Jake Warner + Zachary Smith

Jake Warner: [00:00:14] Hello everyone, and welcome to the ninth episode of Cycles podcast. We're here today with Zach Smith of Equinix Metal. Zack We've known each other for five, maybe six years now, and it's been well over.

Zachary Smith: [00:00:25] Half a decade. Well over half a decade.

Jake Warner: [00:00:27] Yes, definitely. So good, though, to finally have you on our podcast and happy to or really excited to dive into some of the initiatives that Equinix Metal is working on. And and just talk about the partnership between Cycle and Equinix and a number of things there. So I guess starting off, Zack would be awesome to hear a little bit about yourself and maybe an introduction to Equinix Metal and then we can go from there.

Zachary Smith: [00:00:54] Sure. Awesome. Well, thanks, Jake, and it's great to be on. I think we were if I was trying to, like, go back into my back in the day time frame, we were probably hanging out when Kubernetes was not obvious and there was a big debate going on. Was it going to be like, I don't even remember? What was the Docker version of an orchestrator at the time? Well, they had Swarm.

Jake Warner: [00:01:13] It was, yeah, it was.

Zachary Smith: [00:01:14] Going to be Was it going to be Swarm or Kubernetes? Like, I think that's like the time when we met.

Jake Warner: [00:01:19] Yeah, it was it was late 20. It was either late 2015 or early 2016. Yeah, But.

Zachary Smith: [00:01:26] So things were still messy. But luckily change is constant. So here we are. So anyways, my name is Zach. I a little bit about me. I live in New York City. I've been working in the Internet infrastructure, space or industry since 2001. I came to New York City from Orange County, California back in 1997 to go to school. I had dropped out of high school early. I couldn't stand being in school and decided I wanted to get on with my life. So I came to New York City and I went to Juilliard. I was a musician and I was going to be a concert bass player and playing all the big orchestras. In about three or four years into that, I decided, Nope, I didn't like being a starving musician and I didn't want to wait around for those types of things. Being impatient is a quality I've seemed to repeat and landed my way through a series of jobs into the early days of cloud computing partner with a a gentleman named Raj Dutt on creating what was one of the first virtualized compute clouds using Xen. And that business went well, kind of sold it eventually as the big hyperscale clouds were coming around in 2011 and vowed never to return to the world of Internet infrastructure. Two years later, my brother is sitting at a bar with me and says like, Zach, come on, let's let's go fix the hardware for for developers thing. And so we decided to get into highly automated hardware for a developer generation.

Zachary Smith: [00:02:55] We start a company called Packet, ran that for a few years, raise some capital, did some things that I was not very good at, made some mistakes, made a couple of right choices. And in March of 2020, Packet was acquired by Equinix. And so I now work for Equinix where I run the Equinix metal Division. And basically it's a business unit of sorts. And if you don't know about Equinix, we're the world's largest data center operator. We were founded in 1998 by Jay Adelson, who is later famous for founding Big, and he created Equinix as the neutral place for the Internet to kind of live and grow. And so Equinix actually stands for something like Equality in the Internet Exchange or something like that. And it was it was this time where it was kind of in the late nineties where it was not really clear where the Internet was going to live. Was it going to be proprietary within Telco's, right. Like the incumbent kind of nationalized carriers, or was it going to be these private Internets that were coming around like AOL or whatever that had their own content and their own things? And Jay really thought that there needed to be a place where these different sides of the Internet could meet in a neutral manner. And so Equinix is business was born, and we've kind of continued to prosecute that business since 1998 in terms of building ecosystems, that first one being the world of networks over.

Zachary Smith: [00:04:22] I think it's something like three or 4000 networks live at Equinix across the world. We then helped create other ecosystems like financial services or vast majority of the financial transactions in the world happen at Equinix and then later Clouds, where we helped Amazon create Direct Connect. I believe that was in 2010 or as a way for enterprises to have a neutral way to reach multiple clouds. And so you can kind of see the business is really been around creating ecosystems. To give you some quick stats, we operate about 240 data centers in 65 markets around the world, and that means we've got a whole lot of pins on the map, places, buildings. We're actually a real estate investment trust. So we operate as a real estate company that then gives you access to these interconnections, about 400,000 interconnections we have at our facilities. What I do at medal is I help make it easier. I remove friction from the process by allowing you to access kind of a hardware as a service, or we call it a digital infrastructure at software speed layer that allows you to reach all of the different locations throughout Equinix with physical, opinionated hardware being connected to all the right networks. And we do that with a nice programmable interface so that companies like Cycle can build awesome software on top of it and consume that as a base and a foundation for their end user services. So yeah, that's kind of what I do.

