Greg Irwin 0:18

Everyone, good afternoon. Thank you all for taking some time. Thank you all for turning on cameras. This is amazing. Yeah, my name is Greg Irwin. I'm one of the partners at BWG, I run discussion groups for a living. The idea and format here is super simple. We're going to share some stories good, bad, and hopefully not too many ugly about what people are doing with infrastructure. And we're partnered here with Equinix. And with Pure Storage, and the basis for the conversation is Bare Metal about Bare Metal offers. So we're going to start in the beginning of the call, and basically have Zac Smith and Jack Hogan talk about the offers, talk about the use cases, tell a story or two. I'm thinking guys, you know, that's 1010, maybe 15 minutes, maybe what I want to do is basically pivot and talk about how other people are basically, again, collect the stories, how other people are architecting, those hybrid architectures of marrying a cloud tier versus a fixed managed or on prem tier, and see how that works. In reality, maybe there's some real challenges managing it, maybe it's incredibly valuable in terms of bursting workloads, maybe bursting workloads don't work so well, because of egress charges. There's lots of stuff that I know, everybody has to deal with. I do have one goal or one initiative I'd like to encourage here, which is, if you look across this group, it's pretty fantastic. The professionals that are joining here, so I want to encourage everyone to take advantage. And please aspire to make one new contact across this group doesn't have to be Pure doesn't have to be Equinix or BWG, but reach out or let us help reach out on your behalf and expand your personal network and meet a new, a new peer who sits and does similar things to you. So please take that initiative you're putting in the time, it's probably the best value that we can offer is expanding your your expert network. Okay, speaking of experts, let's let's go over to Jack first, Jack, this is fun. This is probably our I don't know 10th time in a group. Take a moment and please introduce yourself and Pure. Yeah, sure.

Jack Hogan 2:46

So I definitely like to introduce myself by starting off with a question which is what is Pure’s customers trying to solve what business challenges are they trying to achieve? And that's actually my role at Pure. I've been a career CTO on the customer side, I've been here now for four. And my role is really to synthesize and work with the likes of a steam group like this and really understand the big challenges of today. And you know, Pure Being a data storage company and and really revolutionising over the last decade, what it means to access and make use of data that sits in the form of all flash storage. We're not taking that to all new levels, because the business challenges have shifted, the desire to achieve outcomes, is now greater than ever, the pressure from boards the pressure from just business reaction time to be flexible, and have flexible infrastructure to react, react and respond to new opportunities, new threats, new challenges. And I think the course of the last call it 18 months with the impacts of the global pandemic, I think have really highlighted a lot of the big pain points, that those of us that have been career technologist and in career operator roles are really starting to grapple with. And that's what we're here to talk about here today. That's what Zac and the folks at Equinix have been doing with Pure now for a couple of years in their former incarnation. But this concept of being able to provide flexible infrastructure at the level that our teams are able to use that, you know, the big skills gap that exists of making a jump straight into hyper scalar clouds is never more pronounced those. Those individuals are hard to find. They're expensive, they're hard to keep. And platforms like this are addressing those problems that we really have all got to come to grips with, which is those things that I like to find is the past, present and future challenges that any technology leader is faced with those past challenges being a lot of the legacy platform I can't tell you how many CIOs I talked to large scale Fortune 50 type CIOs that say, we're 50% along our journey to the cloud, and you ask them about the other 50%. And they say, We have no idea how we're going to solve that. Every Greenfield new application is being developed in, in new cloud-native and Kubernetes, and container work spaces, but cost management and challenges that exist in that kind of future state of Greenfield and then that that data explosion that's happening right in the middle, that presents a challenge. And that's the that's the solution that Pure and Equinix have come together to solve is really that continuum, and allowing folks to really reimagine what it means to enter a cloud model, and achieve the benefits of of cloud operating capabilities. So I really have experienced it, if I was going to say and characterize what we are doing with Equinix. This is the Goldilocks solution that I sought for 20 years of my career being in in many in many clouds in many data centers. I'm having dealt with many multiple m&a activities. This is that solution. And so we're excited to really have a great group here to chat about.

Greg Irwin 6:14

Hey, Jack, thank you. This is going to be fun. Zac, step up to the mic here and give give an intro to the to the Hollywood Squares.

