We are Live! Talking Chip Sanctions, Qualcomm, Micron, IBM, Oracle, NVIDIA, and Snowflake

By Patrick Moorhead - July 5, 2023

On this episode of The Six Five Webcast, hosts Patrick Moorhead and Daniel Newman discuss the tech news stories that made headlines this week. The handpicked topics for this week are:

  1. US-China Chip Controls, NVIDIA
  2. Qualcomm Snapdragon 4 Gen2 for Entry Phones
  3. Micron Earnings
  4. IBM Buys Apptio
  5. Oracle Optimizes DB for Arm, Ampere; Larry Questions Intel?
  6. NVIDIA Snowflake

For a deeper dive 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 ask that you do not treat us as such.


Pat Moorhead: Hey, this is Pat Moorhead, and we are live again here for our Friday Six Five podcast. It’s my favorite thing to do for the week, grinding it out Monday, Tuesday, Wednesday, Thursday, hit Friday morning and man, I am in. It’s like nirvana. Gosh, is it really Friday here, Dan? Did we almost make it?

Daniel Newman:It has been a week. I can tell you, Pat, this was a busy one. As you know, we had some strategic conversations this week, a big meeting around the Six Five, because of The Futurum Group

Pat Moorhead: Well, you had a big Futurum Group offsite meet-

Daniel Newman: I had a Futurum Group offsite.

Pat Moorhead: With everybody in like a real company.

Daniel Newman: We had 20 or so of our executive and leadership team come to town, and that was two and a half days. I did a golf outing, our first annual customer appreciation event here in Austin. It was 1,000 degrees at Top Golf, but we had a tremendous turnout from our customers. It was really invigorating to see that and do that, first annual, and we’re six years old, first time kind of hosting an event. So all that queso was expensive though, Pat. I think that’s why his analysts were used to letting the vendors pay.

Pat Moorhead: No, I know. Just especially getting Connor in there. He probably eats like-

Daniel Newman: It was great to have Connor there. He was their freshman-

Pat Moorhead: He probably ate like a gallon of it himself, but I did see he was in the picture looking good on that swing.

Daniel Newman: He did look pretty good on that swing. I think I hit it further than him though.

Pat Moorhead: Did you? I looked like that. I looked that good when I was like 20.

Daniel Newman: Yeah.

Pat Moorhead: Anyways, hey, if you’re new to the Six Five, I’m going to ask you what’s wrong with you, but hey, we cover six topics, five, 10 minutes each, hit a little news, context of analysis worthy. We’re one of the only programs on the entire internet that does tech analysis, and I have a lot of fun doing it.

Daniel Newman: Hey, does good tech analysis. That’s even a smaller subset.

Pat Moorhead: There’s a lot of news programs out there.

Daniel Newman: A lot of news and-

Pat Moorhead: But not a whole lot of high-quality analysis.

Daniel Newman: A good bit of-

Pat Moorhead: We like to think we do this-

Daniel Newman: A good bit of crappy analysis, you can find it, but it’s just, it’s crappy. It’s like the Ford Fiesta type of analysis. This is the Ferrari of analysis.

Pat Moorhead: Exactly. Exactly.

Daniel Newman: Oh, you like that?

Pat Moorhead: Yeah. A little disclaimer, we do talk about public companies. Don’t take anything we say as investment advice. Do your own research, get a professional, we’re not. Anyways, we have a good show for you today. We’re talking China chip sanctions, impacts on companies like Nvidia, Micron, AMD. We’re hitting the Qualcomm Snapdragon 4 gen 2 for entry phones. We’re hitting Micron earnings, little bit 7:30 AM IBM buys Apptio on Monday morning. We were hitting that pretty hard. Oracle made some big announcements about optimizing its database for Arm and Ampere. Talked about AMD. Larry questioned Intel. So we’re going to dive into that a little bit.

Daniel Newman: Woo hoo.

Pat Moorhead: Yeah, I know. Can you believe it? And we’re going to follow up with Nvidia Snowflake Alliance. Imagine that, another company aligning themselves with Nvidia. Can you imagine being their… Anyways, we’ll we’ll get into that. But hey, let’s hit this first topic, US China chip controls. Dan, you and I were both burning up the video news services. I was on CNBC twice talking about this. I think you were on multiple ones. TD Ameritrade-

Daniel Newman: CNBC, TD Ameritrade. Yeah, yeah.

Pat Moorhead: I did CNBC within an hour of each other.

Daniel Newman: Yeah, you were on the one show and I was on the next show.

Pat Moorhead: How about that?

Daniel Newman: One of these times we got to get them to put us on together. It really does feel like the Six Five join CNBC, John Fort make it happen. You know what I mean?

Pat Moorhead: On John. You can do this.