Jake Warner: [00:05:46] Yeah. No, it's really interesting to hear tell you the story change over time of just back again, back when we we started working together back when obviously it was packet and for those who might be watching this and are are relatively new to Cycle, Equinix Metal formally packet was our first partner as we were building out Cycle in the early days and back when we started we had taken more of a container as a service approach, right. Like where it was just, hey, you know, you upload your container and we were trying to be like the Digital Ocean for containers, but on top of bare metal infrastructure. And then as time went on, we had more and more companies reach out and say, Hey, this is great. Like, we like how simple. It's like, we'll make some of these things, but we won't own the infrastructure and so that's when we had to kind of tear everything down and rebuild it back up with, with Equinix where it was where.

Zachary Smith: [00:06:35] One of the options.

Jake Warner: [00:06:35] Yeah well well we're provided with supported but more so that our users were able to create accounts and directly own that infrastructure and then we'll just able to manage that infrastructure through those API keys. And that was the thing that changed our business because we had so many people starting to go through that process. But so I have a number of things that I want to dive into with some of the initiatives and things that you're working with with Equinix metal. But before those, I would like to just ask more of a another question related to moving from Packet to Equinix model and specifically, how has your day to day changed from from running a company where we're, you know, it's you and the board to now do a position where you're part of this giant, giant company and I imagine your day to day is very different now.

Zachary Smith: [00:07:22] A little bit different. Well, I mean, so, you know, luckily I was a little bit prepared for an acquisition being that I learned a lot. My first go around when we when Roger and I sold Voxel in 2011 to Internap, that was my first kind of round and being acquired and you know, I wasn't it wasn't a huge company. I think it was like five or 600 people in Internap was at the time and we were like 100 people. But it definitely was this like cultural shift that we had to ago. And I had a lot of just personal things I had to deal with. You know, you're building your own business, you're running your accountable for it. Maybe we were very scrappy. We were bootstrapped the whole time, so we were always out of money and all these other things, right? And then like having to change that, we're not. Everybody had that same gritty determination and love for the business that you did. They they wanted it to succeed. They were working hard. It wasn't about that. But like founders have a almost obscene amount of passion for their business.

Jake Warner: [00:08:21] It's our baby.

Zachary Smith: [00:08:21] Yeah, it's your baby. And then trying to, like, expect everybody else around you to also have that not necessarily the same. And so with with Equinix, we had built Packet, I think we had focused on my brother and I and Aaron, who was our early co-founder. We had focused on building not only a great product but a better company. And so one of the things we really focused on early on was culture. So we had a strong remote first culture, really open and transparent, really allowing kind of like the broader amount of people within the company to participate in the growth of our business, the positive and the negative. Right? In my first business, I was kind of like I shielded from all the negative and kind of like took that all on myself. You know, Raj and I spent a lot of late nights trying to figure out how we're going to make payroll in, in, in packet. We were very transparent in that regard. So really open with that and allowing people to be part of it. Of course, we couldn't do that when we acquired by a public company. We had to be a little bit closed doors, but certainly trying to be transparent as we built the business and that created a strong amount of like I'm going to call it accountability throughout the organization. And then we also practice this thing called EOS, Entrepreneur Operating System, which was a framework we used to run the company who what people are there in terms of do they align to the mission? You know, do they have the same value systems as you do? Do they have the right skills and the right time of their life to to run that that job? And then how we actually ran the company by great metrics, by having repeatable great ways to hold meetings and solve issues.

Zachary Smith: [00:09:51] And so that was like very, very part like a big part of of, of packet, right. But, but what I would say was I'm not, I'm not going to say that this is unusual in startups. A lot of the things that you have and we had gotten to like 150 people or something, so you passed the stage where you could remember everybody's name, you know, in the hallway, like you could still do it, but it wasn't the same, right? And we were the virtual. Holly, We're always kind of a remote first business, but we had gotten bigger. But I would say that one of the things that you you have in startups is you have a much clearer sense of the of the mission, because you're usually not like, hey, we're the number one company in the world At X, you're like, Well, what we're trying to become is this. So there's this big focus on the future and the vision, right? And I had a board member or an investor selling to me one time in early 2014. We started Packet.

