C2C Fortnightly News: Cloud Disruptions and Strategic Acquisitions - NC2C006

Chris Miles/Will Collins: 0:00

But with McDonald's that makes sense. I don't care for their burgers much.

Chris Miles/Will Collins: 0:03

I prefer their real estate, I prefer their real estate, yeah.

Alex Perkins: 0:08

Where's the movie right? We had a movie about McDonald's. Now we need the Equinix. The Equinix founder story.

Tim McConnaughy: 0:15

The McReal Estate yeah.

Chris Miles/Will Collins: 0:19

Welcome to the Cables to Clouds podcast. Cloud adoption is on the rise and many network infrastructure professionals are being asked to adopt a hybrid approach. As individuals who have already started this journey, we would like to empower those professionals with the tools and the knowledge to bridge the gap. Hello, and welcome back to another episode of the Fortnightly Cables to Clouds News. If you're watching this video, you might notice things look a little bit different, and that's because I am currently in I was going to say Louisville, Kentucky, but I'm technically not in Louisville Kentucky. No, I'm in Sellersburg, Indiana, which is one of my old stomping grounds, and I'm here with our good friend Will Collins, a friend of the pod, guest on the pod many times, and he's welcomed me into his home. We have Tim and Alex here too as well, which is nice.

Tim McConnaughy: 1:24

But we're not in the cloud gambit studio no, we're not in the cloud gambit studio, the studio where the magic happens.

Chris Miles/Will Collins: 1:32

It is a very nice setup in here, very, very professional compared to, uh, what I have at my house, so I'm very jealous. Um, but no, it's been a great week We've been here. Alex was actually in town.

Chris Miles/Will Collins: 1:44

We all went to have dinner at Yvonne Sharp's house another friend of the pod and even recorded a Cloud Gambit episode there, which Gambit is already out by the time you hear this. So make sure to check that out. We did a nice roundtable with Alex, will, yvonne and myself, so, yeah, that will be something to check out. So with that, let's get into the news. I'll throw it over to you, tim.

Tim McConnaughy: 2:23

Yeah, okay, so kind of starting the news, this week we have a blog release from Cisco, who recently finally closed the acquisition of Splunk. So they put out kind of a better together blog post between Chuck and oh, I forgot his name the EVP of Splunk. We'll put the link in the show notes. I don't remember his name. One of the most interesting things about this blog post is how unsurprising any of the quote unquote insights are from the acquisition announced.

Tim McConnaughy: 3:01

Uh, we on the pod were saying that basically the cisco was buying splunk for the, uh for the ai piece of it, uh for specifically for data set and and uh, stuff for the model training, also for observability, and interestingly this went into the security bu. Uh, so long and short. Basically, this blog post covers all the reasons for the acquisitions and what the future looks like with Splunk inside of Cisco, and it does a first, second, third on down the list and covers basically everything we already had said back when this was announced. So it is fun to be vindicated there. But yeah, I mean, there's nothing really surprising here. So let's see.

Chris Miles/Will Collins: 3:45

I would have thought a $28 billion investment went towards AI.

Tim McConnaughy: 3:49

Yeah, I mean, that was the number one, that was the number one reason, and I think we called it all the way back then. Right, we said that was going to be why Cisco paid $28 billion to this company.

Alex Perkins: 3:59

Yeah, and it was Gary Steele is the guy you're thinking of Gary Steele.

Tim McConnaughy: 4:02

that's right.

Alex Perkins: 4:04

Yeah, like you said, this isn't really much of anything. It's interesting how much they're pivoting towards security. It seems like everything is just being integrated into security with them.

Tim McConnaughy: 4:16

Lately, observability and security are their two new golden tickets, along with AI, which is everybody's golden ticket. Right, so hard pivot from Cisco. All right, let's move on to the next story. It's not even really. This is another one that's kind of like almost like not even a story, but Microsoft has announced that they are also on board with the same with AWS and Google, with the egress. Yeah, I know, right, you know, with the European Data Act. No way this to do it. So it's a slightly different than the way that AWS did it, which was kind of hand wavy, you know, hey, we'll just kind of cancel out your egress charges for you to leave Something more, in line with the way Google's doing it, where you know, if you leave and close your account, we'll give you a reimbursement. Do?

