More like VMwhere it's at? (Remixed and Remastered)

00:00:00
Ned: Where it’s at—oh. [sings] I got two turntables and a microphone where’s it’s—oh, hey… I didn’t see you there. Welcome to Chaos Lever, a weekly tech news podcast where we look at what’s going on in the tech world from a historical lens. This is a bit of a different episode because I actually spent last week in Cancun enjoying fine beaches that are there and having many delicious drinks. And because of that, Chris and I did not record a new episode. But that’s okay because we have a wonderful backlog of episodes, and some of them are very pertinent to what’s going on today, especially with Broadcom’s acquisition of VMWare. Chris and I had a long conversation about VMWare: it’s history, what’s going on with it currently, and what to expect with the Broadcom acquisition. How accurate were we? How much do you know about the actual history of VMWare? Well, tune into this episode and check it out, and I am going to go have a couple more drinks. [sings] I got two turntables and a microphone. Where it’s at.

00:01:15
Chris: Autopilot.

00:01:15
Ned: [laugh] The increasingly, improperly named autopilot.

00:01:19
Chris: Yes. Auto—

00:01:20
Ned: Yeah.

00:01:21
Chris: —auto-drive into the river.

00:01:23
Ned: It’s lesser-known name, doing business as…

00:01:25
Chris: Auto-crash.

00:01:28
Ned: [laugh].

00:01:28
Chris: Wait. No, that would be Microsoft Office.

00:01:30
Ned: Ouch. Nice. Got them. I will say—and I hope I don’t get struck by lightning—Microsoft products don’t crash on me very often anymore.

00:01:40
Chris: That’s fair. I think they’ve definitely gotten more consistent in at least not blowing up the entire runtime.

00:01:46
Ned: Yes. Now, if they could just fix the thing where Outlook turns—the icon on the taskbar turns red when I have a message but then stays red no matter what I do until I close and reopen Outlook, that would be nice. They could fix that.

00:02:02
Chris: I’m convinced that that, along with other things—spoiler alert—

00:02:06
Ned: [laugh].

00:02:06
Chris: —is Microsoft’s not so subtle nudging of customers into the cloud-based only products.

00:02:12
Ned: Ugh. I hate everything about the words you just said.

00:02:15
Chris: [laugh].

00:02:16
Ned: Shall we?

00:02:17
Chris: We shall.

00:02:18
Ned: Okay. So, obviously, the big headline from last week, in terms of the tech community, was the fact that Broadcom has entered into an agreement to purchase VMWare for 61 billion dollars.

00:02:32
Chris: And seemingly out of nowhere. The way that I read it, this whole thing was start to finish, like…

00:02:36
Ned: Unsolicited.

00:02:37
Chris: Yeah. And, like, four weeks, or some absurdly fast thing. Like, basically—I’m blank on the guy’s name—the CEO of Broadcom—

00:02:46
Ned: Yeah.

00:02:46
Chris: —went directly to Michael Dell and was like, “Here’s what I think we should do, and here’s why it’s really important.” And Michael Dell was like, “Oh, I don’t know.” And then Broadcom guy was like, “Oh, you’re going to get 25 billion.”

00:02:57
Ned: Did, did you hear the beeping? That, that’s the money truck backing up—

00:03:00
Chris: [laugh].

00:03:00
Ned: —to your third vacation house so you can build a fourth.

00:03:04
Chris: And then, you know, Michael Dell had to go to urgent care for the neck sprain—

00:03:08
Ned: [laugh].

00:03:08
Chris: —from nodding his head so enthusiastically.

00:03:10
Ned: Whiplash, really.

00:03:11
Chris: Yeah. It’s sad, sad. He’s the real victim.

00:03:14
Ned: He is in all of this. So that’s the big news. And I thought rather than starting in on directly talking about VMWare and what’s going to happen in the future, we could do something a little more reflective. We could—we could gaze into the past and talk a little bit about what VMWare has meant to us. I kept it close to my heart all these years. So I’ll tell a little my VMWare story, and then Chris, you can tell a little bit about yours.

00:03:44
Chris: Yeah.

00:03:45
Ned: And then we can braid each other’s hair and make friendship bracelets.

00:03:49
Chris: You don’t have any hair [laugh].

00:03:51
Ned: [laugh] I don’t have any bracelets either. So I first encountered VMWare in ESX 3.5, I want to say. It was right around when 3.5 dropped.

00:04:05
Chris: I suspect that’s going to be pretty common because that was the one—the first one that people actually took seriously. Because VMWare 2 did exist.

00:04:13
Ned: Yes.

00:04:14
Chris: That would still be VMWare GSX, mind you.

00:04:16
Ned: Yes. ESX was when they hit 3, right?

00:04:21
Chris: Right.

00:04:21
Ned: They renamed it. So I had known about virtualization prior to that because Microsoft had, like, a virtual server product that they hadn’t paid a whole lot of attention to, but it kind of worked if you needed it to. And I was aware of other virtualized environments because I’d been doing a lot of networking at the time. And so I’d worked with stuff like [GS-3 00:04:42] to build virtual Cisco environments so I could get my CCNA. So I was aware of the idea of emulators and virtualization in general, but I was not aware of VMWare until it kind of crept into my periphery. And then I started putting it on some physical servers—some older, physical servers we had in the small data center. Because I was working in an organization that, I mean, we had 20 physical servers, maybe. So having a spare one was actually like a mitzvah [laugh].

00:05:12
Chris: Yeah.

00:05:13
Ned: Oh, I can throw ESX on this and see—oh, my goodness. Now I can spin up eight VMs. And what it really, really helped with me, at the time, was I was in charge of disaster recovery testing, and the fact that I no longer had to wait to spin up physical boxes and lay down the operating system and try to do an in-place recovery of server—this would have been, like, 2003—was awesome. Now I could do it all virtually. Everything happened much faster. And, if I was going to encounter an error, I would encounter it faster and be able to get over it. So that was my first introduction into VMWare. And then, from there, I started rolling it out across the environment. And I don’t believe we ever had official licenses, so I think we were just riding the, the free version. So we didn’t have vMotion for a good, long time, but then it came to a tipping point, where we did need to introduce official licensing because we now had production stuff running on these ESX servers. And we bought our first SAN. We didn’t have SAN before then. It was all local storage in, like, a NAS—and built it up from there. And then, that really helped me jump into my next job, where I was in charge of much larger and more complex VMWare clusters.

00:06:32
Chris: Right. So mine was slightly different, primarily because, unlike you, I didn’t start my career at, like, a gas station.