Zachary Smith 6:24

Well, Jack and Team No, I'm pretty shy, so careful. How am I get stage fright here. So I'm Zac Smith, I'm a managing director of Equinix Metal. I've been here at Equinix. For about a year and a half came along with the acquisition of a company I founded called packet where our mission has been and continues to be around automating fundamental infrastructure primitives, or as I like to say moving physical infrastructure at software speed. And what we do here at Equinix, if you don't know is we are the world's largest data center operator with a rich history in creating ecosystems. And we do that by having 240 buildings all the way around the world. 65 markets just closed on India, Canada, if any of you guys are there, Mexico, building in pretty much I think we have 40 construction sites underway right now. So huge physical reach platform of helping our customers kind of go everywhere. And then we we pair that with with our interconnection capabilities, both physical interconnection with cross connects our i x products and fabric, which allows you to get to the largest number of cloud on ramps, and our other 10,000, enterprise and service provider customers. Extremely fully programmatic, it also connects all of those sites. So as people modernize their network, and move from one location to another, or from a siloed, maybe enterprise data center into an ecosystem-based approach for a hybrid multi-cloud, the connectivity options really have a huge impact on TCO and your overall, let us call it performance and spend profile that you have with them with your networking and your data. And so what I'm excited about what I'm working on, I wrote it down, it was kind of inspired by Jack. So I was like, I better take notes. What we're here to do what Equinix Metal is to remove friction from our customers who are trying to access this huge global reach, they're being put in all these different ways, which is both move more places, because the world is flat, your users are all over the place. Your customers or employees are demanding awesome, you know, kind of global performance and security and localization while you're at it, and data privacy. And you're also being asked to do it faster, which is Hey, make it more automated, do it with less, figure out how you can align your costs with the outcomes of the business. And so what I'm excited about is beyond providing less friction, removing that friction from our customers accessing that global reach and being able to have their infrastructure opinion wherever they need it. We also think we can bring a significant paradigm shift to what I call automating layer zero, taking that data center and hardware and turning it into automated data center automated hardware capabilities. So you can start to practice repeatability and DevOps at the lowest layers of your IT stack and turn that into highly repeatable and automated primitives. We think that sustainability is one of the biggest issues facing our, our company and our customers today. So we have a target of net-zero carbon emissions by 2030, we already 100% green power across the board. That's not enough. In order to do that we see all of our major customers investing in sustainability. And we think we can be a huge portion of that. One of it is by working with partners like here to make sure that the consumption of our technology is aligned with your outcomes. Instead of just trying to sell you things like more storage or more computers, we want to make sure that we get the benefit of being your partner when customers get output or use or value from those. So aligning the outcomes we think is part of the movement as a service is really around sustainability. And then of course, the investments we're making across our energy footprint and kind of carbon footprint all over. And then the last thing that we're interested in is really making sure that people have access to the right technology in the right place at the right time. And that's becoming great pretty hard. We operate data centers from Osaka, to Singapore, to Buenos Aires, to Stockholm to Atlanta, and is hard to put the things these physical things that I know you all do, you build and run and operate this kit. And you know our job, whether it's through, you know us directly or through one of our partner ecosystems like Pure, we want to make sure that the right thing is available for you to use best of breed technology in it as a service model in a neutral way. So that's what I'm excited about. Thanks so much, Greg, for giving us a chance to intro

Greg Irwin 10:31

perfect. So let's kick it off with some questions. And I'm gonna ask everybody, you know, let's, let's get busy on your keyboards, please, drop one question into the chat that you have to understand the offer. And I want to, I'm going to ask Zac, first. Sure. Zac, if, if Equinix? is offering Bare Metal, how is that different from us? I mean, are you basically now and is provider in that sense? What how should somebody think about this different from a service that they would buy from Amazon? Sure.