Chris Miles/Will Collins: 5:17

you have to close your account with all of them.

Alex Perkins: 5:19

Not with AWS.

Tim McConnaughy: 5:20

Yeah, AWS did not say you had to close your account and actually I'm most lenient, I think.

Alex Perkins: 5:25

I think AWS even said like they, you know, put out the message basically this isn't burning bridges with us. You're welcome to come back at any time. While GCP and Azure were like no.

Chris Miles/Will Collins: 5:35

Yeah, like get out Goodbye. Yeah, I thought it would be the opposite.

Tim McConnaughy: 5:40

So did I. You know, I thought so as well, but given the way, given the landscape we're in right now, like the way that the companies are jacking and where they are kind of in the rankings and kind of who's the up and coming and it actually doesn't surprise me as much. I think if you'd asked me six months or a year ago I would have definitely been surprised, but AWS is a little bit on the back foot these days, so it doesn't surprise me that they're like hey, no, you can leave, but please feel free to come back, and the other two would be a little more hark-nosed about it.

Alex Perkins: 6:09

So anyway, that's a good point. If this was like two years ago, I think AWS and Azure might've switched.

Tim McConnaughy: 6:16

They would have thrown their weight around a lot more in a different way.

Chris Miles/Will Collins: 6:19

Yeah, I agree, Will Collins our guest go ahead and cover the next stories that you added, will. These are both very interesting. So we're going to look at the Chips Act and look at some of the work Intel's doing. So if you remember, during COVID there was a big supply chain issue with semiconductor stuff. If you lived around here and you drove past any field you'd see like lines of F-150s for miles An absolute disaster. So I think the point is trying to bring as much of that manufacturing and supply chain work to the US as possible, and so the US Department of Commerce has proposed up to $8.5 billion in direct funding through the CHIPS Act to advance Intel's commercial semiconductor projects in Arizona, new Mexico, ohio and Oregon, and I personally don't think you're going to see the value with this short term. Obviously it's going to take a little while, but hopefully whenever COVID 2.0 comes around we'll have it all figured out.

Tim McConnaughy: 7:22

Don't you put that evil on me, Ricky Bobby.

Alex Perkins: 7:27

I think this is great. It's hard to find a negative in this, I think. Obviously I'm sure there are, but I'm just saying I think overall this is like a net positive right Actually spending money towards bringing some of this stuff to the USS and not just outsourcing it forever and ever and ever and having all those supply chain issues that we saw during COVID.

Chris Miles/Will Collins: 7:50

Yeah, these semiconductors are in everything. We bought a new dishwasher and a new washer and dryer in like the past five months. There's chips in everything now, so when you have no supply of chips, you can't get anything. It seems like.

Tim McConnaughy: 8:06

Yeah, I think my biggest concern with this, and with all, with all uh government programs generally these days, is that, uh, nothing seems to survive for more than a couple of years. Basically, if you don't like the weather, wait for the next election. So I am concerned about the long-term uh uh commitment, if you will, to this program. I agree that it's needed. I am a little bit murky, and I think so is the bill on the actual supply of the supply chain and how they're going to actually make sure that we're not sourcing from adversaries and all this other stuff. Adversaries and all this other stuff.

Tim McConnaughy: 8:45

You remember there was other legislation that was passed not too long ago talking about how we need to make sure that we're not giving our adversaries money, and it was all very high level, and I think at the time we were wondering, well, how are we going to enforce this? So this is a little bit of that as well. It's kind of intentionally vague, but no, it's good legislation. It's in the right direction. Strategically, it's a good idea for America to try to own as much of the supply chain as possible for things like this. Yeah, let's see how it goes.

Chris Miles/Will Collins: 9:12

All right. Next up, the Hindenburg Shorts Data Center for Equinix, alleging inflated profit metrics. Hindenburg is just a research outfit that basically does short sales. That's the bread and butter of their business. And to give an example, what a short sale is, say that Chris and I are friends and Chris likes to play video games. And I say, hey, I want to borrow one of those video games. I borrow it and I'm like man, this game really stinks. Chris has got terrible taste. I am going to sell this for the current value because I think the value is going to go down over time. Well then I buy it back at a lower value, give the game back to Chris, pocket the access, and then I do that again and again and again. That makes money. So that's essentially what a short sell is.