Daniel Newman: Come on John.

Pat Moorhead: You can do this, dude. No, but hey, let me do the rundown here. So there was an article written, I think by the Wall Street Journal that said the Biden administration was going to provide higher levels of scrutiny. They’re already putting higher levels of scrutiny on companies that do AI training and inference. And so people, as you would expect, brought up Nvidia and talked through that. So what we’re looking at, and I just want to put this in context here. This is really a 20 year conflict, okay? 20 years ago, China pretty much banned IBM and Cisco and even Microsoft from critical infrastructure. And they still ban Google, Twitter, Facebook, Instagram, and I’m pretty sure WhatsApp. So it’s really an asymmetrical relationship where you had China banning all these companies and then US banned Huawei and ZTE for military reasons.

And then the Biden administration this year banned 59 companies who sell to the US but also are a key supplier to the Chinese military. So an interesting conflict going on, but nobody should be confused that it’s new, it’s not it. It’s actually, like I said, it’s 20 years old if you want to really put it in context. So I think we’re kind of in a standoff and what that is doing is that’s it impact Nvidia a little bit, but not a lot. NVIDIA’s CFO came on and basically said, “Hey, not really a big hit here short term.” Should make a really good point, which I do agree with is I’m hoping we are balancing the military element or the public safety, US security. But the fact that China is going to, if we ban certain semiconductors from China, they’re going to go and try to develop their own.

Now China has been trying to do this for a long time. In fact, I licensed China when I was at AMD with x86 technology. Now it was 46 technology, but still you could take that and duce, they did nothing with it. And that’s because the US companies like AMD and Intel and Arm were just moving way too quickly.

So here we are in this standoff, but China will continue to invest. I can see them figuring out the designs. The one thing that plays into this is can China get leading edge? They can’t get leading edge right now from the highest performance chip making equipment company. So China might be able to design it, but can they actually make it? Can China do generative AI? Yeah, they can. But probably at a 10x less efficient way of doing it, they have to probably throw 10x the hardware at it to be able to do what US chip makers can do. And then there’s the software. But I think that China could create its own software ecosystem if it wanted to. If nothing else. It has millions more developers than in the western worlds so I left you a lot of oxygen in the room, Dan. Go for it.

Daniel Newman: I like that you self proclaim whether there’s oxygen in the room. I think you did a pretty good job of covering a lot of it. But fortunately the fun part here is this is very subjective, it’s very qualitative. This is not necessarily a spectrum in light and weight. There’s a lot of ways we can all sort of chew on this. And I’m kidding buddy. You’re always the host with the most, I would argue that you’re the world’s best moderator.

Pat Moorhead: Well, no, I think that’s already taken by Jason.

Daniel Newman: I think he trademarked it. But I think I’m proclaiming you are the best. He proclaims it about himself. So just think about, remember that Jason when you’re listening to the Six Five pod. In all serious though, we did have quite a media circus. Pat, this stuff always fuels a lot of interest. This is what the world wants to hear about. It wants conflict. And there is conflict here. And this is not driven by AI, but it’s been reinvigorated. It’s always triggered by something. It was triggered by supply chain, then it was triggered by the CHIPS Act and the desire to start to build a more resilient supply chain around the world. And then in this case, with the power that we know AI is going to have over industry. So there’s two basic arms races. There’s the legitimate military industrial complex of arms race to be the most powerful and have the most capable national defense ecosystem in the world, which the US and China both seek to have.

And then the second one is there is the power of money and industry and the United States and China are boxing in any given day for the number one, number two economies in the world. China has shown over the past decades that it can quickly close the gap on the US’s global economic standing. But the US has long enjoyed some technological advantages that come from largely the Silicon Valley of what we’ve created, what we’ve built. And yes, we have created some risk through outsourcing a lot of the manufacturing to Taiwan, which is a territory of debate, let’s just put it that way. The way we in the west see Taiwan and the way the east sees Taiwan is not the same. And that’s the thing about the myopia that people have in this world is all of us here in the US and even some of the chip makers and Pat, you and I have talked to a lot of executives at times, it’s almost with hubris that they kind of think that everything’s kosher and Taiwan is going to be fine.