00:06:40
Ned: We had four pumps.

00:06:42
Chris: [laugh] One of them was diesel. I started at a university, and the first major computers that I worked on that were not, like, a workstation—

00:06:52
Ned: Mm-hm.

00:06:52
Chris: —where big, big computers, so, like, Sun Fires.

00:06:57
Ned: Mm-hm.

00:06:57
Chris: And the mainframe still existed at that time. We actually had two because, of course, we did.

00:07:02
Ned: Well, redundancy.

00:07:03
Chris: And a lot of our backend services, the connectors that existed, ran in BSD jails.

00:07:08
Ned: Hmm…

00:07:09
Chris: So, from the beginning, the idea of segmenting one computer or one computing platform into many little ones that were separate and completely protected from each other, was just there as an obvious and logical thing.

00:07:23
Ned: Mm-hm.

00:07:23
Chris: So it fell right into that trap of the UNIX nerd understands everything, and the Windows people waste resources.

00:07:31
Ned: Which is still true.

00:07:32
Chris: Yeah. Unchanged.

00:07:34
Ned: Yes [laugh].

00:07:36
Chris: But that was the biggest—I mean, that was the biggest win that I was able to demonstrate. And I showed it with 3.5, but nobody took VMWare in that environment seriously until 4.

00:07:44
Ned: Yeah. That’s really when we started using it for production workloads where I was because 4 introduced, I think, vMotion. Maybe 3.5 had vMotion, but 4 is what really solidified it. And maybe it was multi-processor vMotion. It was something like that, that 4 brought along, and it brought along the much improved vCenter.

00:08:03
Chris: Right. And that was the real big deal was that you can now—I mean, you still didn’t have the actual, automated orchestration or any of those types of tools. They came down a lot later.

00:08:12
Ned: Right.

00:08:12
Chris: But the fact that you could, in production, move a machine from here to there—

00:08:16
Ned: Mm-hm.

00:08:16
Chris: —was an absolute gamechanger when it came to—

00:08:19
Ned: Yes.

00:08:19
Chris: —planning, organization, maintenance windows.

00:08:23
Ned: Everything.

00:08:24
Chris: Everything, yeah.

00:08:25
Ned: But it still was not ready for primetime because I remember trying to move some of the SQL workloads that were on physical machines onto VMWare and the financial engineering team having an absolute fit because their performance dropped. And I did everything I could to tune VMWare—you know, getting down in the guts of the storage system to make sure that it was—had all the proper flags in it and everything. But, at the end of the day, local disk on a physical machine was going to kick VMWare’s butt back in, you know, 2006 or whatever it was [laugh].

00:08:56
Chris: Yeah. There was a couple of drivers. The, the EMC driver I remember in particular, which I’m blanking on the name—

00:09:02
Ned: PowerPath.

00:09:02
Chris: Yes. That was a gamechanger, but that didn’t come from VMWare.

00:09:05
Ned: No.

00:09:05
Chris: That was—that was a plugin from the—

00:09:07
Ned: From EMC.

00:09:08
Chris: —storage provider.

00:09:09
Ned: Right.

00:09:09
Chris: Exactly. And that brought a lot of parity.

00:09:11
Ned: Yes.

00:09:12
Chris: Because then you could connect it to enterprise storage, virtualize that storage, but just present it to VMWare and still have a performing environment.

00:09:21
Ned: Yeah. Once the fiber channel support really started rolling out for it, it was a gamechanger.

00:09:25
Chris: Yeah.

00:09:26
Ned: But, you know, before then, yeah, the SQL server guys were not happy with me, and we ended up moving it back to the physical box. But, you know, that’s—I’m not bitter about that at all.

00:09:34
Chris: Well, I mean, and also they are just, like, never happy.

00:09:38
Ned: Right.

00:09:38
Chris: So…

00:09:39
Ned: So VMWare didn’t invent virtualization.

00:09:43
Chris: No, not at all.

00:09:44
Ned: And they certainly weren’t the first to implement it. What they did have going for them was they came around at a time where Linux was becoming a real thing outside of tinkerers. It became something you would want to run production workloads on. x86 was becoming the commodity platform of choice in data centers, and they had a virtualization solution that supported—it was a Linux-based solution that supported x86 processors and could run both Windows and Linux workloads on it. It was perfect timing for them, and they rode that wave.

00:10:23
Chris: Yeah. Yeah. And it was—allowed people to still deploy what they thought they needed.

00:10:31
Ned: [laugh].

00:10:31
Chris: And VMWare just gave them what they actually needed.

00:10:35
Ned: Right. The idea that, when you got the request for a 12 CPU and 128GB of memory, you could “give them that”—I’m putting air quotes there—And then you could clearly see they were using, using one processor and 4GB of memory. And then, you know, the theory was, you could resize it for them later, which that often did not happen. But it grew this whole larger ecosystem with tools and software solutions to help you with your virtualization environments. All the backup products to now support virtualization, and new backup products were invented from new companies that were specifically around virtualiz—Veeam wouldn’t exist without VMWare.

00:11:21
Chris: Right. And it also gave system administrators, who were either unable or unwilling to migrate off of an old platform, a way to survive. Because you had an old machine on failing hardware—

00:11:35
Ned: Yep.

00:11:35
Chris: —you could vMotion it—

00:11:37
Ned: No. P2V it.

00:11:38
Chris: P2V it.

00:11:38
Ned: Yeah.

00:11:39
Chris: Which is funny. Because in my head, I’m hearing the word “PlateSpin, PlateSpin, PlateSpin.”

00:11:42
Ned: Yeah, you are [laugh]. Well, that was the thing. I can’t tell you how many times I went into a consulting engagement and found them running Windows NT boxes on VMWare because you could, and it was running some backend service that no one was going to re‑platform or re-write, so it was going to live on that NT box until the company died, basically [laugh].

00:12:02
Chris: Or they upgraded to XP.

00:12:05
Ned: [laugh]. I don’t even know if that would—yeah, sure, why not.

00:12:08
Chris: [laugh].

00:12:08
Ned: We’ll go with that. So I guess what we’re trying to say is VMWare rode a wave, brought some innovation to the industry, and then continued to try to innovate for a certain amount of time around their core virtualization technology. They tried to branch out into storage with vSAN. They tried branch out into networking with NSX. And then, more recently, they tried to branch out into things like security and also wider area networking with their SD-WAN products, like, I think VeloCloud. Isn’t that the one they bought?

00:12:41
Chris: It sounds right.