Chris Miles/Will Collins: 10:05

So Equinix shares fell nearly 5% in the afternoon trading as the short seller said that more than 260 facilities globally was using an accounting trick to boost its adjusted funds from operations. So yeah, just an interesting story, and I think this is going to maybe happen more and more. There's going to be some, maybe some shady stuff going on as these big companies try to show that there is an ROI on all this investment we're putting in the infrastructure to get ready for AI. We still have to recoup that investment and try to get whatever multiples that we need as a business, and so there's a lot of back-end stuff that seems to happen in these circumstances, and this is not unique, obviously.

Alex Perkins: 10:47

This is like the last episode. Two weeks ago. I kind of called out Arista claiming that they had taken over the data center market, and it was. They're adding like hyperscaler purchases into that, so it's not really a fair representation of the enterprise data center market. So it feels like there is a lot of like hand-waving going on, especially around all this AI stuff. Like you called out just how people are actually going to make money off of all this and it's easy to say, oh, we're going to see a billion dollar profit next year because of AI, but without actually explaining having realized or given yeah.

Alex Perkins: 11:26

Yeah, exactly, it's just something to say on a financial report these days.

Chris Miles/Will Collins: 11:33

Yeah, I've tried so many of these AI apps and other than like Google using like Yvonne had an example of the facial recognition with Google Photos Like that's a great application of AI. But all the new stuff that tends to be getting released now are things that enterprises are not going to buy. Like, who's going to buy this stuff? A lot of it's figuring it out, science projects nothing burgers, I think at this point, Throwing shit at the wall to see what sticks.

Tim McConnaughy: 11:57

basically, exactly A lot of people don't know that Equinix is a real estate investment trust, so a lot of people are not aware of this. So what that means is that Equinix's actual business is real estate right. Their pops, their locations, commercial real estate right. Just like a lot of people don't know that McDonald's is a real estate company and not a burger company, right, so you got to think about it through that lens. With McDonald's, that makes sense. I don't care for their burgers much, I prefer their real company right. So this is, you got to think about it through that lens, but with mcdonald's, that makes sense.

Chris Miles/Will Collins: 12:27

I don't care for their burgers much prefer their real estate where's the movie?

Alex Perkins: 12:35

right, we had a movie about mcdonald's. Now we need the equinix.

Chris Miles/Will Collins: 12:38

The equinix founder story the mcreal estate, yeah all right, uh, next up, we have two articles. I think both of these are actually from tech crunch. Um, it's simple. So, uh, the first one, we have, um, an organization called inflection. Um, so this article is basically detailing that, after they've raised about 1.3 billion dollars uh, microsoft being the lead investor um, they, they were having some difficulty competing in the AI market against these giants like OpenAI and Google.

Chris Miles/Will Collins: 13:13

And eventually.

Chris Miles/Will Collins: 13:14

Microsoft just ended up absorbing most of the team as well as the technology. So Inflection is a company that kind of focused on creating a more I think they were saying a more personal, conversational AI, but just really didn't make an impact like they wanted to. So now the co-founders and several of the team members have moved over to Microsoft under a new AI division. So the remaining leadership, you know, has this challenge of pivoting the company towards, you know, an AI studio business which is a highly competitive market with Microsoft's role in raising questions about the tech giant strategies in the AI space. So I thought this one was interesting. Obviously, if they get that much in funding and then really can't make an impact and eventually just get absolved by by Microsoft, how do you guys feel about this one?

Alex Perkins: 14:11

It takes I just want to say that this is now an AI podcast. Give us money and then we're going to fail and hire us, because that seems to be all you have to do these days. 1.3 billion, yeah. And then they just get eaten alive Like this is crazy man.

Tim McConnaughy: 14:28

But like $1.3 billion and then not have a product essentially right, Like your product, apparently is not competitive.

Alex Perkins: 14:33

Yeah, then you just get hired as VPs at Microsoft, Like I just don't.

Tim McConnaughy: 14:37

Yeah, we are now an AI podcast. Please give us money so that we can involve ourselves in the AI stack, please.

Chris Miles/Will Collins: 14:49

Well, imagine the absolute gigantic amount of capital and investment pumped in from Silicon Valley into Web3 or all those companies right now.