Zachary Smith: [00:10:44] It was 2015 . I was probably complaining a little bit to him about how hard it was, and he said, Zach, you're like, You're in the best spot ever. I'm like, What are you talking about? I don't have any customers. Nobody knows who we are. Like, we're like, we're competing in this market for like, mindshare and we don't have money. And he's like, No, no, no, Zach, you have the best advantage. It's like you don't have any customers, which means you don't have any past. All you can do is spend all your time thinking about the future. And so you're in creation mode all the time. And that is a extremely unique skill and a precious gift that you need to use because eventually you'll have a bunch of customers and you'll have a whole lot of things and you'll have something to report that is so like what happened last quarter, and then you'll spend most of your time thinking about the now in the past and not enough time thinking about the future. And so I would say that culturally we had strong decision making, very, very rapid decision making at Packet and also a very big focus on the next thing, always the future. And that is probably changed the most for me at Equinix. First off, we have nearly 13,000 employees, so we're much bigger than 150 people that we had at Packet. And so my role as a senior leader is very different leading through the pandemic, leading through change of integrating companies and aligning culture.

Zachary Smith: [00:12:10] That is a big new challenge for me. And then I would say those other two things are really the biggest differences, which is I spend a lot more of my time focused on what's happening now and what just happened. We are a public company. We report a lot of information to our investors, like there's a lot of that, right, Which wasn't a big part of what I was doing in a startup. And and then I would say that other thing is, is building consensus, right? We're in a company where there's a lot of cross functional requirements in order to activate a larger business. And 13,000 of your colleagues, you need to gain their alignment, which is not as easy as saying, I've decided we should go this way, which is kind of what I had a little bit of the ability to do within a startup. I was like, Well, I've decided or it's been decided that this is the way it's going to be. And you know, like you could do that. You could turn really quick if you wanted to. In a big company, it's like an aircraft carrier is like very powerful but doesn't move very quick. And so getting everybody to work together and align on that thing is the other thing I spent a lot of time doing. And those two those two would be the biggest differences for me.

Jake Warner: [00:13:17] Well, it's really interesting to hear that. Obviously, you're you know, you're much further ahead than than than we are. But it's very interesting because, like, you're the you're the first founder of a non customer that we've had on the podcast yet. And so it's really interesting in the first founder of a provider that we've integrated with. And so it's really interesting to hear just like a founder to founder some of the stories because it's like you spend so much time as you're going through these processes and building these companies where as many times it starts to feel lonely and it's like, all right, like, hey, I'm doing this and it's getting results. But is it am I doing it fast enough or am I doing it right? Or you know, there's just so much time that you kind of spend lost in your head and trying to figure out things that whenever you hear other founders talk about the hardships they went through as well, you're like, All right, this is not just me. This is this. It's a hard thing to do.

Zachary Smith: [00:14:12] And it is very hard and it's hard for the people around you, right? If you have family or whatnot. I think it's a very, very difficult ask. It's not without it's it's not free. And I think also personally, I've I've really benefited from being really open about the challenges as a leader that I have. Right. That's one of the things that Equinix has done really well, is to focus in on personal development and mental health, which has been incredibly important during this pandemic. And I've leaned into that actually before. The first thing that I did when I knew that we were going to sell Packet was I got myself a therapist, right, so that I could talk through all the emotional issues I was going to do with getting rid of my baby and becoming a bigger part of a company and rebranding the business and doing all the other things I knew I would have to personally unpack and at least like recognizing at a founder that there's a lot of stuff in there, right? And so as you go through, whether you grow your business, whether you take on investors, you go through ups and downs, you sell the business, all these types of things. Stuff that you as founder kind of have to deal with, even if you choose maybe to just move from founder CEO to just like founder and not CEO.

Zachary Smith: [00:15:22] These are areas I think that it is good to talk about with somebody. I'm lucky enough to have an identical twin brother who I work with. So we have a very close relationship and can kind of talk through. But even I went to access some other help that I could talk through these issues with because they are unique. And frankly, when when when I became part of the leadership team at Equinix, it was something that I felt very lonely on because most people were not coming from founder led journeys. They didn't have the same kind of emotional connection to things that I maybe did not because they didn't care about their work. It was just a different situation. And what I, what I needed to do is learn how to react to that differently and unpack those things on my own, because I couldn't expect everybody to know what how I felt or my reaction. So that was an important part of dealing with the the founder led, you know, issues that I had to work through. I still do.