00:12:43
Ned: [sigh] But, ultimately, I think they became too entrenched in the tools that they had created. And kind of to tag off of what you just said: that it enabled administrators who didn’t want to change to keep doing things the same way. That became a real problem later when people wanted to do things a different way and VMWare couldn’t make that pivot anymore.

00:13:07
Chris: Right. Yeah, it was a complicated situation they found themselves in because the core product is and always will be ESX.

00:13:14
Ned: Yes.

00:13:16
Chris: And the thing about ESX, and to a lesser extent but similar platform area in management, is that it’s simple. It just works.

00:13:25
Ned: Yes.

00:13:25
Chris: Its infrastructure is a service in your data center, effectively. Everything that you just talked about that piles on top of it, brings complexity.

00:13:32
Ned: Mm-hm.

00:13:33
Chris: And it brings new ways of operating. And a lot of people didn’t like that. And I feel like that VMWare as a company didn’t do a good job with those two streams.

00:13:46
Ned: Right. I am reminded of—remember Bimodal IT?

00:13:49
Chris: Is that when you do two modes of IT at once?

00:13:52
Ned: It—no, it’s when you have two people riding unicycles, and they hold hands to perform the…

00:13:58
Chris: Aw…

00:13:58
Ned: Isn’t that?

00:13:59
Chris: Yeah.

00:13:59
Ned: No. So Bimodal IT was the idea that you would do—half of your IT group would be, like, cloud-focused, and the other half would be more traditional focused, and you would just function in that bimodal., which is, like, kind of true. But I think the problem that VMWare ran into is they had been really catering to that first mode of how people wanted to operate, and then something came along that may have changed the way people operate a little bit. And now we’re getting into the 2012, 2013 era with the rise of AWS, Azure, and to a lesser extent, GCP.

00:14:33
Chris: Absolutely never heard of them.

00:14:34
Ned: Okay. Good. So this has become a 400-hour long podcast.

00:14:39
Chris: [laugh].

00:14:40
Ned: Which is good because my chip says we have 402 hours to record.

00:14:43
Chris: Perfect.

00:14:44
Ned: Excellent. So the cloud era came along, and what the dominating thing about cloud is it’s self-service. So you’re now expecting people at the developer level—you’re now catering to developers to a certain degree, and them being able to spin-up infrastructure on demand. It’s also API-driven. So you are probably not always going to be working in a GUI. You might be, but you also need to have an API interface behind everything that is accessible and well documented. And it’s immediate. You can just spin it up and spin it down whenever you want, and it seems like it’s a virtually limitless pool of resources. Now—

00:15:24
Chris: Unless it’s Azure.

00:15:26
Ned: Nice. VMWare’s reaction to that was to first try to create vCloud Air.

00:15:34
Chris: [sigh].

00:15:35
Ned: Which we talked about. Like, the amount of capital, like CapEx, you have to spend to build and operate a cloud.

00:15:42
Chris: It’s eleventy-trillion-zillion-willion-million-tillion-quinspillion dollars.

00:15:48
Ned: It’s right around there. I think you missed a zero. But the problem was that VMWare is not willing to make the level of investment to build out all the hyperscale or infrastructure to run a cloud service.

00:16:01
Chris: Right.

00:16:01
Ned: So they gave up, and they farmed that out to HVC or something. I can’t remember the exact company. It’s the one that had the big fire two years ago in their data center, so good choice.

00:16:12
Chris: [laugh].

00:16:12
Ned: [laugh]. But they farmed it out somebody else—basically sold off vCloud Air, and for a while it just seemed like they declared cloud bankruptcy and also chose not to update all of their existing products to support the cloud way. And, when they did, it was a complicated mess to roll out. The closest thing was vCloud Director.

00:16:35
Chris: Right. And the goal there was orchestration—

00:16:37
Ned: Mm-hm.

00:16:38
Chris: —across multiple clouds and your own, local data center.

00:16:41
Ned: Right. It was embraced by MSPs, and that’s really the only people that could afford to run the complexity that was vCloud Director.

00:16:49
Chris: Which is going to be a recurring theme from now until we get to the current date and time.

00:16:55
Ned: Right. So I think cloud was the first punch to the VMWare gut, if you will. And they were not prepared for it. They did not handle it gracefully, and they doubled down on-prem, and the way we’ve always done things. And then came containers and Kubernetes. And guess what they didn’t do?

00:17:14
Chris: They didn’t Kubernet.

00:17:16
Ned: They did not Kubernet well. Now, I read a few posts that make the point that, for many enterprises, they have not yet embraced containers and Kubernetes. That is still off in the distance. It’s off in the future.

00:17:29
Chris: It’s a fun buzzword.

00:17:31
Ned: It is.

00:17:32
Chris: But most C-Suite people only understand the Container Store, and they’re like, “Oh, that’s too expensive. We can get a better deal at Target.”

00:17:42
Ned: I mean, they’re not wrong. There’s a lot of value at Target.

00:17:45
Chris: That’s true.

00:17:46
Ned: This podcast has been brought to you by Target.

00:17:47
Chris: [laugh].

00:17:49
Ned: [laugh]. So I think, in many ways, they missed the boat on cloud to a certain degree. And then containerization came along, and they tried to do vSphere Integrated Containers, which was a mess.

00:18:03
Chris: And very complicated.

00:18:04
Ned: And very complicated.

00:18:05
Chris: A lot of moving parts.

00:18:06
Ned: Right. And then they did—

00:18:09
Chris: Don’t forget Photon. That was in there somewhere.

00:18:11
Ned: —Photon—which was very successful for them. In fact, a lot of the deployment software for vCenter and other components of the VMWare stack are all Photon-based.

00:18:22
Chris: Right.

00:18:23
Ned: But nobody else adopted it.

00:18:26
Chris: Yeah. Which is a shame because the opportunity was there.

00:18:30
Ned: The opportunity was there, but I think we have to go back to what made the could so successful and what people like about Kubernetes at a managed level is that it’s relatively simple to deploy and start using.

00:18:41
Chris: Right. Back to the whole hook of ESX: it does what it does, and it just works.

00:18:46
Ned: Yes. Deploying one ESX server is very simple. And I’m going to say ESX because I don’t feel like saying ESXi, but we all know what we’re talking about. Deploying three of them, not super difficult. Adding a vCenter—okay, not bad. Still I’m not, like, at a over-complicating the thing, but then you add vSAN, and you add NSX, and then you add all the other add-ons, like, vRealize Operations Manager. And, suddenly, you have this ridiculously complex thing that’s really hard to wrap your brain around, and it’s really hard to deploy or even try.