Zachary Smith 11:05

Absolutely. So I think of in a few different ways. I've been in the public cloud journey for a long time started when the first virtualized clouds in 2004, based upon Zen, so it's come a long way since then. But I think we have a new paradigm that's emerged, which is, over the past five or six years, a huge amount of innovation has occurred at the software layer, where whether it's platforms like VMware, or Pure Storage offerings, software, defined storage capabilities across all sorts of environments, or things like workload portability, and Kubernetes. Software is now extremely portable, including the IS layer. And so for me, what we're doing here at Equinix, is offering two fundamental things. Number one, data centers are service, you don't have to build the world's best data centers, with all the right people and all the right places to have access to those. So we do that for you. And the second thing we're doing is delivering hardware as a service. I call this automated colocation, we're bringing you the right things, the storage, the storage, and server appliances in the right way, connecting them to networks at a fundamental layer, so that you can put the right software on top of them. And so what we're doing here is giving you fundamental operated, physical infrastructure, eg colo, as well as the Compute and Storage primitives, and bringing a partner on top with their software capabilities, like Pure best of breed software, to give you that foundational storage, technology and automation, we think the same thing with for example, I'm sure many of you are users of VMware, or maybe OpenShift, or Kubernetes. And we think that that's the IS layer. So we think that there's kind of a new paradigm emerging where it can be much more efficient, and frankly, lower cost. Data Center and Infrastructure as a Service hardware as a service with an IS we're best of breed software is portable

Greg Irwin 12:49

Perfect, perfect. So we're gonna do this, we're gonna, I want everybody to ask questions, so that we can cover the specific things that you're dialed in here and you want to listen for now, the downside of that is we might bounce around a little bit, but that's alright, we'll keep that on our toes. And, you know,

Zachary Smith 13:06

we'll do plenty of coffee.

Greg Irwin 13:08

There'll be a little messy. But let's do that. Please keep questions coming. And we can also try and answer them here on the on the side. question here from Matt, Syracuse, Northwell. Health, I talk a little bit about how you drive change culturally, in terms of cloud, particularly when you're talking to the infrastructure team, whose job it is to keep highly available infrastructure available, you know, ready for the organization.

Matt 13:37

just preface that with one thing is that your development teams are always out front, right? They're ahead of everybody else, and they're pushing the rest of the company. So development teams can develop infrastructure as code. But that's kind of a dangerous thing to allow developers to do that. So we want to retain that within the infrastructure teams. Obviously, infrastructure teams need to get up to speed. And you know, they don't have the same skill sets that developers do. So just looking for some advice and insight here.

Jack Hogan 14:11

Yeah, man, that's it's a really great question. And I start with, you know, my personal experience here, which is the just the general concept of DevOps. And I use that as really two separate terms, but we've we've merged them. And what really is happening is that the Dev side is pulling away in terms of getting onto the edge driving further and further innovation. And you end up with an Ops team whose real job is to ensure the security the uptime, the enterprise capabilities of the infrastructure, and the further they pull away, the bigger the cultural challenges and they start to resent each other in the way that we're working with Equinix is actually now allowing the infrastructure operations teams to use the same level of skills, the same level of training and really catch up to the Dev side of the house, while bringing in the really proper constructs of security audit by improper automation. And in thinking through what disaster recovery plans or business continuity plans need to look like. And so bridging those back together, so by bringing the physical infrastructure, bringing your storage arrays bringing physical compute layer a layer, the traditional operation side is now able to understand how to operate that they know what it means to manage bios on a server, they know what it means to have separated VLANs. And in different network layer segregations. It's so it's the same really common plane that they're used to working with. But they also can deliver that same automation on the front end for the for the Dev side of the house. And so it really is that bridge between the two worlds. And that's where, you know, we talk about it in the form of reimagining your cloud journey. And it's not, do I ultimately go to a cloud model, it's allowing your existing staff, your existing trained operations team to catch up to that world use the basic constructs, they're using us to Zac mentioned, the primitives, like those primitives are physical servers, single tenant isolated the ability to segregate on security. And so it's, it's a great way to bridge that challenge that really starts to exist as an acceleration from the the Dev side really increases. So as I mentioned earlier, it's kind of that that happy middle ground, allows for the full automation infrastructure, at the speed of software. And, and really, you know, that ability to meet the needs of the developers as they need it, and, you know, allow policy driven infrastructure management to really exist in a physical sense.