Jake Warner: [00:16:21] Yeah. I mean, I appreciate that, though, because like was one of the things that like I mean, I haven't, you know, I haven't worked with a therapist, but one of the things that we've done within our company is we have probably our board meetings are probably more frequent than most startups. We our board meetings are the first Monday of every month. And but our board meetings, we put a lot of energy into making sure that they're not formal. Like like, yes, we'll spend like the first 5 minutes talking about metrics and things like that, and then we'll take the the the next 45 minutes or so and just have a discussion about what's happening and things like that. And it's really nice because on on our board we have we have, I mean former operators, right? So like current and former operators. And so it's really nice because like these are people who have been there kind of I don't want to say that because everything's I mean, every company is different, right? But they understand the stresses that founders go through. And so it's nice to be able to have advice of people who obviously, you know, are involved and want to help, but.

Zachary Smith: [00:17:20] Got to have a safe space to talk about those things. So that's your outlet.

Jake Warner: [00:17:24] It is. And like I've realized, like a lot of people have said, like, hey, we've had many people that I've talked with over time that have been like, I could never be in a company that has monthly board meetings. And I'm like, Yeah, but this is this is my this is my way of like, like finding mental peace, right? Is if I, if I can if I can have open and honest conversations with these people, it helps me keep keep stable.

Zachary Smith: [00:17:46] But that's great. That's I'm glad that you have that with your board. That's really awesome.

Jake Warner: [00:17:50] It's real nice. But so, you know, I guess diving into some of the I have written all these questions that I wanted to dive in to for some of these things that Equinix metal, I mean, I've loved the conversation because it's a conversation that I haven't been able to have with anyone on the podcast and I think is an important conversation to have. But shifting to Equinix metal, so recently I saw the I know that you had put out a a blog post in early October about it was about cooling in some of the some of the sustainability efforts the Equinix is currently working on. And so I'd like to to learn more about those. But specifically before I guess I dive into the other questions is in the article you had mentioned a number of times the Open19 project, and for anyone who's watching this and is not familiar with the Open19 project, would you be able to give a high level briefing on on what that is?

Zachary Smith: [00:18:43] Sure, yeah. So Open19 is a project within the Linux Foundation where we work on a standard for kind of the mechanical aspect of putting computers in racks. So what we work on is what's called a brick with that we remove the power supplies from a server and we basically use a highly efficient power shelf with blind make connectors on the back of these kind of what we call brick cages. They're basically just sheet metal with the right cutouts where we can plug power and network with blind mate connectors. And now what we're working on is a version two that increases the power fairly significantly, much more efficient. 48 volt power does some things on the network and adds in a pluggable liquid cooling standard. And what liquid cooling is, is like one of our biggest environmental challenges within data centers beyond we need a lot of power is that we create a ton of heat. And the current way, especially in a retail data center versus, say, a hyperscale data center, a hyperscale data center, you can kind of be very opinionated about the the hardware. Like you might have thousands of machines that look the same. So you'll design your cooling systems and everything very, very efficiently around that thing. In a retail data center like Equinix, we got customers doing all sorts of things. One rack next to the other could be two kilowatts on one and 20 kilowatts on the next one. And we can't dictate.

Zachary Smith: [00:20:04] Here's what your hardware looks like. You have to buy this thing to put it in our colo. And so we have challenges related to cooling. Definitely we have power issues, which, you know, certainly worthy of a conversation. But our biggest issue is actually cooling or maybe set a better way. Removal of heat. And that's because chips are getting hotter as we move to things like machine learning, accelerators, ddr5 faster networking, new processors going to seven nanometer, all these things create so much heat. If you've ever sat with your MacBook on your lap for a while, you will understand exactly what I mean. But try that with dual sockets of 400 watt things, with all the fans and all the DDR things and all the memory and all the accelerators and suddenly you're pulling multiple kilowatts per box. And so what we've been working on is a standard within Open19 to allow us to use liquid like a closed loop system to run it over the processors. Right. Or the high, high heat generating components and capture that heat and remove it in a more efficient way. What's really neat about it is not only can we dramatically lower what it cost to air condition the things and put cold air all over the place, we can we can basically remove waterways because we spend millions and millions of gallons of water that we have to evaporate today and kind of standard air conditioners.