00:19:22
Chris: Right. And even in terms of maintenance, the amount hands-on keyboard time just continues to go up at an exponential rate.

00:19:28
Ned: Right. So because of that, people turned to Software as a Service, or Infrastructure as a Service, or Anything as a Service because I don’t want to manage this complexity. Please, dear God, somebody take it out of my hands.

00:19:41
Chris: Are you trying to imply that VMWare is directly responsible for Salesforce?

00:19:46
Ned: No. But an argument could be made. I think what it’s more indicative of is that people got used to the cloud experience where, if I want to try something, I click three buttons and it’s up and running in my [tenant 00:19:58], and I can try it.

00:20:00
Chris: Right.

00:20:01
Ned: If I want to try a new product or feature offering in my VMWare cluster, well, I can’t just put it in my production cluster. I have to spin up a whole separate cluster. And that’s, like, a four-day exercise in frustration, so I’m not even going to bother. Could they have built a product that was—as a Service-based and deployable in your data center? Yeah, they could have. Did they?

00:20:25
Chris: No.

00:20:26
Ned: Well, there’s vSphere Cloud Foundation, which is a product that exists. And if you have eleventy-billion dollars, you can buy it and put it in your data center. But guess who it was really targeted at?

00:20:38
Chris: MSPs [laugh].

00:20:39
Ned: There you go. And we got there. So, really, they’re catering to MSPs at this point—

00:20:43
Chris: Yeah.

00:20:43
Ned: —and not, not the little guy—the little guy who is going to just move to the cloud, essentially. And that catering to the little guy thing is going to come back around in a little bit. So you just had this situation where VMWare had a solid product that worked very well, but they hadn’t embraced the shift in the way people wanted to consume and manage their infrastructure. And now we’re ten years later, and they still haven’t embraced it. And they tried to do the whole Tanzu thing. That did not go well.

00:21:16
Chris: No. No. Because, again, it was very complicated.

00:21:20
Ned: Mm-hm.

00:21:20
Chris: It required a different instillation. It required a different license scheme—

00:21:25
Ned: [laugh].

00:21:25
Chris: —which I’m sure we’ll get to. It was poorly documented. And, if you were not a capital E vExpert—because the V is not capital—you were lost.

00:21:38
Ned: When Tanzu first dropped, I tried to deploy it in my home lab. And I know a thing or two when it comes to VMWare. And a day and a half into the instillation process, I gave up. I just—not worth it to me.

00:21:53
Chris: Yeah.

00:21:53
Ned: Because, if I wanted to mess around with Kubernetes, I could deploy vanilla Kubernetes on those same servers in a couple hours. I had, like, a nice instillation script. And if I wanted to use OpenStack’s installer, it would be even faster. So why would I go through all the hassle of deploying this thing just to try it out?

00:22:12
Chris: Right. And the trouble is the other idea as to why would want to do that, as a customer, is I can manage everything through the mythical, single pane of glass.

00:22:23
Ned: Single glass of pane. Yes. That was the idea is, I am going to have this one interface that I log into and I’m going to manage all of my containers and all of my servers and all of my vast, virtual infrastructure real estate through this one interface. So, like, I mean, okay, but couldn’t you have extended that interface to interact with Kubernetes clusters that were using upstream Kubernetes bits on their own physical servers—

00:22:53
Chris: Right.

00:22:53
Ned: —instead of trying to mash the two together on the same clusters with the same operating system? I just feel like there is other ways they could have handled it.

00:23:02
Chris: Yeah. I mean, my theory on this goes back even before Tanzu. Like, NSX was a nightmare when it came out. Same exact thing.

00:23:09
Ned: Yes.

00:23:10
Chris: At a—when it first came out, you literally could not install it yourself. Like, you weren’t allowed.

00:23:14
Ned: Right. You had to bring in someone directly from VMWare or someone who had gone through the two-week NSX training.

00:23:21
Chris: And I believe when it came—first came out, you had to be a VCP. Like, you had to be a significant VMWare expert—

00:23:27
Ned: Right.

00:23:27
Chris: —in order to get your hands on the installer bits, because NSX was that difficult.

00:23:31
Ned: And you know what that tells me? It was too complicated and not ready.

00:23:35
Chris: Exactly.

00:23:37
Ned: This is a recurring theme.

00:23:39
Chris: [laugh].

00:23:39
Ned: So I think we painted a pretty clear picture of where VMWare started, why it got the dramatic boost, and then kind of what happened afterwards. There’s a whole bunch of financials that also go along with this. You know, when EMC picked them up for under a billion dollars—that was a thing that happened—

00:24:00
Chris: Dark days.

00:24:00
Ned: —and then Dell bought EMC, and then basically rode the VMWare train to payoff their debt from buying EMC and then spun out VMWare when they were quote-unquote, “finally free.” And that has lasted a grand total of six months before being scooped up out of the public market by Broadcom. So, in a lot of ways, VMWare was the revenue—not even revenue—the income-generating engine that paid off debt for EMC and then Dell for a long period of time. So there’s a lot of financial machinations going on here, which maybe brings us to the deal at hand.

00:24:39
Chris: We might as well.

00:24:39
Ned: Okay. Let’s shall. Okay. So Broadcom is buying VMWare for 61 billion dollars, which is going to be a combination of stock and cash. They have taken out a 32 billion [laugh] dollar loan to make up for the cash portion of things, and then the rest of it is going to be stock. Now, Broadcom released some slides to their investors explaining how to justify this massive purchase, and they said that they are currently—VMWare is performing at 4.7 billion dollars of EBITA, which is word that was introduced to both of us at an all-hands meeting, and I still don’t understand why.

00:25:28
Chris: Because it sounds good.

00:25:29
Ned: It does not.

00:25:29
Chris: It’s all capital letters.

00:25:30
Ned: It is an ugly word.

00:25:32
Chris: It’s not actually a word.

00:25:33
Ned: Ugh, acronym.

00:25:34
Chris: There we are.

00:25:35
Ned: So, Earning Before Income, Tax, Depreciation, and I can’t say the last word.

00:25:42
Chris: Amortization.

00:25:42
Ned: Thank you. I literally cannot say that word.

00:25:46
Chris: [laugh].

00:25:46
Ned: So Broadcom is targeting, in three years, to be at an annual 8.5 billion dollars in EBITA, which there are two ways to raise the amount of money you are making. You can introduce new products or charge your customers more, or you can cut costs.