And then there’s this irony in the east where it’s kind of almost like they’re just sort of waiting, lurking, watching and you can’t help but just be nervous that the decision could come at any time to reclaim what the East believes is it’s own. And of course almost all the leading edge is produced there. So that’s creating a lot of friction, a lot of tension that the US and both China are fighting for. So you look at the Micron situation and what an interesting decision for China to go after Micron. And here’s my thesis and Pat, they’ve done nothing to actually validate the claim. They just kind of block made a decision. And I feel like they almost picked a company that was in an area that China has the most capability to do on its own. Like “Yeah, we’ll go after the memory folks. Because we can do that.” And I would argue Micron has some really great innovation that China and its companies like SK that they have relationships with that they continue to get volume from yield can make a lot-

Pat Moorhead: And they have their own and-

Daniel Newman: And they have their own. And they have their own. Yeah. But having said that, going after an Nvidia is too risky because China can’t do anything right now to compete with the US and an AI without having access to Nvidia chips. Now they don’t have access to the 800 or the most advanced chips, but they still want to be able to get the legacy stuff because the legacy stuff that they are still able to get beyond is still better than anything they can manufacture there today.

Pat Moorhead: That is true.

Daniel Newman: So they do not want to pick a fight where they cannot compete. So they randomly are sort of saying “Here, let’s do this micro-aggression and go after a company that’s kind of more nascent to us where we can get what we need from other channels more easily.” And of course, don’t be dumb society to think that China doesn’t have ways to work around chip controls. They’re still finding ways to get the most advanced technology into the country. It’s happened for generations. This is not the first time. I’m not proclaiming Pat, you look like you want to say something about that.

Pat Moorhead: No, I’m just agreeing. I’m agreeing with you and I’m glad you brought that up. People have been backdooring band stuff into China forever. Now when they get caught it’s super ugly. We saw what happened when Huawei was accused of funneling their technology to Iran. They literally jailed the head of Huawei, the daughter of one of the rotating CEOs for two years.

Daniel Newman: Yeah, this isn’t something to take lightly, I’m just pointing out it’s a very complex subjective qualitative debate that needs to be had. These are additional controls. And let’s just be candid, the most powerful country in the world for the future of the semiconductor industry is the Netherlands. So you can go smoke your weed and drink your coffee and you can wait for the Dutch to basically make the final decision on who the most-

Pat Moorhead: Hold on man. Tell us all why that’s the case. I know why it’s the case, but tell everybody.

Daniel Newman: Because ASML holds the keys. The only manufacturer that can enable the EUV to be able to go to these smaller and smaller process nodes is ASML. We’ve actually weaponized the Dutch and they’re the most powerful country in the world. And right now they’re on our side. But they-

Pat Moorhead: I think China will just-

Daniel Newman: They’re more powerful than Joe Manchin. They’re literally more powerful right now than Joe Manchin. They can literally decide the winners and the losers. I don’t think anytime soon they would be defecting to support the Chinese. And as long as the demand stays this high in the US and in Taiwan, everyone from Nvidia to Micron will be fine because the demand locally will augment and offset and allow for growth even if they’re losing 20 or 25%. I could talk about this for another hour, but we’ve got to do five more topics and we just did the six 20. We’ve literally been on this topic forever.

Pat Moorhead: Well, you and I both spent a lot of time this week going through this and there’s a lot to talk about and it’s a complex issue. It’s not a product launch.

Daniel Newman: That’s fine, and I could do a whole show on this. We could literally do the-

Pat Moorhead: Why don’t we do that Connor?

Daniel Newman: We should sometime-

Pat Moorhead: Let’s.

Daniel Newman: Make it happen. Let’s talk to our producer. We should do the Six Five deep dive, Six Five deep dives. I love it. Six Five, Deep Dives.

Pat Moorhead: About that… Somehow. I think I saw that on a slideshow. But hey, let’s move on. Let’s to the next topic. Talking of product announcements. Qualcomm Snapdragon 4 Gen 2. Dan, what’s going on here?

Daniel Newman: Well first of all, someone wrote a great article this week in Forbes talking about the battle for AI needs to also be fought and won at the edge. There was a great Forbes article, I can’t remember who wrote it, but I feel like I know it. It’s

Pat Moorhead: So bad.

Daniel Newman: But what’s right Pat is, I had a interview there. It was the CEO of Silicon Labs, Matt Johnson, and we were talking because they’re very IOT centric. And remember when IOT was cool, we used to talk about it. Now it’s not really all that cool, but AI is going to be hyper dependent, not just on these kind of training workloads for enterprise and LLMs and the data center, but think about all the data that resides at the edge. The data set that sits in cameras and vision and phones. And again, edge has by the way a continuum of definitions because it can be anywhere from the sensors in the fields of a farm. But an edge can also be, in some definitions, can be something like a PC or a workstation or a phone. So Qualcomm’s made quite a bit of headway in terms of helping the world understand its critical nature to this part, the edge, the device.