Zachary Smith: [00:21:29] And we can make that a closed loop system where we actually don't waste any water. It's really important in certain parts of the world where we just we're we're really challenged by our water use. And then would think the other way is that we can actually take that heat at such a high temperature that we can use it to create power or actually sell it back to the grid, or municipalities that need heat like to power a turbine or to heat buildings or related. So there's some really interesting opportunities. We've even looked at how to do that with vertical farming right next to it because they need heat. So there's lots of interesting ways that we can take technology, which is currently, I'm going to say, very energy intensive, very taxing on our environment. And our goal at Equinix is to be net zero by 2030. And that doesn't mean just buying offsets for our power that we use. That means actually net zero from a carbon standpoint using science targets. In order to do that, we've got to innovate with the entire ecosystem and that's why we've chosen to do that with an Open19, because it's an open standard that OEMs like Cisco participate in and power partners like Schneider Electric, because we can't do it on our own. We need to work with a collaborative ecosystem to build a standard so we can roll out these technologies much, much faster and more ubiquitously.

Jake Warner: [00:22:45] And I mean, all of that is really fascinating. Like it's sustainability, something that's always been very important to me. So it's nice to know that one of the companies that we've partnered with is taking that story and sorry, it's taking that that that initiative and that that goal and not, as you said, just buying offsets. Right. But actually working towards becoming truly net zero. And in that same article that I was reading that you had written, you had talked about taking the extra heat and actually powering residences with it over in Helsinki. And how did that come to be? Is that something you can spend?

Zachary Smith: [00:23:20] I don't know the specific of that project. I mean, what's fascinating is it's happening all over the world right now, right? Like suddenly what went from being two years ago and I've been involved in Open19 for a while, but what happened two years ago used to be pretty siloed, like, oh, we're trying to do this because it's better. And now it's like an imperative. It is everywhere. It's being through regulation, through social constructs, through our customers. I can't say a single one of our big customers like our Fortune 500 in every RFP. Now sustainability is up at the top, is not somewhere in the bottom of the RFP. And so we're seeing just this big movement all over to find and creative. New ideas because we're going to have to use all the ideas. By the way, there is no magic bullet. And I and I think that beyond how do we make our operating of technology, clouds, data centers, etc., more efficient, there's certainly like one thing we were discussing recently with Intel and a few other of our partners was that even you like Cycle could really help us in managing our energy use. For example, when we have renewable access, we want to be able to power more intensive jobs, but when we don't have renewable access, we would like to express it to you somehow, to let you know so you could do something about it. Right now, you have no idea based upon the power things that we have in the grid or our own generation capabilities, whether it's a good time or a bad time to run this intense machine learning job or re index all your search queries.

Zachary Smith: [00:24:49] Right? But like maybe we could come up not with a bespoke integration for every single thing, but like that's a place where I think we could do really well within the industry to have unique ways of telling you that. And you could build those into workload managers like Cycle or whatnot to help your customers use the existing resources at the right time or just more efficiently. Obviously you're helping them use it more efficiently today. But we could also help make it so that you could use it at the right time. And then I think one of the other things that we could do long term is is changing the business model. I really like the movement that we've had over the past 15 years towards cloud where basically the operator gets paid when you use the things. So that's good because it aligns our incentives. The challenge is that there's a whole part of our industry, including OEMs like server manufacturers or silicon manufacturers, that don't get paid when you use it, they get paid only when you buy it. And when you look at the actual carbon impact of a computer and it varies, but like let's just use round math, 60 to 70% of the energy use. Over a computer's lifetime is making the computer. And so only about 30% is like using it and maybe recycling it at the end. Mm hmm. And so, so much of it is in supply chain process technology, building the chips, sheet metal, mineral extraction, all these kinds of things that go into building the technology.

Zachary Smith: [00:26:17] Well, what if we could make it so that the industry was incentivized to use the technology longer? Versus throwing it out after the current Cycles, which is usually 3 to 4 years, because people always want the latest things and we're incentivized due to the economic model to do that. And so I think there are some business model things we can do. There are some technology things like giving you information and allowing that to proliferate through workload managers and standards and and also just measurement. Right now it's really, really hard for like we don't provide, which is like an initiative we have internally. You don't know how much energy your infrastructure with Equinix or your customers infrastructure at Equinix actually takes. You have no idea what's the carbon impact of doing this. And so it's very hard to say, Hey, Jake, Cycle, you guys should do better at that when you don't even know what it is. And we've kind of like, what is the phrase like sufficiently advanced technology is no different than magic or whatever that phrase is, which is super awesome. Like we've made it magical enough that you don't know actually how much energy is being used to do X or Y, or you can turn on that server. And that's a problem that we need to solve to give measurement and availability and transparency throughout our industry.