00:26:06
Chris: Right.

00:26:07
Ned: Guess what Broadcom is going to do?

00:26:09
Chris: Gamble at Monte Carlo?

00:26:12
Ned: Not quite.

00:26:13
Chris: No.

00:26:14
Ned: Both—they’re going to do both.

00:26:16
Chris: Nice.

00:26:17
Ned: So, if you think about the current crop of VMWare customers, they’re kind of stuck with that technology, and the vast majority of them are just going to have to grin and bear it when the licensing costs go up by 50 percent.

00:26:28
Chris: Which they will.

00:26:29
Ned: Which they will.

00:26:30
Chris: Immediately.

00:26:31
Ned: Because that, that is what they’re focusing on. And then, also, hey… a lot of redundancies when you get acquired, so there is going to be a massive reduction in staff. And we know that because if we look at previous purchases by Broadcom—Symantec was one, Computer Associates was another—Both of those had massive reductions, and we’re talking they laid off 70 to 80 percent of the workforce at both of those companies after the acquisition.

00:27:00
Chris: Right. Because they are using them solely as cash-cow.

00:27:04
Ned: Right.

00:27:04
Chris: Meaning their products are static, and they will stretch every single dollar they can out of the existing contracts, which will last far longer than they should because they’re so entrenched.

00:27:16
Ned: Exactly. Broadcom had a separate strategy and investor meeting in 2021, and the slide deck showed that they are focused on the top 600 customers across all of their platforms. Which means if you are not a giant customer, one of those 600, they are not going to focus on you at all.

00:27:38
Chris: Or care about your needs.

00:27:40
Ned: Mm-mm.

00:27:41
Chris: Return your calls.

00:27:42
Ned: Mm-hm.

00:27:42
Chris: Buy you sandwiches. Remember your birthday.

00:27:44
Ned: They are going to slash the sales staff down to—

00:27:47
Chris: Wash your car.

00:27:48
Ned: Uh-huh.

00:27:49
Chris: No. You were saying something important, too, I’m sure.

00:27:51
Ned: Yeah. I mean, blow out your birthday candles for you. They’re not going to do that.

00:27:54
Chris: They won’t even bring you a cake.

00:27:55
Ned: I know. It’s going to be pie.

00:27:57
Chris: [sigh].

00:27:57
Ned: And you’re going to like it.

00:27:59
Chris: That’s…

00:27:59
Ned: Mincemeat pie. Ugh.

00:28:00
Chris: Woah. Woah.

00:28:01
Ned: Woah, I know [laugh]. I’m so sorry.

00:28:03
Chris: I didn’t realize we were going to get nasty.

00:28:06
Ned: [sigh] So they’re going to fire most of the marketing, most of the R&D, most of the sales. And then these are just general predictions that have been substantiated by many people. There’s a lengthy article by Brian Madden, who is a former employee in their VDI product, Horizon ONE, and we’ll include that link in the [show notes 00:28:31], but he had some rather unkind things to say, bearing in mind he’s not happy with VMWare in general, but also he’s sort of laid out the case for why there’s going to be these massive layoffs.

00:28:41
Chris: Right.

00:28:43
Ned: So what does this mean for VMWare? What does this mean for current VMWare customers? And what does this mean for the tech industry as a whole? Chris?

00:28:54
Chris: So I would say the first thing to know is we’re lucky that VMWare ESXi 7.0 came out when it did.

00:29:03
Ned: Mm…

00:29:03
Chris: Because it was a legitimate big deal.

00:29:06
Ned: Yes.

00:29:07
Chris: Nice upgrade.

00:29:08
Ned: Mm-hm.

00:29:08
Chris: A lot of strong under-the-covers features as well as some of the stuff we’ve been talking about, Tanzu, et cetera, et cetera.

00:29:13
Ned: Right.

00:29:14
Chris: However, I think it is likely that it is also the last major upgrade to ESXi that we’re going to see.

00:29:22
Ned: Right. Don’t expect any crazy improvements. It’s all going to be small, incremental updates from here on out. So, as far as VMWare goes, we can consider most of the products they have now are not going to change significantly over the next ten years.

00:29:38
Chris: Right.

00:29:39
Ned: If you like the way their products work now, great. If you don’t, too bad.

00:29:46
Chris: [laugh] And it’s going to put people into an interesting situation because over the past five to ten years, licensing costs have consistently gone up. As one of the only major—and I’m putting these in huge air quotes—“innovations” VMWare has had is in the way that they charge you for the stuff you use or don’t use.

00:30:06
Ned: [laugh].

00:30:07
Chris: Enterprise Plus. That’s all I have to say. Enterprise Plus Plus Plus.

00:30:11
Ned: Isn’t’ there a platinum tier?

00:30:12
Chris: Super Ultimate Platinum Explosion Enterprise Plus Secret Top-Secret.

00:30:17
Ned: Jubilee.

00:30:19
Chris: Dot com.

00:30:20
Ned: Dot org [laugh].

00:30:23
Chris: But the thing is, if you have the stuff already, you’re not going to upgrade it. You’re probably not going to pay for the license, so you’re not going to get additional features going down the line—

00:30:32
Ned: Right.

00:30:33
Chris: —and you’re just going to have an environment that it falls further and further away from current, when it comes to patching. Because my guess would be that they’re going to pull an HP here, which means that if you don’t have a valid license, you can’t download even security patches.

00:30:48
Ned: Oh… that could happen. You’re not wrong. Which is great for everyone.

00:30:53
Chris: Yeah.

00:30:54
Ned: Really, really want to spread around the, you can’t get security on your products. I’m sure that’s fine. That’s not going to impact anything.

00:31:01
Chris: I do wonder, you know—because, like I said, I think it’s a very static, stagnant environment, and it’s stable enough. So people could conceivably run indefinitely on ESXi as it stands right now.

00:31:13
Ned: And they probably will.

00:31:14
Chris: Yeah. But for new people, and there are new people. New people do exist.

00:31:19
Ned: That seems unlikely.

00:31:20
Chris: I have heard rumor that people make new people every day.

00:31:22
Ned: No.

00:31:24
Chris: Other virtualization products that have an enterprise scale are going to all of a sudden become a lot more appealing.

00:31:33
Ned: And I think that brings us into the larger tech industry.

00:31:36
Chris: Yeah.