And one of the things that the company does, while it’s very well known for its kind of ultra premium tier devices, is Qualcomm has a pretty strong ambition to be democratizing more than 40% of the world that really still isn’t fully connected all the time. And part of that ability to do that comes down to, yes the critical infrastructure, but also creating devices that can be more affordable. While we all like to kind of drool over the new Apple devices, those devices cost months of salary, months of pay for people in lower economic standings and in other parts of the world. And of course, even here in the US and in Europe, across Asia, there is a subset of the population that do not want to spend $2,000 or getting close to 2,000 US dollars to buy a smartphone. And so Qualcomm, I think it was about a year ago, launched the Snapdragon 4 Gen 1 and the idea was is to help the manufacturers, the OPPO, the Xiaomis, the Samsungs to be able to create more compelling lower-end entry devices that still offer most of the features.

And this is what I think is really good, is the company’s continuing to make it lower power, giving it better battery life, higher performance, improved audio, an improvement in the 5G RF modem, which by the way is a big part of making 5G beneficial. You and I have another conversation to have at some point about why 5G is stalled. We really do, it hasn’t actually stalled, but the interest, the excitement, the enthusiasm has kind of hit a wall. It came with some updates to the ISP. We all know that people, whether they’re buying low end cameras or phones… Cameras, I just call them cameras because for a lot of people why they have their phone is, it’s an all day long camera. It’s the ability to document your journey and share it on social media. You need connectivity, you need a camera.

And then of course what was kind of cool about this Pat was it’s I believe the first in the series to be built on, it’s the TSMCs 4nm. So we always talk about 5, 7, 5 and 3. And they’re actually building this on four. The other thing that they did that I thought was pretty good was they increased the frame rate. So the display is going to be really good for high definition visual and then quick charge.

And Pat, I think you and I can both agree nothing more frustrating than sitting around waiting, waiting, waiting for a charge on your device. As devices age, that tends to get worse. So the quicker we can get them charged back up for use, the better. So it’s a fairly comprehensive offering. I think that the real key here is Qualcomm’s definitively won the top of the market, but they have some opportunity for strong growth in the middle and lower tier. This is a more and more compelling device. It’s really go into that media tech space, go into those lower end devices that are being pumped out by some of legacy Samsung and they have a chance to gain market share there and grow. So it’s an encouraging step forward. As the company continues to try to show the power of AI at the edge, Qualcomm continues to make that more democratized for phones across the price spectrum.

Pat Moorhead: Yeah Dan, that was a good breakdown and this is all about optimization and cost reduction. And if you see what the company did before between Gen 1 and Gen 2, they removed the hexagon DSP. They also removed a millimeter wave 5g that is pretty expensive and they moved from TSMC to Samsung. So I like this move, but like you said, the profit dollars are being driven primarily by the high end and the mid-range, but you have to drive the volume and scale with a very competitive lower end. And that’s exactly what the company’s doing. The company can derive higher prices because you do get the Snapdragon brand particularly in areas like China and Southeast Asia, the Snapdragon brand really is premium. It reminds me a lot of how Intel Inside used to be 20 years ago here in the United States for PC.

So hats off to the marketing folks. The brand folks have done a really good job, particularly outside of the US with Snapdragon. So good breakdown of that. Let’s move to another topic. Micron earnings, and I talked about this again twice on CNBC and I would say all in all it was a mixed bag. There was this excitement that the company essentially said in the memory market, the demand trough is over. They said they had increased confidence that the industry passed the bottom for both quarterly revenue and year on year revenue growth that came directly out of the call. They beat on EPS, they met on revenue. Initially after hours the stock went up. And then I think as people kind of peeled it back, people looked in and said, “Okay, so revenue is still off versus last year by over 50%.”

So let me go down the list of the business units. Compute and networking were down 64%, embedded off 36%, mobile off 58% and storage off 53%. There were some really good commentary on AI and the company did a really good job showing how much the order of magnitude memory you need. And oh by the way, higher performance memory. And then if you look at HBM three, which is driving a lot of these accelerators out there, that’s another big adder to AI. China, they gave an update on that pretty much sticking to their talking points that they brought out that we covered it on the prior podcast saying that the worst part impact is a low double digit percentage of worldwide revenue. I’m going to take that as between 10 and 15%, not necessarily hitting in the quarter that was released, but in previous quarters.

So that would affect the outlook on one side in data center, generative AI, they talked about as big, but again, everything else in the server market is blah. In fact, they on the call, their CEO used the word lackluster. So again, kind of mixed. PCs were what you would expect, smartphone, what you would expect. And my final thing I’m going to talk about is risk profile. Investors hate risk. And the story of Micron kind of goes like this. They brought industry leading technologies, DDR5, HBM3 things like CXL, but demand fell off for end product demand. The smartphones, right? The servers, the PCs, and then all of those manufacturers, the Dells, the HPEs, the hyperscalers, they reduced their finished goods and in some instances even returned memory. Okay, this is true for not only Micron, but it’s true for Hynix, it was true for Samsung.