Jake Warner: [00:27:28] Do you think that as time goes on and more and more manufacturers and providers are building like application specific hardware that will help extend the lifetime of that infrastructure?

Zachary Smith: [00:27:43] Yeah, I think that the big opportunity right now is we kind of not to use my warship. I'm not into actually battleships, but not to use my battleship analogy, but a typical processor basically has everything built into it. But you really only need very little of it. And the challenge is right now is like super hard. If you wanted to optimize software against one specific function, you've got to buy the whole big processor just just to do it right. In reality, accelerators and purpose built technology I think represents a massive opportunity not to make like 20% better, but like thousands of times better, you know, more efficiency on doing certain things. And I think you can see like what certain technology companies like Apple have done related to their M1 processor. Right. Hey, we we realize that everybody is using JavaScript all the time. Well, let's build JavaScript engine into the hardware so that like 70% of what people do on their web browsers or whatever goes better and like, how could we create a way where that actually is a reality? Because right now if you whose like in the software supply chain somewhere wanted to optimize against technology, it would be so difficult for you to actually take that technology and get it in a market that you wouldn't really be able to do it on the time scales that made sense for you. And so one of the other goals that we have at Equinix and my job within metal is not to be like a provider of bare metal cloud, but a distributor and operator of technology. So that technology is in the right place at the right time that it can be used by the right customer. Now that's, in my opinion, more of a logistics problem than anything else.

Zachary Smith: [00:29:18] And today we basically build technology and this is what's so cool about Open19 is today. If you've ever been into a data center and seen all the racks get made, you know, hundreds of cables all put in in very special ways and zip tied and made all special, we have done everything that's all about getting technology into the data center. But we never have done anything to allow you to move that technology to another other place. And the packaging, like we throw it all out, it's not reusable. So then how are you going to get it from like Helsinki to London? Like, you might actually need to use this specialty set of cards that you have and you need it in a different place, or maybe for a different customer. And right now, those things are all built just a one way street. And so if we can make it like they call it, like a circular economy or a circular model, well then we design it. So it was reusable. We designed it so that you could take that thing easily with low amount of friction to put it somewhere else or even to a secondary market. So I do believe that within the next couple of years, like the next phase of workload is going to present an opportunity for developers who have already figured out how to make their workload portable. That's great. Well, now how do we get it on the right thing? Now that's the next big way that we can save our planet, our energy usage, and also meet the needs of our customers, which say, I want more access to greater technology.

Jake Warner: [00:30:38] Yeah. And I mean, I agree with everything that you're saying. Like, it just makes so much sense. And it's neat to see, you know, things like this conversation are happening already with, as you were talking about with Apple and their M1. Well, first, have you have you had have you actually had an opportunity to test one of the new M1 Pro or M1 Macs?

Zachary Smith: [00:31:00] I didn't try one of them yet. I just got one to go buy.

Jake Warner: [00:31:03] I got one last Thursday. Incredible. Incredible. Amazing building. Building containers on it. I've noticed like a 80% decrease in compile time. If if it's an arm container. Right. If you if you're building an intel container on it, it actually it takes more time. But that would make sense, right? There's emulation built in. But the point that I was getting at there was specifically of having those those efficiency cores built into the processor as well. It's just really neat. Like I'm I'm one of those people. Like I'm I'm a data nerd in terms of like watching like how all my different machines are doing. And so I run I stats if you've ever run I stats on your computer. Oh, yeah. And so we track the individual cores and things and it's really interesting to watch. Like, like if I'm sitting there doing a compile, I'll see all the cores doing their thing. But if I'm just, if I, you know, once it's done and I'm just sitting back and not doing anything, you see everything shift over to those energy efficient cores. And it's just so neat to actually watch little And, you know, I mean, as you were talking, I forget the exact quote that you mentioned earlier, but the magic of it all, the fact that this computer is automatically making the decision that, hey, I don't need this heavy compute, I'm going to switch over to these cores that use so much less energy. And that's something that, you know, a user does not have to make that decision. They don't have to build a specific application to say, hey, you know, when you're done with this, move yourself over to an energy efficient core. The the the operating system or.

Zachary Smith: [00:32:28] The kernel or whatnot.

Jake Warner: [00:32:29] Knows how to do that.