00:31:37
Ned: So I want to sideline that for one second and just mention that all these layoffs are also going to accompany by people just leaving the company voluntarily. They’ve been frustrated for a long time, they’ve seen a lack of innovation or haven’t been able to work on the projects they want to. You are going to see an enormous number of start-ups happen over the next 18 months of people who were formerly at VMWare, and now are pursuing whatever the idea they had at VMWare that they couldn’t get realized.

00:32:08
Chris: Right.

00:32:09
Ned: So that’s exciting.

00:32:10
Chris: Which is kind of funny because that’s one of the way that VMWare as a company originally got started.

00:32:18
Ned: [laugh]. Sure.

00:32:18
Chris: People working at VMX were like, “This is nonsense. We don’t need to charge $200,000 per node. We can do better with x86.”

00:32:27
Ned: Yep. And the cycle repeats itself. So that’s one good thing. But I think the thing that you just alluded to was this is a benefit to the other competitors in the hybrid marketplace.

00:32:39
Chris: Correct. Because I think everybody sees that this is a financial play. It is not a technical play.

00:32:47
Ned: No.

00:32:47
Chris: Which means that the idea of buying VMWare suddenly got fraught, where it never was before.

00:32:54
Ned: It was a slam-dunk before. Now, yeah, you’re thinking twice. You’re evaluating other options. And so, what are your other options when it comes to running on-prem self-hosted infrastructure?

00:33:08
Chris: Right. I mean, one of the things that I think is going to be interesting to see in terms of the response, is the private cloud products from AWS and Azure.

00:33:19
Ned: And GCP with Anthos.

00:33:20
Chris: Right. And I forgot about—I always forget about that.

00:33:23
Ned: I like to bring it up to bother you.

00:33:25
Chris: So these are trillion-dollar companies, literally in some cases.

00:33:29
Ned: Yes.

00:33:30
Chris: In all cases? In all cases.

00:33:31
Ned: Yes.

00:33:33
Chris: But they never really did that to a level that I would consider serious.

00:33:38
Ned: Well it’s—so Microsoft always had to walk the tightrope of—a lot of their [laugh] stuff runs on VMWare, so they can’t piss off VMWare too much, but they also have their own internal products that they want to push. All of Azure runs on Hyper-V.

00:33:52
Chris: Right.

00:33:53
Ned: And so if you think about the scope and scale of that, being able to bring that same hypervisor down to your on-prem environment is kind of attractive, But again, you want that cloud-like experience, and that’s where something like Azure Arc or Azure Stack comes into play. Okay. It’s still running that same hypervisor, but now it’s managed by Azure, essentially.

00:34:13
Chris: And you get to say that magic phrase.

00:34:16
Ned: What?

00:34:16
Chris: Single pane of glass.

00:34:18
Ned: Ugh. Oh, God, it, it burns. Outpost, same idea.

00:34:22
Chris: Mm-hm.

00:34:23
Ned: Anthos, same idea. You use the control plane that exists up in the cloud to manage your on-prem resources. Okay. So we got that. I think another big player and benefactor is going to be Red Hat with their OpenShift product.

00:34:38
Chris: Right. Probably the biggest opportunity they are ever going to have.

00:34:43
Ned: Right. Now, OpenShift by itself is extremely focused on containers. It’s Kubernetes under the covers, with some pretty add-ons and a nice UI. But, at the end of the day, it’s basically OpenStack running underneath everything, and then Kubernetes on top of it, and then some developer tools. So it doesn’t run virtual machines, per se, but they have been working on lots of virtual machine extensions, in case that’s a thing you might want to do later. Now is later.

00:35:15
Chris: Right. And that’s actually an interesting point. Is this—so, like, forget about the pure replacement for VMWare—

00:35:21
Ned: Right.

00:35:21
Chris: —is this the impetus to move more customers into containers?

00:35:26
Ned: I think it absolutely is. And the toolsets to do it—even if you just want to take your clunky VM and bundle it up as image and deploy it in, [laugh] you know, in a container platform, you can do it.

00:35:39
Chris: I mean, the thing about that is it’s been doable for a while.

00:35:42
Ned: Yes.

00:35:43
Chris: People don’t often do it because it’s silly in terms of how containers are used and why.

00:35:48
Ned: Right.

00:35:48
Chris: But, as a technology, does it work, yes/no?

00:35:51
Ned: Yes.

00:35:51
Chris: Yes.

00:35:52
Ned: Can you vMotion it? No?

00:35:54
Chris: It’s…

00:35:56
Ned: It’s complicated.

00:35:56
Chris: Let’s not ask questions.

00:35:58
Ned: Okay. But, yeah, okay. So we have the major cloud players smell blood in the water. Red Hat is going to make a big push on their side, and there’s one other eternal foe with VMWare that I heard cackling off in the distance when the buyout happened, and that’s Nutanix.

00:36:14
Chris: Ohh, yes. AHS?

00:36:16
Ned: AHV.

00:36:16
Chris: H—oh, I was so close.

00:36:20
Ned: You were. I’ll give you half credit, a quarter credit.

00:36:23
Chris: Three-quarters credit?

00:36:25
Ned: Not likely.

00:36:25
Chris: 210 percent credit?

00:36:27
Ned: Okay.

00:36:28
Chris: Perfect.

00:36:29
Ned: AHV is their Acropolis hypervisor platform, and they have been trying to get people to move from VMWare to that for all of the reasons that are beneficial to using it. You know, it’s an integrated stack, et cetera, et cetera, whatever. They have been bashing vSAN basically since it’s inception. And if you want to get into an argument with someone, just put a post on Twitter that says “vSAN is better than Nutanix,” and wait.

00:36:58
Chris: [laugh].

00:36:59
Ned: It’s entertaining, but not enjoyable.

00:37:01
Chris: And AHV also has a major advantage over ESX in that you don’t have to pay for the license for it.

00:37:10
Ned: Right.

00:37:10
Chris: You roll it in with what you buy from the vendor.

00:37:14
Ned: Right. And so…

00:37:15
Chris: Or you can actually run the Community Edition if you want to.

00:37:17
Ned: You can. You can. You have both options, but they do have hardware-integrated solutions.

00:37:22
Chris: Right.

00:37:23
Ned: And that’s…

00:37:24
Chris: Which, when you’re talking about ease of use and quick turnaround from shipping to data center to production, it’s nice to have just an “on” button.

00:37:36
Ned: Yes. And it sort of mirrors what people expect to have in the cloud.

00:37:40
Chris: Mm-hm.