Tim McConnaughy: 14:55

Oh, I keep seeing articles about how Web3 is going to solve all our privacy issues. There are still people out there beating that drum. I mean, I don't know where they are now, but you're absolutely right. But there's definitely still articles that keep showing up promising Web3 is going to deal with all the privacy issues that we're dealing with.

Chris Miles/Will Collins: 15:12

I think with AI it's got to be. To be successful, it has to be a must-have for business to move forward in the future. If it's not a must-have, then who's going to pay for it? Because budgets are getting locked down pretty tight right now, Cloud spend really high. A lot of companies blew out their cloud spend and they learned from their mistakes. Hopefully.

Alex Perkins: 15:35

It's a good point, like I mean, this is like the second time, right, that you've basically brought up the same point and there has to be value right now, and I think that maybe that will change eventually. And there has to be value right now and I think that maybe that will change eventually, but right now it's not going to be successful without providing some kind of tangible, actual value to companies that are going to use your product.

Tim McConnaughy: 15:58

Yeah, that's an ebb and flow, though. Right, when this first started. That's how these guys got $1.2 billion. Right. It's because everybody would say the word AI and VCs, despite the high interest rates, were still throwing money at everything. Guys got 1.2 billion dollars. Right is because everybody would say the word ai and vcs despite the high interest rates, we're still throwing money at everything, right. So so this is just a natural evolution of that. Can you sell me this pencil?

Alex Perkins: 16:17

it's ai powered, yeah I haven't seen that everywhere yeah, to sell me the pencil, yeah it's, it's AI powered.

Tim McConnaughy: 16:25

Have it sell itself, yeah, exactly.

Chris Miles/Will Collins: 16:29

Yeah, so on the tail of that, talking about cloud spend, we have another article, like I said here, from TechCrunch, about another startup called Polar Signals. So they've raised about $6.8 million to help companies reduce their cloud costs by writing more efficient code cloud costs by writing more efficient code. So their aim is to create a more resource-efficient operation of apps within cloud that could lower expenses. So I think the idea is that you basically plug in your code and they analyze it and they can give you back how to make this run more efficiently within cloud and cut your costs. They kind of advocate for this continuous profiling concepts to monitor and organize your code, performance, resource usage, et cetera, and then they say it could lead to significant savings. I believe this is all based on an open source project called park or parse I don't know how you say it, but yeah, this one, this one's super interesting. Of course it's AI-powered, because what isn't? But yeah.

Alex Perkins: 17:31

I thought this one was interesting. Yeah, it's a little weird to me that I just feel like a lot of companies are going to claim to have something like this, like an AI tool that assists you with writing more efficient code. It just strikes me as a little strange. Co-pilot yeah well, there's so many. There's just assists you with writing more efficient code. It just strikes me as a little strange. Co-pilot yeah Well, there's so many, there's just AI agents. It's weird that there's a company saying that this is their exact model.

Tim McConnaughy: 17:55

That's their model, yeah.

Alex Perkins: 17:56

It's like this is a side thing of some other AI companies that they would just claim that they have, so it'll be interesting to watch.

Tim McConnaughy: 18:03

I think the reason that and who knows if it's real right, because the idea is that you feed in your code and you have it continually look at your code and continually watch the instruction.

Tim McConnaughy: 18:15

What's different here is there's an observability play to it, where it's supposed to watch your workloads at work, using your code and stuff to help profile the efficiency of the code, if you will. So I think that's what they're considering to be the differentiator. So, like Copilot, for example, you plug in your code and it obviously will try to make it you know instructionally. It'll make it efficient in like number of lines or make it more elegant in some way. This, I think, is leaning hard into the idea that the less instructions I have to send and get the same outcome, the less compute that will be used and therefore the money will be saved. I think the proof is going to have to be on the pudding. On this one, they need to be able to demonstrate how much am I saving and is the value of buying your product going to even that out or do better?

Chris Miles/Will Collins: 19:08

I think that it's a serious problem. I was talking to Darko Mezarosh from AWS and we ran it for probably about an hour about how cloud comes along and Lambda can scale practically infinitely, it seems like, and so the industry's traded in speed in favor of optimization. Nothing is optimized these days, and unless you have a real person writing it and thinking through every little thing. It still seems to be that way Because a lot of the code generated from the chat GPTs of the world is just simply not optimized.