And what happened is that kicked off a price war between folks, right? That’s just now settling down on the memory side. On the call, the company said memory price degradation went down 10%. Think about that. Dan prices going down 10 to 15, 20% a quarter sequentially. It’s tough. And now we have this China risk that I really do appreciate them being very crystal clear on what the worst case scenario is. But listen, I’m a believer in Micron. I think Micron is going to do well. I think the R&D I’m hopeful that it doesn’t cut down to the bone and cut its ability to not only create great technologies and great products, but also you have to market and sell this.

So I’ve seen a lot of companies make a big mistake by cutting all communications off, cutting all marketing spend. And I’m hopeful that the company won’t do that because it will backfire. It’ll hurt them in a year. But I’m also realistic enough to say, “Well the company has negative gross margins right now, Dan.” So you have to make the right cuts. Just make the right cuts, not the wrong cuts. It’s easy as a pundit. But don’t forget, I spent 12 years as an executive at a chip company. I do know how this works.

Daniel Newman: And of course the problem is if you cut too deep, this is a boom and bust industry. So you need to cut deep enough that you try to leave yourself in a good position for the austerity portion of any glut and any downturn. But then again, when it turns you know how hard it is to bring in talent. And then so you have to keep your best talent. You can never cut the PhDs, engineers, scientists, the people that are developing. So then it’s where it tends to really bleed out is the marketing. And then what ends up happening is you make those cuts in the sales marketing and some of the operations areas and then when the market turns you’re chasing your tail. So the companies that can squirrel their nuts during the uptime and know that you’re kind of averaging out over time tend to do a little bit better.

Pat, basically their EBITDA, or let’s just call it NOI was like 50% of their revenue in losses. So they had 3.8 billion in revenue or whatever that was. And they had 1.9 billion in losses. It cost them 6 billion to do 4 billion. I mean that’s a pretty remarkable result. And not to the good, just being really candid. Now, I’m not saying Micron doesn’t have a great opportunity around ai. I actually believe they do. I believe they have some great innovations, 232 layer man. Some of the things that they’re doing very, very proactive for data center and are going to make them a very compelling partner to many of these new future AI data centers that will all require more memory capabilities. But the business performance pa, look, I get it that you did better than you predicted, but the prediction was bad and the result was bad.

Given the current market, losing $2 billion in a quarter is hard to swallow in any circumstance. But like I said, if the company was communicating well, what they really have done is they had some great quarters during that boom run up and they made a lot of money. And this is the burn to try to keep the company in a good position to be able to capitalize on that next wave. And that’s kind of the subplot of what I heard. It’s kind of they believe the trough worst of the trough is here and now we’re going to start seeing the company rise out of that trough. I mean clearly memory is not benefiting or accelerating at a rate anywhere close to the GPU space. Their guidance wasn’t even up. It was really just in line. The result was 1 hundred million above expectations.

But when you’re in multiple billions, that doesn’t necessarily look at look to people like a holy crap moment. It’s like, “Okay, it’s good. It was good, it was better than we thought. Good.” And so the real question you have to ask about Micron is do they have the technology and the innovation? I believe so. Is China going to be a huge negating factor for the company? If these sanctions are held up, it’s going to make a difference. But is if these US chip makers continue to see the acceleration, they will likely be able to offset that same as Nvidia. By the way, Nvidia is offsetting it because the demand is so small here in the us, they can deal with it. So I think Sanjay Mehrotra came out and said the right things. I think he’s in the right place. I think the losses are just, they are what they are and it’s hard to look at them and not be like, “Ugh.”

But the company, if it can turn the corner, if it’s played its cards right, if it’s built the infrastructure out memory will scale in parallel at a fairly good clip. And then of course Pat, as demand rises in this business, prices rise, demand and price or closely tied, that means margins rise. That means profit rises and we get out of these discounting wars. And by the way, they also can get out of this by continuing to out-innovate, meaning their technology does something the others doesn’t. And so when it’s all things equal, this industry is incredibly cutthroat.

Pat Moorhead: We could see 100 point swing when there’s not enough memory and then everybody gets mad at the memory suppliers and the storage suppliers when they double their prices. But it’s like, look at this negative gross margins folks. So been covering this for 30 years and there has always been a reaction, there’s always been a memory oversupply and a memory undersupply. It’s like there’s never the right amount, it’s weird. There is no three bears scenario with memory. Hey, let’s move to the next topic. IBM bought a very interesting company called Apptio. What does Apptio do? What is it relevant? What does it mean to IBM?