Zachary Smith 16:51

And maybe I'll add something on to that, Jack, you did a great job explaining it. One of the reasons why we built Equinix Metal, the way we have is that we really feel that, you know, taking an opinionated verticalized approach up the stack, saying, Hey, you can have automated developer experiences, as long as you build everything around this entirely new paradigm is actually like an incredibly, largely been huge investment. Right. And so what we focus on something that's a really, really common activity for people using Pure Storage, and Equinix Metal as a service would be something like automating your VMware cluster, like bringing up a vSphere cluster from scratch, with a basic script like TerraForm. You know, it's you can copy something off of our GitHub, you can use it in like a couple days, suddenly, you have the ability to create and recreate and start on a journey where you're saying, hey, all I'm getting to is maybe a vSphere environment that I I'm pretty comfortable with. But I'm doing it in a programmatic way that I can start to introduce, I'm going to say it infrastructure is code, just getting back down to like an IT system level and doing Hey, I'm on my port configurations, my VLANs, my bring up my ESXi nodes, the adoption of vSphere and my storage units to like, programmatically calm. And that's a journey, I think that most people are going to be pretty comfortable with in terms of the primitives. And if we can help you along that path to automate those primitives that provide the stepping stone for modernizing organization, adding new skills, but not re-architecting your entire IT stack around a new model, which is frankly, quite a leap, and maybe not even appropriate in many case scenarios.

Greg Irwin 18:22

I'm going to come back to jack with a question in terms of the offer, Jack, how should an infrastructure manager think about this offer from Equinix? Relative to the other ways that they can consume Pure? Because you you guys have been doing variable purchasing of equipment on prem for a while, and people could stand that up in any third party? How is it different? Hear what you're doing with that?

Jack Hogan 18:49

Yeah, it's a great question. And a great distinction here. So Pure pioneer, this whole concept of as a service hardware delivery, we introduced here's a service over four years ago, we brought the ability to truly turn that into a service rather than what we see a lot of our competition do, which is really glorified leasing. Pure is actually delivering the full physical lifecycle management of the hardware rights. We back what we're doing with Equinix. With that same underpinning, that same capability to deliver SLA is deliver a class of service, but bring it up a much higher level and an a much higher value, which is now you don't have to think about do I need to go procure a data center? Do I need to have staffing to run that? Do I need to buy my own switching and interconnects all the digital infrastructure services components are now included in the price per gigabyte that you pay? And it starts to look very much like what you would get out of a public hyper scalar. But you're not making the trade offs. You're not making the trade offs of multi-tenancy of QoS QoS challenges, you're not making the trade-offs of ultimate egress and locking challenges. This allows you to take What you do in your data center today extended into a model now, where all of those hosting components are included. And then you can connect that directly up to a cloud. And you can do that. And ultimately mountain we've got customers that are mounting compute hosts out of a public hyper scalar, all the way down to the storage. Or even better, you can consume directly adjacent to that the compute nodes that you can get in a Bare Metal sense from Equinix. Metal. So it gives you all of that same flexibility of purchasing model the same just in time only paying for what you're using, and, you know, Pure, very different than public, hyper scalars, or even some of our competition, we only charge for the data that's written onto the platform, although we make additional capacity overhead available for bursty workloads, temporary growth of data. And so we bring all those benefits, but deliver it in a full-service infrastructure on-demand model, that really brings a whole new level of level of value, because you're not having to commit to data centers. I mean, think about it a data center commitments, typically a five-plus-year commitment. In this scenario, you can consume on a one-year minimum term commitment, start with 50 to terabytes of storage. And you can scale that as your business grows, you can use it as a temporary environment, you can use it as a permanent environment, you can consume compute, in a reserved sense, if you know you've got regular workloads, or you can consume that on demand. One of the greatest use cases I see is the what I call pilot light DR environment where you can place your data there have that ability to stand up a vSphere environment, as Zac said, build the scripting for that, but only pay for the compute nodes if you need them. Like that's, that's that perfect scenario for DR. You're not having to duplicate infrastructure, you're not having to set up your own data center as your DR Data Center. You know, these are a lot of the big things that this consumption model and pairing with the partnership here really delivers is a key value to the business outcome.

Greg Irwin 21:54

Who has the support model? So if there's something wrong if this if the system isn't performing, and I need to make a phone call, who am I calling?

