there. What do you say? It’s the over under. And Pat, I totally agree with you by the way, highest valuation, I think second highest ever SaaS deal, highest ever private deal. And definitely a multiple that’s got, I can’t wait to hear what David Sachs has to say about this one. That’s going to be a good one on the other bestie pod. The second bestie pod out there.

Patrick Moorhead: I know, I know. And I’m interested to hear what Logan Roy might say about the deal but it starts with an F and it ends with an O.

Daniel Newman: F off, it starts with the F.

Patrick Moorhead: Yeah, exactly. He’s going to build it himself. So F off. Hey, let’s move to just a completely different plane of existence here and this is getting into the software tools that you build IP with and that you build and design SOCs with. And Cadence came out with a super exciting, two announcements. First one is called Verisium, and this is an AI verification app suite that sits on top of what’s called JedAI, which is the joint enterprise data and AI platform. So essentially bringing AI to verification. You might be like, “Pat, this is so nerdy, I don’t get it, I don’t understand it.”

Well, here’s the deal. So in the last five years, the cost of an SOC going from 10 nanometer to 5 nanometer has tripled. And inside of that you have five costs. So you have IP qualification, right? That’s like a CPU, qualifying the CPU you might get from ARM, it might be X 86. There’s money that goes into the architecture itself. There’s the physical IP. There’s the verification to make sure the physical IPs work and the architecture work. And then you have software. And we all know that software has gone up exponentially when you’re looking at SOC. So this tool essentially can reduce debug verification times up to 10 X by leveraging all of the knowledge inside of the enterprise into any type of design that you’ve ever done. So if you did a prior design, it learns from this.

And so it pulls data from five different platforms. And by the way, most of the audience has never heard of this, but jasper, excelium, palladium, protium, helium, okay, pulls it into JedAI so we can learn from the future. And then barisium is five different tools for debug, semantic differential, pin down wave minor outer triage, which again, I don’t expect you to know, but I think it gives credence in depth over just saying it’s AI, right?

Two big quotes that I thought were big coming from two of the top five largest semiconductors out there, Samsung Semiconductor who’s either number one or number two depending on the day and Media Tech, the number four silicon vendor out there in the world. So congratulations to Cadence. This is a really big deal. And quite frankly, if you could think about increasing time to market, having the best PPA, that’s performance power in area and that kind of is the deal here, this is in incredible value to these customers.

Daniel Newman: I don’t have a close relationship with Cadence, so I’m taking this as a 78% learning opportunity and a 22% because I have an opinion on everything. But the long and short is that the ability to expedite the development of AI tools is going to be critical and AI silicon is core. We talked about this with LinuxONE, we talk about this constantly on devices. We talk about this in data center applications, Pat, so these are SMLs and the companies that are the companies behind the companies that make things happen. It sounds like Cadence is really going to be an enabler for a number of the key silicon providers to be able to speed up and expedite the design of silicon for things, whether it’s ADAs, whether it’s data center, whether it’s on device, Pat, and so I’m going to dig a little bit more into this. I don’t have a ton of commentary here, but it’s impressive. And you’re doing a Forbes piece so I can plug you?

Patrick Moorhead: Yeah, I’m going to do a Forbes piece and which I’m, again, I think, I feel like my role in this is to simplify things. If you look at the grand scheme of semiconductors, the tool vendors get no credit, right? Even though you couldn’t do any of this. And it also relates to even parts in China that were developed using Cadence and Synopsis, which came under the rules where the Chinese are going to have to create their own design tools and methodologies. And when Pat Gelsinger talks about using industry standard design tools, he’s talking about companies like Cadence that enable Intel to do their own foundry because these are the tools that customers use to take into a Foundry. So good analysis here, good adders. Let’s move to something completely different and that Tesla self-driving is under fire again, what’s going on here?