Zachary Smith: [00:32:30] Yeah. Well, I think that this lies one of the bigger opportunities we have is to unleash that what Apple is able to do in its verticalized closed ecosystem. Right. Is to be able to do more of these things in open ecosystems. And I think that's why that's why I spent a lot of time in Linux Foundation and related. But I'm also super excited about Risk five and different kinds of I'm going to call it open ideas that allow this to happen. And I think we have a duty to kind of connect the rest of the supply chain and make it available. But I don't think it's unreasonable that in five or six years you could kind of order a specific type of processor or thingamajig or that the thing that you want. And we were able to put it in market really cheaply all around the world and run it for you, and that could make your application or run time or security regime or whatnot, orders of magnitude more efficient. I think that's going to have to be a reality. We have to consumerize the supply chain of getting technology in the right place and then reusing it for the right purposes or recycling it and having a circular side. Along with that. We're not there. It is really far away from it, but I think it's totally within the realm of possibility.

Jake Warner: [00:33:40] No, it makes so much sense and.

Zachary Smith: [00:33:43] We're going to get you your M1 pro on demand available everywhere. Just let me talk to the guys over at Apple.

Jake Warner: [00:33:50] Well, I mean, so I mean, it's a really relevant conversation right now because before with Cycle, today's Cycle only runs on Intel. And I think you know, why, you know, is there really ever a time to shift away from that or to rethink that? And and then with the M1 with how well things have been running on the M1 and then the efficiency cores and things like that, like knowing that, hey, I can use less power if I, if I don't need to, you know, need performance. It started to make me realize, like, you know, there's a lot of improvement or there's a lot of work that needs to happen on Cycle to start supporting other architectures and being able to support ARM and things like that is something that I think could be really powerful for this for exactly what you're saying of being able to have. You know, application specific infrastructure I think would be very powerful. And so one of the things I've been working on recently is a because like, like Cycle OS, it has like a static kernel built into it, right? And so I've been working on a way to have kernel modules downloaded after the server starts up. And so that way when the agent checks in and says, Hey, here's who I am, here's the infrastructure that I'm running on, and then we can push kernel modules back to it that way.

Zachary Smith: [00:35:00] The initial download.

Jake Warner: [00:35:02] Is still very, very small, but the thought is that as time goes on and we have more specialized hardware in these servers, we can then push those those kernel modules down to those machines and.

Zachary Smith: [00:35:13] Take advantage.

Jake Warner: [00:35:14] Of it. Right? Exactly. And so that's been one of the I mean, as we go back to our founder conversation earlier, that's one of the big transitions that I've been working on or going very quickly is I used to be like spending 90% of my time writing code and now it's like 20% of my time. And so, you know, it's weird. Write code was what I've done for a long time. What you did. Yeah. And so now that it's that 20% the things are a bit.

Zachary Smith: [00:35:40] Weird and now you just do podcasts all the time.

Jake Warner: [00:35:43] Yeah. I mean, I mean I enjoy it though, right? Because like, you know, there's just there's a lot of good information out there that needs to be told louder, right? Like, more people need to be aware of some of these things that are happening and and but yeah it's it's it's been it's been a big transition into less of Jake as a developer and Jake as ita CEO and I mean I enjoy most of it some day it's not but that's part of the I guess growing.

Zachary Smith: [00:36:09] That comes with it.

Jake Warner: [00:36:11] But so let's see I, I guess so out of the question that I had asked, I think we've so far we've we've gotten through almost all of them kind of just automatic. Right. But the one that I hadn't I mean, I guess we've kind of touched on it, but maybe we can dive in a little bit more is so as we've talked about building ecosystems and communities, which is obviously building these ecosystems and communities and things like that, and focusing on sustainability is, I think, important to both of our companies. But is there any ways that you're enabling your customers to directly get involved with some of these sustainability efforts? I mean, I know that you talked about like, you know, making it happen kind of with a magical component. And I guess you already did talk about if we can make that point where you can tell your customers or when it makes sense to run more intensive processes depending on workloads and things like that. But is there any other initiatives that I guess we haven't talked about yet that are specifically focused on empower or giving your customers direct access to some of those sustainability efforts?