Tim McConnaughy: 19:37

Yeah, I mean when I was an enterprise engineer and I'm sure anyone who's ever worked in enterprise. So all of us basically at some point will understand the concept of having to overbuild the network to deal with a shitty app that makes like 10,000 database calls when two would do right Long-running SQL queries Right.

Chris Miles/Will Collins: 19:59

I think eventually we're going to live in a data center wasteland. It's just going to be data centers everywhere.

Alex Perkins: 20:05

Can't, you know, have all these data?

Chris Miles/Will Collins: 20:06

centers to power the solution for global warming.

Tim McConnaughy: 20:10

Yeah, so the problem is real. The point is the problem unoptimized code, god knows with lift and shift, the problem is a serious, real thing. My question then becomes okay, so I lift and ship my crappy app into the cloud, I run this thing on it and it's supposed to help me fix that, but like to what? To what end? Right, at some point, you know, I'm going to have to refactor God knows how much of that, and, and that costs millions of dollars, usually to refactor an app, right, so. So where's the ceiling on this, or where's the floor?

Chris Miles/Will Collins: 20:41

And not to mention we're talking about probably extending the trust of giving another company your entire code base as well, right, With all your production apps and things like that. So I think GitHub's one thing, but it seems like one hack away from a big lawsuit.

Alex Perkins: 21:01

All right, I'm going to cover the last two. So we got another article from TechCrunch. We seem to love them this week. Last two so we got another article from Tech Crunch we seem to love them this week. This really was just a quick summary of NVIDIA's keynote at the GTC conference that they did this week. I actually I did go and watch the entire keynote from Jensen Jensen Wong. It is everything you'd expect. I'd highly recommend go watching it, but it is two hours long. You need to. And jensen, uh. Pauses for applause many, many, many times.

Chris Miles/Will Collins: 21:32

So it's not like him no, not at all.

Alex Perkins: 21:35

What?

Chris Miles/Will Collins: 21:35

a rock star. I mean, he's almost retirement age, isn't he that leather jacket?

Alex Perkins: 21:39

he's just killing it. Yeah, um, but you know, at the beginning they basically talk about a new chip that came out, um, called the blackwell, and it's it's sort of there's a lot of claims about how it, the Blackwell, and it's sort of there's a lot of claims about how it's like the most physically, it's the largest size physically possible, and they basically combined two of them, two of these new GPUs, with a CPU, and if you watch the video like it goes through all the different stages and how these pieces all connect together, basically it turns into this big data center, ai data center, which now Jensen is calling an AI factory. So if you start hearing that term a lot, you'll see it's because of NVIDIA is trying to coin this new term of AI factory, right? So that's why I said well, you're saying we'll live in a data center wasteland, it'll be an AI factory wasteland.

Chris Miles/Will Collins: 22:25

I'm still getting over migration factories.

Alex Perkins: 22:29

For the rest of the keynote, I'm trying to think of some of the things that stood out. There are these new things that came out with called NIMS, which we've discussed this a little bit. They're basically containers like right, If you think of like the Kubernetes space that are. They're containers where you can pre-train them, you can pre-train an AI model on and then you basically have that portability where you can run it wherever you want. They showed a lot of really cool examples of how this works A lot of partnerships, obviously, and companies that are kind of already using these things, and they have this thing called Omniverse Cloud, which is like a central location to pull from all these tools that NVIDIA creates for AI, and all these NIMs can connect into the Omniverse Cloud. You can run your applications like over this whole thing.

Chris Miles/Will Collins: 23:17

I'll try to find a little bit of a, I think it's like NVIDIA inferencing machine or something like that. Yeah, but it's literally basically a container. I was like am I supposed to just know what this means? Like a NIM.

Tim McConnaughy: 23:33

You had to watch the two-hour keynote.

Alex Perkins: 23:35

Yeah, I would definitely find like a highlights video because I watched it all but I was fast-forwarding through a lot of it just to try to break up some of the applause pauses.

Chris Miles/Will Collins: 23:44

That'd be a cool episode for you all in the future. It's going through CPU versus TPU versus DPU, digging in and knowing what the difference is in their applications, like TPUs for inferencing and whatnot. That's a great idea.

Alex Perkins: 23:59

Pete, if you're listening, we might need to borrow you for an episode. All right, so last one, this is basically a blog from Broadcom and this is everybody's favorite topic. That's been kind of dying. Highly recommend reading this article with an open mind, or try to.