Daniel Newman: Yeah, it was a big move. It kind of consumed the weekend Pat. So I think both of us were, we had had some early insights into this, so we had to keep ourselves on low about what was going on here. But overall, Pat, this was the news of the weekend. It was originally estimated at around a $5 billion deal. It came out on Monday morning, $4.6 billion deal. It’s the biggest deal that IBM has done since Red Hat. So while IBM has been gobbling up and making lots and lots of smaller acquisitions, this is the first one they bought from Vista Equity Partners. Vista bought it in 2018 at around 1.94 billion. Put in the show notes, Futurum’s practice leader and VP Steven Dickens wrote a Forbes piece about it this week. I believe you did too. Covering what the company is doing, Pat, this really is a FinOps play.

It’s a hybrid cloud play. It’s bringing together the fact that as companies continue to grow in scale and are trying to define and fully understand and appreciate the IT architecture that is going to be hybrid and multi-cloud, that they want to A, understand their cloud spend B, understand their full IT and workload placement and subsequent spend. And then C, understand the labor realization and labor allocations related to all of the above. Now that’s really what Apptio is. It’s really three products. One focuses on public cloud spend. One FinOps, one focuses on the hybrid cloud and overall IT and then the other one focuses on labor and on the public cloud spend, they have a really fast time to value product. As you obviously go more and more towards prem and towards personnel, it takes a little bit longer for that time to value to be realized.

But I think the thing here is not only are these tools important, FinOps companies understanding spend, optimizing for these periods of austerity and these market uncertainty. I know because of the magnificent seven, everybody feels like the economy is great, it’s really not. It’s great yet. AI’s given us a bit of a prop. And so companies as a whole need to figure out how to optimize spend and doing so makes a difference to every part of the business. So I like it for that reason. Pat, I also like it because I BM has become very transparent in its focal area. As we’ve talked about this endlessly. We had Arvind Krishna on the Six Five actually keynote of our last year, Six Five summit, not this year, the one before. And he talked about it, it’s hybrid cloud and AI. That’s what the company’s focused on.

It shed the weight of Kyndryl. It really zeroed in it’s turned its profitability. It was one of the best performing companies through the downturn because it’s steady, it performs and it really has gotten to a point where it’s look and feel is that of a company that has the potential to grow, albeit mid single digits perhaps is its more realistic outlook. But that’s still really good. Tools like this though Pat give them access to every customer, which has been a big part of the services consulting portfolio, the companies. How do we get to the companies that don’t using our, maybe say an IBM cloud, maybe they’re not using our cloud for financial services or they’re not using any of our public cloud services.

Red Hat was a huge win there. It get got them so much access to so many customers that are using other third party hyperscalers. This is another one where companies that are using all kinds of different clouds, multiple clouds on-prem are looking to figure out how to optimize the business. And of course, given some of the thoughts about generative AI automation lowering costs, another area to potentially lower costs is getting a better handle on your labor. This is something of consulting and services business can and should do, and what Apptio does has a software that helps you do that faster. So solid acquisition on a good trajectory.

Pat Moorhead: Yeah, we were up at 7:30 in the morning on Monday talking to IBM, senior vice president of software and CCO, Rob Thomas-

Daniel Newman: Chief Commercial. Yeah.

Pat Moorhead: … and also AJ Patel, who’s the chief operating officer at Apptio to really get insights on this, which I really appreciate. And there was only so much the two companies could talk about in future integration. But I look at this quite simply. And first of all, y’all know my favorite topic is the hybrid multi-cloud and the fabrics that are going to make that happen, whether they’re data fabrics, app fabrics, networking fabrics, security fabrics, that is the way the world is going to operate. Where enterprises want a choice between where they put their compute, where they put their storage, it’s going to be on the edge, it’s going to be in the public cloud, it’s going to be in SaaS, it’s going to be in COLO, it’s going to be everywhere.

And what Apptio does is lets you better manage the finances, the spend at where you are putting all of these workloads and just, first of all, I love it because this is a FinOps hybrid multi-cloud fabric. And regardless of where companies are putting their compute, IBM can monetize that and create value. And this adds to other types of services that operate like this. Instana, AIOps, Red Hat, Redshift is a hybrid multi-cloud fabric for applications. So I like it and I should have seen this coming so kind of silently, IBM has put together, not silently. When I put Red Hat, that’s huge. But when I look at what they’re putting together here at IBM, they have their play that says, “Hey, if you want a full stack from us, you want to buy everything from us, we have that. You want to go hybrid, multi-cloud, we have that too. And we can monetize every part of it.”

So I like this, read my article I wrote on Forbes for the deep dive. I have some quotes from Rob and AJ putting this into even better perspective. Hey, let’s move to the next topic here. Oracle made a couple proclamations. First of all, Oracle they were optimizing their database for Ampere and their database for 35 years has been x86, and now it’s going to be Ampere and Arm. And there were also some comments that Larry made about AMD and Intel. Again for years and years. You have to pick your binary compatibility point when you write software and Oracle database it’s been around, I started in Tech 32 years ago, and guess what? I got trained on Oracle when I was at NCR in 1990.