00:37:41
Ned: So I think what’s most likely—and I’m kind of surprised Nutanix didn’t get scooped up by Google. I really thought that was going to happen at some point, and it didn’t. But I think what we’ve really got here is just all the existing hybrid players are going to come running at VMWare customers, especially the smaller ones that aren’t going to get the white-glove treatment anymore, and start eating into its market share. And Broadcom may find that VMWare is not nearly as profitable of a purchase as they once thought it was. Now lastly, I want to address what do people who currently work at VMWare do. Because I know some of the listeners that we have, just from a pure numbers perspective, may work at VMWare.

00:38:23
Chris: There’s a chance.

00:38:24
Ned: There’s a chance. And you might be feeling a little despondent now or betrayed or upset, and those are all completely valid things to feel. And I think you should sit in those feelings for a little while and let yourself be bothered by it.

00:38:37
Chris: Yeah.

00:38:38
Ned: If you feel like you need to grieve about this, do so. And then don’t do anything off the cuff. Don’t do anything spontaneous. Ask yourself, do you hate your current job at VMWare? If the answer is yes, get out. There was no reason not to before, and there’s even less [laugh] reason not to now.

00:39:01
Chris: Rubrik is a mile away.

00:39:02
Ned: It really is.

00:39:04
Chris: No, it really is.

00:39:05
Ned: No, I know. I’ve been there.

00:39:06
Chris: [mocking] You’ve been there.

00:39:07
Ned: I’ve been there. They have pretty plants, and a really cool tree.

00:39:11
Chris: [laugh] They’re not the only—

00:39:11
Ned: So if you ever go—

00:39:11
Chris: —people on Earth that have a plant.

00:39:12
Ned: If you go to the Rubrik campus, they have this really cool tree. And I just recommend—if you happen to be there, check it out. Yeah. Yeah. So, if you don’t hate your current job, if you like your current job at VMWare, maybe stick around for a while. You’re going to get a severance package eventually, and in the meantime, you’re not going to have to work very hard.

00:39:31
Chris: No. They’re going to probably dissuade that sort of behavior.

00:39:34
Ned: [laugh] They sure are. Do not panic, and talk to someone. I have seen an outpouring of support on Twitter and LinkedIn saying, “if you want to talk, my DMs are open. My messages or whatever is open. Reach out to me. I’m happy to chat.” And there’s just a bunch of people—former VMWare employees or just VMWare-adjacent folks who are happy to support and provide guidance and advice. Talk to those people before you do anything rash.

00:40:02
Chris: Right. And that’s been happening in public in ways that really kind of surprised me. You know, I made the joke about this being a wake or sitting shiva. And in reality, in the tech circles that I have been observing, it really feels like that’s what’s happening.

00:40:18
Ned: Yeah.

00:40:18
Chris: You know? The VMWare what was is being discussed, and people are…

00:40:24
Ned: Speaking of it in the past tense.

00:40:25
Chris: Yeah. But they’re doing it in a way that feels like they’re getting things off their chest.

00:40:29
Ned: Yeah.

00:40:29
Chris: They have some feelings about what has happened in such a short period of time to a company that they probably felt a connection to, and a mission to assist the company in solving or serving.

00:40:44
Ned: I—we were both deeply invested in the VMWare ecosystem for years. I mean, I built half of my career using VMWare.

00:40:52
Chris: My middle name is VMWare.

00:40:54
Ned: It’s odd but true. I don’t know how your parents had that prescience, but good for them. Is that a Germanic root?

00:41:02
Chris: It comes—

00:41:02
Ned: VM, var?

00:41:03
Chris: [laugh] V-M-var.

00:41:07
Ned: [laugh]. So I, I guess, just to wrap-up the whole conversation, you know, we wish everyone at VMWare well. We hope that you land on your feet. This is the natural churn of the tech lifecycle. And, if you’re a current VMWare customer, just be aware that now your decision needs to be a little more nuanced about what you want to do next.

00:41:27
Chris: And the most important thing about that is—I mean, you said “don’t panic” to the employees, but “don’t panic” to the customers as well.

00:41:33
Ned: Right.

00:41:33
Chris: Because the options that we talked about are legit. They’re all backed by giant companies that have done huge amounts of research—

00:41:40
Ned: Mm-hm.

00:41:41
Chris: —and development, and they are legitimately production worthy.

00:41:45
Ned: And they’re all going to come knocking at your door giving you free demos.

00:41:48
Chris: And sandwiches.

00:41:49
Ned: And lots of sandwiches.

00:41:50
Chris: And cakes with candles. It’s going to be your birthday every day.

00:41:53
Ned: It [laugh] really is, and they’re going to blow those candles out with you. It’s a partnership, so enjoy that. Shall we lightning rounds—or not really lighting round?

00:42:02
Chris: [laugh] Yeah. Lightning… Microsoft? Electrocute Microsoft round.

00:42:07
Ned: [sigh] Aw… All right. So Microsoft Build happened last week, amongst all the other—

00:42:14
Chris: [laugh].

00:42:14
Ned: —Sturm un Drang that was going on. And I wanted to mention a few things that I found interesting about it. So, since I am not a developer, TM, dot com, dot org, et cetera.

00:42:27
Chris: Dot JS.

00:42:28
Ned: Yes. There was a bunch of stuff they announced that I didn’t care about, you know, AI, ML, blah, blah, blah, blah. But some cool stuff that caught my eye. The first thing I noted was that Azure Container Apps is now generally available. It’s been in preview for a while. Essentially, the idea is if you wanted run your containers but not manage Kubernetes, Azure Container Apps is perfect for you because they manage the Kubernetes cluster on the backend, and you’re not even managing at the Kubernetes level. You are just deploying manifests.

00:42:58
Chris: So it’s one, single container and they handle all the annoying details in the background?

00:43:03
Ned: You’re still deploying multi-container applications, but you just don’t have to worry about the underlying nodes that are running it.

00:43:09
Chris: Mm… okay.

00:43:09
Ned: So it’s one step removed from AKS.

00:43:11
Chris: That was just going to be my next question. Okay.

00:43:14
Ned: Yeah. Azure API Management now supports GraphQL passthrough. So, if you want to set up GraphQL-based APIs, you can do it through the API management. That’s useful if you’re a GraphQL kind of person. Single-node Azure Stack HCI is GA, which, when you say those words out loud, sounds strange.

00:43:35
Chris: [laugh].

00:43:35
Ned: Is it a stack if there’s one? Is it hyper-converged if there’s one? And yet, here we are. It’s aimed at edge locations, retail locations, somewhere you don’t need more than one node.

00:43:48
Chris: Right.

00:43:48
Ned: You don’t need that redundancy, but you do want the Azure Stack HCI experience.