Daniel Newman: Well, Pat, there’s a number of different pieces out there. I sort of focused in on one in the Verge that talked about, and like I said, while we don’t do a ton of news, I think it’s important to just give a little backdrop here. There’s another Tesla owner coming after Elon Musk this time, basically for misleading and deceptively marketing it’s autopilot and full self-driving advanced driver assistance features that are software add-ons. This is a individual plate tiff, gentleman, I believe a gentleman named Briggs Matsko. And by the way, there’s another class action going on as well.

So there’s an individual and a class action going back to about 2018, people were paying $5,000 for what was kind of a full self-driving, I believe. In the newer year now that’s gone up to $15,000, but yet it’s not working all the way. He’s basically talking about how this full self driving has been promised for many years in 2018 or ’19, Musk said Tesla was going to make a cross country trip. Here we are in 2022, this hasn’t happened yet. We of course have heard about crashes all over the country. I won’t say they are endemic or anything yet, but there’s been a number of them that have happened. And obviously there’s been some deaths. And this is part of, I guess, innovation, when you can’t do it in a virtual world. But that’s a big problem.

So the company has basically been, it’s been facing scrutiny for this for some time, Pat. We had a really interesting discussion with Luminar CEO, Austin Russell, who has also brought attention to how the computer vision only approach that is being used by Tesla isn’t meeting the needs. And I know Tesla got real upset about some of the demos. You had talked about another gentleman, I can’t remember his name, that came out just a few weeks ago. And also full head on assault on Musk between having to buy Twitter and no longer wanting it, his full self-driving under fire and now this lawsuit, plus I think California has another lawsuit, right? That’s the one that’s… Is it California, Pat? And I’m… I don’t want to a hundred percent here. Musk is kind of like a constant, he’s like a fireball in the media right now.

And overall, Pat, I guess this is where I’d come in with an assessment and say, “How do you break the standard, take risks and chances and do so in an industry like this where safety so sensitive? How do you make promises for a capability?” But obviously we know now that capability is not quite as capable as it’s been promised. We’ve got a whole world kind of with their eyes on Musk and we have a pretty significant amount of data now that says what’s being promised isn’t really what’s in the market. And at the same time, I like Elon Musk, so this is not like… I like that he’s bold, I like that he’s ambitious, I like that he stands up to things. But Pat, I think he might be in a little bit of trouble here.

Patrick Moorhead: Yeah, Daniel, I never really fully understood the veracity of the Tesla crowd until I started bringing some truth to the reality of Tesla cars. I’m not a newbie. Listen, I have run engineering groups, I’ve run businesses before. I know how these things work. I applaud Musk, I think he’s a freaking genius. He is our ages Howard Hughes. And by the way, if you know the history of Howard Hughes, he had some weird peculiarities just like Musk does. And I think there are certain brains and types of brains that enable this. But let’s get to the reality. I mean the National Highway Transportation Safety Association brought out numbers of Tesla crashes in the US where ADAS was enabled. And in July there were 51 crashes where ADAS was enabled at the time and airbags were deployed 46 times out of the 51. So, that means all these crashes were big.

And I get back to what you talked about, which was in 2016, Musk made the claim that he was going to do a coast to coast by 2018. And he said every single car built since October 16 and this was with HW2, will be updated. And Elon doubled down in May 22nd, 2017 and has said, “Hey, we’re still on for the end of the year. Just software limited. Any Tesla car with HW2 by”, that’s the Nvidia silicon, “will be able to do this.” And here we are in 2022, over five years, over six years on the initial claim and we’re not here yet.

So I believe one of the biggest mistakes that Musk made and Tesla made was getting rid of radar and never adopting lidar. And at the time lidar was this massive puck. And you can look at some of the Google cars that are going around as robotaxis, they’re huge. That’s how you know there’s lidar. But companies out there have absolutely brought down the cost to below a thousand dollars like Luminar. You have other companies doing silicon like on semiconductor for the automotive. You have Intel bringing down the cost of lidar. And I think that Musk really needs to reassess. I don’t believe that he has the on chip hardware enough oomph to make it happen. Do I have the exact figures on that? No. Does Musk keep improving the software? Yes. Are there still big issues with it for FSD? Absolutely. I think that companies should have a cooling off period and reassess otherwise the government’s going to step in and the government already is with this.