It’s been around for a long time. And Oracle’s not only x86, Oracle had their own Spark processor that they bought from Sun, but this is the first time we’ve seen it optimized for arm. I haven’t seen any benchmarks of how it performs or how it scales, but I see this as this is a huge statement, right? Because if you look at, I think anybody who’s in the know views Arm as a player in public cloud, native cloud software from the os, the middleware to the Engine Xs to the full stack. I mean we’ve seen this with AWS, right? But there’s always a continued question about what I would call legacy software. The SAPs, the Red Hats, the Oracle database. Well here we are. So it really opens us up. Now Oracle is an investor in Ampere, which makes it even more interesting.

But Larry, at this event that I was invited to but I couldn’t attend, made a statement that says “We’re going to be buying billions of dollars from Nvidia, AMD and Ampere.” And when asked about Intel, Larry’s comment based on a news story was basically talking about Intel x86 legacy. So not great, but I’m certain that Oracle is buying a tremendous amount of technology from Intel. And you don’t normally have people kind of go on the offensive, but it’s Larry and this is what we get. So good news for Oracle. Good news for Ampere and it looks like good news for AMD.

Daniel Newman: Yeah, Pat, I think you covered it pretty well. The elephant in the room is Intel in this one and it’s that this is going to have a sizable impact to the Intel number. And this is not probably something that Intel’s going to be particularly happy to see. I’m sure they’ve known about this. This is not a secret to them. You can generally sense when a buying shift has changed. This changed around X9M. This was not an immediate change, but you started to see this shift. It’s just that’s the elephant in the room. It is a signal of strength though for what AMD is doing. It’s another kind of feather in the cap. AMD has had some big wins as of recent, and of course getting this kind of alignment with Oracle could also be good because we know AMD’s intent in GPU is also significant.

So being able to get in and win on the CPU side could long term mean a significant opportunity. And we’ve been hearing about the Ampere partnerships for some time as well. Pat, this wasn’t really breaking news either, but this is a big win for Ampere and Arm of course, and shows again, the diversity of the silken that’s going to be going into these hyper hyperscalers Pat. They’re almost becoming in some ways the channel for these semiconductor manufacturers. They become the channel because yes, obviously hybrid and on-prem is still very, very significant. But so many companies are spinning up workloads in hyperscale cloud. And so it’s becoming a bit of, you go into the grocery store, you pick your soft drinks or your beer, you get Coke, you get Pepsi, you get Dr. Pepper. Now if you have a preference on the actual silicon, you can go to any hyperscaler and get the various of the majors, the AMDs, the Intels, Nvidias, Arm base, et cetera.

But if you have no preference, it might start to come down to things like price performance. We talked a little bit about AI with some of the execs at Google and obviously they have all the Nvidia variants, they have TPUs, some people really care, some people don’t. So it looks to me like Oracle is shifting and pivoting and this is going to be an area of contention, but this is a battle right now. And so it’s going to be driven by everything from the performance to the business agreements that these companies put in front. In the end though, the question will always be as with the customers and the users of this stuff, as it becomes more democratized, is it commodity or is there true differentiation on the silicon and what drives that decision of which silicon to use for which workload? So I’ll stop there.

Pat Moorhead: Yeah, really provocative stuff. I’m really interested to see what this can do for ARM and the enterprise because ARM and the on-prem enterprise hasn’t really done much at all. HPE has some interesting technology with Ampere in it, but again, more research to do on this. I have a VP and principal analyst of compute, Matt working on this right up right now. I think it’s in my inbox. Okay, let’s move to the last topic. Nvidia ties up with Snowflake and pretty much everyone else. Is this just a big nothing burger or is this, you know what, what’s going on here Dan?

Daniel Newman: Yeah, you know what, it’s a little bit of both. I think Nvidia continues to hug and wrap it’s arms around every major software player in the industry. This is just the next one, Pat. I have been from event, it’s not just software, software, it’s hardware, they’re partnering up with every major SI. Look, they’re staking their claim, good for them. But the tie-up with Snowflake, it’s really all about where compute lies and historically speaking data goes to compute. The long and short story about this partnership is it’s about the compute going to the data. And so as the volume of data continues to grow, if you look at Snowflake’s recent acquisitions, it’s streamlet, Neeva, the ability to build applications. And by the way, Snowflake’s doing a little more OLTP, a little more operationally focused as opposed to just analytic focus. You’re kind of seeing this whole industry shift.