Zachary Smith: [00:37:22] Yeah, I mean, it's a great question. You know, it's always hard to find like the balance in it when you're in physical infrastructure. It's like where can you engage? It's a little bit more difficult to build communities that may be in software, right? Or open source can be a relatively I'm going to call it egalitarian, you know, prosecution because, hey, what do you need to be able to do to play around with this code? Pretty much just a laptop. And if we're talking about, hey, what do you need to do to play around with this liquid cooling capability? Well, probably an entire mechanical team and whatever. And so it's not exactly as easy or straightforward, but I think our participation in Linux Foundation has been really great. And then working with our supply chain and basically moving them, we've been very, very public about our science targets that will get us to to our 2030 goal of net carbon. And we're expecting our partners to come with us in an open way to solve these challenges, not just for us but for all of this industry. And so that's been like one big galvanizing thing is using the the umph of a $70 Billion market cap company and related to say like this is important to us, we're stating publicly that it's key to our our corporate sustainability model and our business and we're expecting our supply chain and partners to come with us.

Zachary Smith: [00:38:38] Funny enough, a lot of our customers want to come with us, too, on that, right? Because they they operate data centers or clouds or they're doing the same things within their own big businesses. So I think that's been really nice just to start the conversation be be a strong part of it. And we work with a lot of the the government and nonprofit organizations around that, the European Union and sustainability. The Singaporean government has an initiative. We're working with the City of Dallas on this, so we're engaged throughout, I'm going to call it the governance model as well of public policy related to it . Software area. We've donated our provisioning capability to the CNCF. So it's a project called Tinker Bell. That's an area where we hope to allow hardware manufacturers and software people to meet for things like, Hey, how much power is this thing using or is there a how do I turn on using my Kubernetes operator the low power mode so that way it can appropriately or even measure how much power is coming out of the processor. So things like that I've been working on. Areas to participate in there.

Zachary Smith: [00:39:37] And then, frankly, the last thing is just an area of passion and building ecosystem's around the similar topics. I don't think, again, like this is the area that I believe is an opportunity of our time. Like we need so many layers of the stack to be able to come in in an open way. But I think there's really an opportunity to kind of organize community based efforts. And one of the things that we operate at at Equinix, these things called ECN, their connection networks, where people have particular topics around social responsibility, around I'm going to call it things like gender equality, know different, different kind of causes that we have within our employee base. And one of the most popular ones around is sustainability. And people really want to have an area to put their effort into. And so I don't have the answer like, but I do think that there's more for us to do in creating a space for people who have particular skill sets in the Linux kernel or in workload management or in Physical hardware or mechanical systems to be able to come together and solve some of these super complex things in an open way. So you've you've, you've continued to inspire me to work on that. But I can't say I have the answer.

Jake Warner: [00:40:48] No, but I appreciate that. And and again, the nice thing is that we're a number of years into our partnership and there's still still, still a lot more to happen.

Zachary Smith: [00:40:57] Maybe what we can do is we can make it so that way within the Cycle interface, we can somehow show how much carbon was used for any particular part, and we could also do things like offset that or we could allow people. I've been very inspired by what Stripe's done with Stripe climate, where they have actually invested in carbon capture technology and they can just turn on a button that says take 1% of my revenue and put it towards carbon capture. And I've been thinking about that within cloud computing, like could we build sustainability into the actual consumption model so that you could see not only how much energy was being used or carbon was being offset or something like that, but also you could do something about it. For example, whether it was use a lower carbon or lower energy data source at the time and move your workload, or you could actually invest in changing that and we could be a platform or together we could showcase that within the Cycle interface. So let's, let's, let's talk about not how I run workload, let's run workload better.

Jake Warner: [00:42:01] No, I'm up for it. And I mean, that's the neat thing is that today there's a few hundred servers being managed by Cycle. And if that's something that we can implement and we can do, well, then it's not just one small workload that we're changing. It's it's not only the hundreds of servers that are connected and managed by Cycle today, but as time goes on, you know, it's more and more what's the word I'm looking for? Like just I don't want to say exponential.

Zachary Smith: [00:42:26] Your network effect. It's your network.

Jake Warner: [00:42:28] Effect. There you go.

Zachary Smith: [00:42:30] I got it.

Jake Warner: [00:42:31] Well, yeah. Thank you so much, Zach, for for, for taking the time to answer the questions and being on the podcast. It's been great. I've really enjoyed our conversation and it's been nice to also have the founder, the founder component as we've been able to dive into those things. And so I've really appreciated that.

Zachary Smith: [00:42:46] Well, thanks, Jake, man. It's great that you're doing it and awesome work with Cycle. Thanks for being a partner in a platform ecosystem partner of of Equinix. So it's it's really cool to see what you guys have built and what you're doing for customers. And hopefully we'll continue to work on that and make it even better.

Jake Warner: [00:43:05] Excellent. Excellent. Thank you so much.

Zachary Smith: [00:43:07] All right, Jake.