Daniel Newman: And we know our current administration, just a quick booma, booma, booma. Probably doesn’t like Tesla that much to begin with. I mean after having 99% of the EV market in the U.S. didn’t even get an acknowledgement when we were talking about electrification. So a little chip on Musk’s shoulder, I mean I think you said it really well. He’s got an opportunity to adopt second and third layer technologies that would provide a much more precise full self-driving capability and for whatever reason, just absolutely steadfast that it can all be done with computer vision. And I don’t understand what that is other than stubbornness, you know you mentioned the Howard Hughes thing, Pat. But anyhow, I just think it could be better. It’s not in… There are lives at risk and that’s not something to take lightly. Even one life when you don’t have to. Sometimes I feel like we’re really callous and apathetic about things like that.

Patrick Moorhead: And just to circle back, I mean ADAS is safer than a human, I realize that. But that’s ADAS, when you do FSD, it’s a completely different ballgame.

So let’s move on here, final topic ARM did a infrastructure update where they talked about an updated Neoverse V2 called Demeter or Demeter, I’m not too sure. And some updates on their V NE platform. So some of my biggest takeaways is for the first time ARM is saying they will have leadership with single socket one thread integer cloud performance. And that’s over Intel and that’s over AMD. Now I don’t think that they know what Ampere Computing is coming up with, but we’ll see. The other thing was that Grace, NVIDIA’s Grace super chip is in manufacturing and the claim there was two times the performance per watt over XE 6 and I think that was a big deal.

Also, there was a disclosure, an AWS disclosure, that 48 out the top 50 AWS EC two customers are running Graviton2. And by the way, that doesn’t mean they’ve moved off of X 86 and moved solely onto Graviton2, but it says that they are running production workloads. I thought that was impressive. And Dave Brown, we had him twice in the Six Five summit to talk about his homegrown silicon. I thought that is great.

And just doubling back, Daniel, I mean the company has just come such a long way and ARM being a British company is they don’t flex a lot. But two days ago in the infrastructure announcement man, they flexed. They’re like, “Hey, we are first with memory bandwidth, the highest level of memory bandwidth. We were first over a hundred cores. We were first with DDR5 and PCI Gen 5. We were the first one with over 502 [inaudible] rate 2017 with the Alibaba Yitam 710.” I mean they were on a fricking roll. And then they’re like, “Hey and don’t forget when it comes to DPUs and network and security offload, whether it’s AWS Nitro, Intel Mount Evans, Marvel Oction, Intel BlueField. All based on ARM IP.” So serious, serious flex going on into the future, but also showing what they did in the past.

And I have to tell you, in my mind, this is kind of this pivotal moment we’re going to have to see the companies that come out with near verse V2 to see if everything lives up to it. But I do know that ARM is a very conservative company in the claims they make and they have to be extra conservative because they don’t make chips. They make IP that chip companies pull in to make their chips better. And I’ll end with all this. I’m interested to see what Ampere has with AmpereOne that’s that’s been sampling since May, that has a custom CPU architecture based on an ARM ISA and some other custom IPs.

Daniel Newman: More companies using ARM to build silicon. So hey, you said a lot, you did well. I guess I’ll lean into here is the ARM ecosystems growing. The ARM adoption is becoming more crystallized. The X 86 ARM debate will rage on for a lot of different particular both in the data center and in the PC space. But ARM is here to stay. I think any debate about that is over at this point. The focus on power performance will always cause the geeks to want to geek out and do benchmarks and tests. The company has already gained a foothold. You mentioned all the different architectures, Pat, that they are based on, ARM already. We talked about effectively every hyperscale cloud provider that’s serious right now has at least one offering that’s built on ARM or an ARM variant from an Ampere or something like that. And so ARM is here to stay.