Zachary Smith 22:04

Yeah, so I mean, basically, you're going to do tier one with with Equinix Metal where the operator, right, so we've got the data center layers, the data center online, we've got your physical infrastructure of servers, okay, we've got your networking class networking as the network on and then we've got the Pure physical infrastructure side. Then after that, you get a subscription, including to Pure One, which is the full management portal and whatnot, to manage your the health of your peer storage environments, regardless of where they are.

Greg Irwin 22:29

Fantastic. All right, Zac, tell us a story. One customer tell us about one customer, one environment and how it's working. Sure.

Zachary Smith 22:40

Yeah, I think there's I can give you lots of stories, because it's been a fun year, started at Equinix on March 2, when they acquired packet and then in March 4, we went into lockdown, and have been a pandemic mode ever since. And everybody thought we were going to have a nice sleepy integration, we were going to just like, you know, go to some cocktail parties and talk about the big ideas for the future. It turns out what we did is every single IT organization was like, oh my god, the internet and my customers and remote working when bonk bonkers nuts. And also I can't go anywhere. Like I can't actually go to these data centers anymore. And then of course, year in here comes the supply chain challenges and friction around, you know, both supports is not just affecting our Christmas gifts, as we all know, it's impacting all of the, you know, equipment and fabrication capabilities and cables and everything else. So it's been a really, really busy year helping more customers move faster with digital transformation and India as a service models because of necessity, or because, you know, that was part of their strategy. Right. And so a good example would be and something that I can reference publicly. You know, we explained this at one of our investor days recently, it was Sysco Foods, right Sysco Foods, you know, you know, large, you know, multinational in terms of food service, you know, deploying Sysco with an s not with this. Yeah, yeah, exactly. I'm sure that the food service side of Sysco, the networking company also has robust needs. And maybe they've been a little bit bored during the pandemic, though, at the corporate cafeteria. So yeah, Sysco and how they designed and automated their, the movement of their network using our digital infrastructure side with network edge, where we offer NFV appliances, I believe it was actually Sysco appliances, as a service. So these are virtualized instances they can remove and rehome their network. And then using Compute and Storage with Equinix Metal, I'm not sure if they were using Pure but you know, this model where they deploy physical infrastructure programmatically and they built all this in the US and then they deployed it in five locations throughout Europe because they had to re refresh and handle kind of like increased logistical load happening throughout Europe. And they did that I believe in under five days. We have the the kind of details about how this happened but they just like literally turned up Frankfurt and London and Paris and Amsterdam and Madrid right and completely automated the IT turned out without any humans going to Europe have their up their infrastructure stack, which is effectively I believe VMware and you know, kind of a Sysco Networking application and storage. So I think this, this is helping, you know, key enterprises. And then also, we've seen, you know, with traditional, I'm going to call it modernization efforts. And then we've seen that the whole new application workload, it's super fun to be working with a disruptor, like Pure Storage, because it's not just that they have this and like incredible block and file access, they also have, you know, a really deep integration into things like Kubernetes, with port works. And so seeing how customers have been leveraging container aware storage, and just moving faster with rewriting technical debt, launching new applications, and then doing so globally, which is really fascinating now that people can use workload managers and data becomes like a powerful part of it instead of a burden that they have to worry about. So it's pretty, it's pretty fun to watch.

Greg Irwin 25:45

What's what was the key driver for Sysco, I leave there lots of value props, you guys talk about, hey, somebody else's problem. It's somebody else's, I mean,

Zachary Smith 25:55

surely alternatives for Sysco, what the alternative for Sysco was, you know, take six to eight months, by colo ship here, send people, you know, do the systems integration with a partner, and get their kind of, you know, I'm gonna call it localized network and applications online, they had to do these in country, or at least in region because of both latency for their kind of distribution centers, and also because of some data privacy things that are happening in the EU, right. And so, you know, instead of doing that, which would have taken them the vast majority of 2021, and 2022, you know, they were able to focus on having it as a service moving quicker, having it worked exactly the same every place that they went, and then focusing back on delivering to the business. So for them, it was kind of a no brainer, I think, from a programmatic and removing friction from the process, which is what kind of what I refer to earlier, remove friction, they're actually doing exactly what they would do in colo, they're just doing a whole lot less friction

Greg Irwin 26:48

You know, I have a couple great questions have come in. Gary, thank you for that I'm going to cover in a moment. One question that guys were getting on every single one of our sessions is supply chain lead times and out of stock on specific skews. So Zac, I don't I don't envy you. Because you if you have to stand by SLAs is on availability of product. So how is it? How is it in terms of being able to get your equipment up and running and online, to be able to serve serve your customers?