00:43:53
Chris: Sure.

00:43:54
Ned: Sure. NGINX for Azure is a managed instance of NGINX for all you lazy bastards out there.

00:44:02
Chris: So we found a way to make NGINX easier?

00:44:05
Ned: Yes.

00:44:05
Chris: Wow.

00:44:06
Ned: There’s also a Dynatrace on Azure, which is a managed service of dinosaurs. It’s pterodactyls.

00:44:13
Chris: That’s nice.

00:44:14
Ned: Yeah.

00:44:15
Chris: That’d be a good product name, which is probably why Azure didn’t use it. [screeching sound] That was horrifying.

00:44:20
Ned: Wasn’t it, though? Microsoft Dev Box was announced, which is basically VDI for developers. So, if you think about what a developer for a non-web application needs, they need more than what’s available in, like, GitHub code spaces.

00:44:35
Chris: Mm-hm.

00:44:35
Ned: They need a full-blown development desktop.

00:44:39
Chris: Oh, gotcha.

00:44:39
Ned: Maybe one that’s running Windows. And if that’s the case, Microsoft Dev Box—it’s basically an Azure VM, but it’s a managed image, and you can create those and then deploy them out to your developers, which means they can get up to speed faster, and also they’re not tied to a single, physical machine. They can just log into this machine and do their work.

00:44:58
Chris: You know who had a product exactly like that but then dropped it because they didn’t feel like it had a future?

00:45:03
Ned: Hm… VMWare?

00:45:05
Chris: VMWare.

00:45:05
Ned: Ohhh… Anyway. GitHub OIDC, OpenID Connect with Azure AD, is a new thing you can do. This allows you to create a service principle in Azure AD and associate it with a GitHub repository. It could be a branch. It could be a tag—whatever you want to associate it with. And now Azure AD implicitly trusts requests from that specific repository.

00:45:35
Chris: How is that different from what you could do before?

00:45:37
Ned: Before you had to put static credentials into GitHub Actions in the secrets associated with the repository or the environment.

00:45:44
Chris: Oh, okay…

00:45:44
Ned: Now you put in the client ID, but you don’t have to manage the password or secret. That will be managed between the OIDC connection of GitHub and Azure AD.

00:45:54
Chris: Now, is that set up for you? Or do you—is it working behind the scenes?

00:45:58
Ned: You have to set it up, so there’s still some leg work to do. But—

00:46:01
Chris: So it’s not a check box.

00:46:02
Ned: It’s not a check box yet. Maybe it will be in the future. But I think this is really cool for—if you’re using GitHub Actions as your primary deployment platform and you’re managing these static credentials, you don’t have to do that anymore.

00:46:15
Chris: Interesting.

00:46:16
Ned: Yeah. Related but not, part of the Build announcement is also that Windows subsystem for Linux, V2, is coming to Server 2022 in the next feature pack. So you’ll now be able to run WSL on a Server operating system for reasons that completely escape me because wouldn’t you just run Linux?

00:46:35
Chris: One would hope.

00:46:37
Ned: And yet, here it is.

00:46:38
Chris: Well, here it almost is.

00:46:40
Ned: Eventually.

00:46:42
Chris: Microsoft Office security vulnerability identified after Microsoft decided eh… it wasn’t a problem.

00:46:49
Ned: [laugh].

00:46:52
Chris: So TL;DR, Office apparently is stupid insecure. That’s not stupid and secure. It’s stupid insecure. Pronunciation matters.

00:47:00
Ned: This is news to nobody.

00:47:02
Chris: Nobody. So for some godforsaken reason, Microsoft Office products have the ability to run outside of executables.

00:47:09
Ned: Mm-hm.

00:47:11
Chris: This alone is staggering, but hey… they also have macros, so…

00:47:16
Ned: Yeah.

00:47:16
Chris: Anyway… this specific vulnerability exploits the Word remote template feature to allow discrete PowerShell commands to be executed on system using MSDT, which is some kind of support tool.

00:47:29
Ned: Mm-hm.

00:47:30
Chris: This bypasses the macro functionality, so it’s a completely new kind of insecure skullduggery. The best part about this is it was responsibly and ethically reported to Microsoft in April. Microsoft summarily dismissed it, saying it was quote, “not a security-related issue.” Basically, it’s a feature, not a bug.

00:47:52
Ned: [laugh].

00:47:53
Chris: And as an added bonus, as of May 29th, EDR still fails to detect this malicious behavior. Superb work all around. Microsoft has finally acknowledged it, and they are fixing it at various speeds across the multitude of existing office deployments, of which there remain many. There are mitigations you can enact right now, including de-registering the MSDT protocol or, you know, don’t—

00:48:21
Ned: Not using Microsoft Office?

00:48:23
Chris: —don’t use Microsoft Office.

00:48:23
Ned: Yeah. An AMD powered supercomputer breaks the exascale barrier for the first time. Some people said it couldn’t be done. Other more sensible people said, “Sir, this is a Wendy’s.” AMD said, “Hold my junior bacon cheeseburger. I’m going in.” This past holiday weekend, the AMD powered 128-node HPE Cray-connected, water-cooled Frontier supercomputer officially passed exascale, notching a sustained 1.102 exaflops per second. An exaflop, of course, is one quintillion point operations per second. That’s a “1” followed by 18 zeros. This is, as you are probably noticing, simply a head-scratching amount of processes, and the success places this system at the number one slot on the top 500 supercomputer list for this metric. To put it into perspective how bonkers this achievement is, Frontier is now faster than the next seven systems on the list combined.

00:49:34
Chris: And my favorite tidbit about this: AMD, while celebrating, announced that in 2023, their next supercomputer was going to break 2 exaflops.

00:49:43
Ned: Yeah, baby. Oh, I know this was a long one. So, hey… thanks for listening. I guess you found it worthwhile enough if you made it all the way to the end. So, congratulations to you, friend. You accomplished something today. And now you can search in vain for your sunglasses before belatedly realizing they were sitting on your head the whole time, hang that head in shame, and pop open a fresh bottle of Clearly Canadian. You’ve earned it. You can find me or Chris on Twitter at @Ned1313 and at @Hayner80, respectively, or follow the show at chaos_lever if that’s the kind of thing you’re into. [Show notes 00:50:18] are available at chaoslever.com if you like reading things, which you shouldn’t unless you like reading movie scripts.

00:50:25
Chris: And pie.

00:50:26
Ned: We’ll be back next week to see what fresh hell is upon us. Ta-ta for now.