I think we’re going to rage on to the debates about the very large specific workloads that have been optimized and optimized and optimized on X 86 and can ARM bring the developer crowd along and bring that kind of, the SAP workload, is that going to ever be as run as well in a data center on ARM as it will on X 86? And I think the answer is that’s where it’s heading. And so Pat, I think ARM’s being more aggressive. They’re a public company now. I think that’s going to put pressure on them to be more exciting with their roadmap and try to build an investor sentiment and strength.

But I also think we’ve all watched Intel and we’ve watched AMD and we’ve watched Nvidia and we know operational execution is everything in this space. So can they get this out? Can they get it out on time? And of course, like you said, just getting it out, sorry, you’re right, Pat soon to be public company, can they get it out? Can they get it on time? They’re going to be held to a much higher standard over those the next several quarters. And then of course who builds on it and how quickly are they able to get what they’re building on Arm into market? But they were exciting announcements. They definitely look like they’re going hard at it in that data center space and competition’s good. I don’t know, you and I say that a lot, I’m not going to deviate.

Patrick Moorhead: Good stuff, Daniel. No, I think it’s the time. There was a lot of criticism around SoftBank’s acquisition of Arm and SoftBank in general. And SoftBank has laid a bunch of eggs, but ARM was clearly a success, gave ARM time to come into the pits and make some serious investment. And some of the biggest IP areas they invested in were infrastructure and automotive. And those businesses are taking off. IOT didn’t, they’re selling off and laying off a lot of people in IOT, particularly on the software side and that bet didn’t work. But quite frankly, IOT didn’t scale quickly as any of us thought that it would.

And I’m looking forward to them going public to see even more granular data. But hey, Daniel, we’ve been going 48 minutes here and I think we are going to wrap it up. Good show talking Oracle, IBM, Adobe, Figma, Cadence, Tesla and ARM. Bestie, it’s great to see you. And it snowed up at the very top of the peaks here of the mountains and I’m going to stay the hell away from that. But I might do some long hiking or maybe I’ll go into Aspen and drink some champagne. I’m not too sure which of those I’ll do.

Daniel Newman: Well, it’s Friday, You ain’t got no job today. And that’s an old movie reference. People probably wouldn’t get it. And we’re really old and it sounds tacky, so I’m going to let it go there, buddy. But enjoy your trip, enjoy your vacation. I’ll see you next week. We will be venting our tails off, Intel On, Intel has a big event next week.

Patrick Moorhead: Gosh, I hope it’s not that, I hope it’s not that week because-

Daniel Newman: Wait, no, that’s the next week. Next week I’ve got Dreamforce and you and I will see each other in New York for Qualcomm Automotive Investor Day, right?

Patrick Moorhead: That’s right. Yep. I’m going to be paying attention to Dreamforce, particularly on some of their growth areas, doing a workshop with Cloudera and spending two days in New York City with you at the Qualcomm Investor Automotive Day. So looking forward to that buddy.

So, that’s a wrap. Thanks everybody for tuning in. Hit that subscribe button if you like what you heard. If you didn’t like what you heard, hit that subscribe button just so you can hunt Daniel and I down on social media and tell us where we’re wrong. Now we appreciate the comments. We’re not right all the time, but we’re right on the important stuff. That’s what I like to say. And we’ve made some pretty good calls, but if you have any comment, commentary, hit us up on Twitter. Hey, we appreciate you. Thanks you for tuning in and have a great weekend.

About the Author

Daniel Newman is the Principal Analyst of Futurum Research and the CEO of Broadsuite Media Group. Living his life at the intersection of people and technology, Daniel works with the world’s largest technology brands exploring Digital Transformation and how it is influencing the enterprise. Read Full Bio