We saw last week, Pat, when we were at MongoDB Local, how they’re doing a little bit more analytical, but they’re not doing analytic, but you can do a little bit more of that there. But of course everybody’s really focused on how do you optimize these applications? How do you make them flexible in the multi-cloud? How do you handle with streaming data? How do you keep that data secure in use, in motion and at rest? And then of course, how do you optimize compute? And so really what’s happening is that the Snow Park is what they call it container services, is giving it the ability to run containerized applications on top of Nvidia GPUs, all without moving any data outside of Snowflake. So basically you can run the app and have the compute and the data altogether without ever leaving the Snowflake environment. So Pat, it’s provocative, it’s thoughtful, it’s smart.

And of course it’s a way to, Nvidia gets more lock in and as it’s a very simple ratio of how many workloads, how much data Snowflake has that’s being used, and then how much Nvidia is going to sell. So Nvidia has built NeMo right into the into container platform. It’s a pretty smart approach, Pat, and I know we’re kind of limited on time here, so I’ll just say it’s not nothing Burger, it’s truly is the convergence of compute to the data. But it is like there anybody that’s not going to be locked in with Nvidia, by the time this thing is done, everyone’s going to be so busy chasing. It’s like, wow, by the time any of these others have their platform, it’s going to be just a big rapid game of catch up. Hopefully the dollars and cents and the huge margins will create a competitive battle. And of course you never want to see margins sucked out, but maybe margin could be distributed a little bit more along that supply chain. Because as far as I know right now in AI, there’s only one company making a lot of money.

Pat Moorhead: This announcement irritates me so badly, I can’t even tell you.

Daniel Newman: Are you going to be okay?

Pat Moorhead: I’m going to be okay, but I might need to anybody have a paper bag I can bring into? So first off, snowflake doesn’t have any of their own infrastructure. Zero. They don’t have their cloud, they don’t have a cloud, they don’t have a data center. They sit on top of AWS, Azure, GCP and more. So it’s like a joke. It’s like nobody talked about this anyways, it’s like, do people not realize that? And second of all, I think the only winner here is Nvidia.

Daniel Newman: Always.

Pat Moorhead: Because you have them locked doubling down on NeMo, which again in Nvidia’s vertical and horizontal integration play, it’s huge. Nvidia wants to get to the point where, hey, you might have some competitive hardware, but we’ve got you locked into the software framework And you’re not going to move. They’ve done this in gaming and Jensen’s going to do this in the cloud. So yeah, I think this is just a big nothing burger for Snowflake. It is yet another feather in the cap for Nvidia. And yeah, Jensen was on stage, but it was local. Okay, Jensen, he had to drive down the street to get there. And Snowflake is every three to five years, the new data thing comes along and everybody jumps on it. Snowflake happens to be the merry-go-round, the circus ride that everybody is-

Daniel Newman: Flavor of the week Pat, flavor of the week. It’s just-

Pat Moorhead: It is. And you know what? There’s going to be another one in three years. And-

Daniel Newman: Well, there’s still companies like Cloudera and Teradata that are managing exponentially more data and for more well designed for hybrid than-

Pat Moorhead: Snowflake does not have an on-prem play, black and white. They’re a public cloud play. They do what they do well. They’re using their run up in valuation to go buy companies to become someday a bigger company, but they don’t do hybrid. So good luck with that strategy moving forward. We have seen with all the public clouds, there’s still room, but it’s not an ending. Hey, we got to go. I am talking way too much on here. Great show Dan. Great to see you buddy. By the way, Six Five Summit has been ungated. Check out YouTube Futurum Group and check out more insights and strategy YouTube, and then we’re going to be slowly broadcasting these delayed live on Twitter and LinkedIn. Thanks for tuning in. Give us feedback, give all the good feedback to me. Give the bad feedback to Dan.

Daniel Newman: Oh man, here’s some bad feedback. Snowflake only loses $2.67 a share.

Pat Moorhead: Oh my God, I love it. Let’s run it up baby. It’s Snowflake.

Daniel Newman: $8 billion in value burning cash.

Pat Moorhead: I love it. Hey, great show. Take care. Love y’all. Smooch, smooch.

Patrick Moorhead
+ posts

Patrick founded the firm based on his real-world world technology experiences with the understanding of what he wasn’t getting from analysts and consultants. Ten years later, Patrick is ranked #1 among technology industry analysts in terms of “power” (ARInsights)  in “press citations” (Apollo Research). Moorhead is a contributor at Forbes and frequently appears on CNBC. He is a broad-based analyst covering a wide variety of topics including the cloud, enterprise SaaS, collaboration, client computing, and semiconductors. He has 30 years of experience including 15 years of executive experience at high tech companies (NCR, AT&T, Compaq, now HP, and AMD) leading strategy, product management, product marketing, and corporate marketing, including three industry board appointments.