Zachary Smith 27:21

Yeah, well, it's not, it's not easy, right. But we've taken a whole bunch of the problem set things what I like to call owner economics. We have owner economics on data centers. So we've deployed over 30 megawatts of space to Equinix. Metal this year alone. That's because we're the world's largest provider and creator of data center space. So we have a lot, and that's good. And being able to manage that on behalf of our customers is something we think we can be best to the world that so we're just gonna lopping off a big portion of that. And that starts to include things that we have with our supply chain partners like Schneider Electric, and related, which is a We spend billions a year on capital investments, we also buy a lot of racks and PDAs. In fact, 10s of 1000s of them. And so having that capability in region just starts to remove a lot of the what I'm going to call low value, but high lead time items. Like if any of you have ever shipped a PDU to you know, in a rush job, right? It's really expensive, even though it's a, you know, $1,500 part, you got to get it there. And then the same thing starts to go with things like cables and optics, right? We're we're now starting to be able to handle a lot of those things, both from a supply chain aggregation standpoint, because frankly, the supply chain is figuring out how do I get all these orders to the customers, of which 33% of them are going to an Equinix data center. And so we're starting to help them with that and start partner and go deeper into the supply chain. So we can make things like spare parts and better delivery of those kind of, I'm going to call them sundries, right available for our customers. And then the harder thing starts to become things like network, where we can offer some scale there because of multi-tenancy. So instead of you know, only using 16 of the ports on your 64 port switch, you know, we're able to help with that, right. So we have some, you know, economic benefits of being able to do aggregation throughout the data center. And the last one is actually in the compute and server-side and storage, you know, being able to partner with Pure, who has, you know, a robust supply chain effort and, you know, strong partners for their own product and say, Hey, instead of just trying to get it to all of you customers and trials and on-demand and more capacity, now we've got a scaled relationship with Pure where we can do that globally. And so we're managing it now in 18 locations around the world. And as we expand that to 30, it's a program we're doing. So if you need it in Stockholm, or Toronto, or Miami, we're going to be able to take care of that for you along with things like very mundane things. tax jurisdiction. Yes. Joining done. Yeah, yeah, I got them. Yeah. Things like tax import duties, you know, having these in all countries that you can just like slow down the process. These are areas where I mean, Jack, I mean, how long do we spend working on importing taxing for our customers probably the like eight months with our tax and legal teams to make sure that this could work everywhere. And so like, these are the issues that we feel that we can remove friction from the system doesn't mean we can solve everything and say, Hey, we've magically made all the supply chain problems go away. But it does mean that we're paying a lot of attention, putting our resources towards it and trying to smooth it out for our customers.

Greg Irwin 30:22

Excellent. Alright, let's go. One more question on DR and failover, I think is a great question. Great point. And thanks to Robert, and, and Gary on that. And then let's go around to some stories. So I'm going to tee up. I'll tell you I have a friend on the line. Adam, Jeff, a I'm going to come to you next. But then I want to go to several others as well. And talk about some stories about, again, some of the real real-world issues or successes of matching kind of this tiered, tiered environments. Jack talked to us about DR Potts, DR Work with these solutions.