Alex Perkins: 24:32

They've made plenty of mistakes, but the whole point of the article is to kind of talk about why they've done a lot of the things they've done and how they've refocused around VCF, the VMware Cloud Foundation, which, chris, we talked about this a couple episodes ago, about the license portability between the GCP VMware cloud thing and the on-prem one. They explain a little bit about that right, how you're supposed to be able to just buy one license and then you can use it kind of anywhere. But it's a very it's an interesting article, that kind of. I'm just glad that he came out and actually talked about some of the reasons behind what they did. I'm just glad that he came out and actually talked about some of the reasons behind what they did, instead of just silence and just slashing all these products and you're everybody's hearing about all these prices being raised and whatnot it's nice to see someone actually talk about the inside and what their thought process is behind some of these.

Chris Miles/Will Collins: 25:23

You guys have any thoughts on that. It's nice to hear, I guess, why things happen behind the curtain. I don't think it's going to make anyone feel any better. True, it at least does provide a bit of reasoning. But at the end of the day, what's done is done right, the blood has been shed, type thing.

Tim McConnaughy: 25:41

Yeah, a lot of damage done. The explanation's for the shareholders, it's not for the customers. That's true when you think about it.

Chris Miles/Will Collins: 25:47

VMware for the customers.

Tim McConnaughy: 25:47

That's true, yeah.

Chris Miles/Will Collins: 25:49

And when you think about it, vmware is a product. It's much more than a product. It's like doors for businesses out there, like you can't operate without doors. Like it's been part of their strategy. It's in, like every data center out there. It's part of like what you breathe, and big changes, of course, are going to have gigantic impacts. And then my free hypervisor got taken away, so that's why I was ticked off.

Chris Miles/Will Collins: 26:13

I'm on the prox box now. So did this article make you feel any better?

Chris Miles/Will Collins: 26:16

No, and I'm not even bleeding millions and billions of dollars.

Alex Perkins: 26:20

You're not one of their top 2,000 customers.

Tim McConnaughy: 26:23

This article's for the shareholders man. I mean, call it what it is, but I mean it is nice at least that he had. Here's our plan and here's why we think we're going to make money. Once all the blood has been spilled and once all the smoke has cleared, please stick withhired or overspent and overdone things.

Chris Miles/Will Collins: 26:50

Broadcom tends to be on the other side of the fence where they've cut things maybe a little too much. You know they're running a really lean shop and really maximizing those profits and gains.

Chris Miles/Will Collins: 27:00

All right. With that, I believe we will wrap up. So thanks again for joining us for another episode of your favorite fortnightly cables to clouds news. I'm not going to make the joke this time, so we're going to break the streak today. What joke is that I don't know if acknowledging the joke is making the joke or if that counts.

Tim McConnaughy: 27:23

What joke are we talking about?

Chris Miles/Will Collins: 27:24

I don't know we're going to avoid it this time. We'll be back with it next, next time, trust me. But yeah, thanks again. If you enjoyed this, please give us like, subscribe, follow all that crap on all the all the socials. Make sure to follow and listen to the cloud gambit, which is we'll have a great episode releasing next week with or no. Actually, by the time this comes out, this will have released yesterday, so if you haven't listened to it already, you're late. You're already behind the times. You need to get up on it.

Chris Miles/Will Collins: 27:53

I haven't listened to it, yet Tim doesn't listen to podcasts.

Chris Miles/Will Collins: 27:56

There's time travel involved here.

Chris Miles/Will Collins: 27:58

Tim is a podcaster who doesn't listen to any podcasts.

Tim McConnaughy: 28:01

Wow, way to just call me out bro.

Chris Miles/Will Collins: 28:02

That's fine. You know what? That's why you're so good. That's why you're so good at it. Alright, with that, we'll leave everyone be. See you later. Hi everyone, it's Chris, and this has been the Cables to Clouds podcast. Thanks for tuning in today. If you enjoyed our show, please subscribe to us in your favorite podcatcher, as well as subscribe and turn on notifications for our YouTube channel to be notified of all our new episodes. Follow us on socials at Cables to Clouds. You can also visit our website for all of the show notes at CablesToCloudscom. Thanks again for listening and see you next time.