Jack Hogan 31:02

Yeah, listen, one of the benefits of this is it's a it's a giant clean slate in terms of how you want to use the platform, it is a Bare Metal infrastructure. So the way you can think about it is it's your own data center, operated as code in another location in the world. So whatever you want to put on top of that, you want to put vSphere you want to put SRM, you want to have some sort of Pure has the ability to take from your current on premises data center, leverage things like active cluster or active DR, which allow the data to be able to exist coexist in two locations. And then you can you can automate anything on top of that. And so the way that we look at DR Is it can, it's a building block, and it's starting with that code, you know, Colo as a service or data center as a service. So you're starting at the lowest level you need to where you're getting the physical infrastructure, and then you can layer in whatever software one of the big benefits of this is, unlike what you look at in public hyperscalers. Are you looking for managed service providers where there's opinionated stacks, there's opinionated software, there's opinionated environments, this allows you to bring in however you want to architect that. So that ability to leverage native capabilities within your storage arrays or operated at a higher level and push the data across the wire. It's a perfect scenario for DR We've got a couple of joint customers that are doing that exact thing. We have a couple of customers that are also taken out one step further. First Data Center, they moved out of the particular customer SaaS company and customer based in the north and the northwest, decided that they wanted to move out of their primary data center and move that into Equinix as an Equinix Metal facility as the primary location for their data, their data set. Then they said, well, our DR data center, which sits out in Denver, they're now moving that into another Equinix. So they've fully virtualized their data center environment, with the same construct of it can be active, passive, active active, you have all of the same flexibilities as if it were your own data center that you're running, as well as all the same constraints, the constructs of being able to segregate workloads have security and all of those key current key capabilities. So it really opens up all possibilities for any type of DR environment you want to establish.

Greg Irwin 33:18

Is, is the idea here, though, that you're going to package. So I mean, over time, I think there can be a view of you want to add on the services. So it's, you know, a pre package DR Offer, or you really want to leave it Bare Metal, and really let your customers and partner and partners come forward with their own offers. What's our strategy?

Zachary Smith 33:44

Here? Let me let me provide some context there. And so I think you'll find that both Pure and Equinix are really ecosystem-based companies. We're very, very partner focused, which is we know what we do best in the world. We're not trying to do all the different things. And so we have a couple phrases here in front of Equinix. But like us plus two partners equals a deal. Like mainly it's because we're we're a digital infrastructure company and physical infrastructure business. We offer primitives, we know our customers are looking for solutions. And so Pure is part of that where they can help bring, you know, kind of best in class storage capabilities. But we know that there might be a managed service partner in the middle, there might be somebody coming in with workload management, we've been working closely with people like Nutanix, VMware, Red Hat, to be able to bring, you know, operated software environments. And so we'll continue to build those ecosystem capabilities. And whether that's by making it easy by giving upstream integrations, like what Pure and Equinix have worked on related to upstreaming, our TerraForm integration so that you can use TerraForm across both our infrastructures and have them be compatible. Or it's actually fully managed offerings where there maybe are MSPs or even direct software companies like VMware offering, you know, kind of fully integrated operated environments like what you might see in VMC at some of the public clouds Right, something to note with Equinix interconnection is built in. So if you also want to use DR against hyper scalar, and burst compute into Amazon or GCP, or Azure, using your Pure Storage through their marketplace, or through that Pure one, I'm sure Jack can explain how it works. But, you know, you can also access resources in in public clouds. And that includes their operated things like we just launched with Amazon's Elastic Kubernetes service Eks on prem. Or as now I think it's called Eks, anywhere. So you can control Amazon workload in an Equinix data center connected to your peer storage device. Like it's pretty cool. So like, we are an open ecosystem-based company, so we don't aim to solve it all. But we aim to have the right ecosystem around it to really make our customers successful.

Greg Irwin 35:47

Right, great stuff, guys, let's let's shake it up a little bit, I want to bring some people into the conversation. And what's a little different about BWG events versus lots of webinars, and I'm sure everyone dials into his, I really like getting other people involved. So I'm going to try and do that. Steve, Steve, Frank, you happen to have your camera on, I hope, I hope you're in a spot where you can share a little story or question with us. But do you have a minute just to give a little intro to this, this group that we've assembled? Sure, Hey, real quick, what's your focus over, it was fun.

Steve 36:25

So my focus is architecture, and strategy within Enterprise functions. So not the consumer, facing part of the bank, but all the back office stuff, compliance, legal audit, you know, areas like that. So, you know, the some of the challenges and issues that I face on on a right now. And one of the main objectives and I've had really over the two years that I've been here is, these are areas of the of the bank that were not a focus of technology in the past. So you know, they were kind of left on their own to come up with their, their own tech solutions to do their business. And so right now, there's a significant need to modernize several of the lines of business within within those parts of the organization, which, you know, it ranges from, like very legacy apps, you know, 10 plus years that have, you know, just really nothing that we could lift and shift into, you know, a cloud or something like that. But even beyond that, is that, you know, so much of it is there's, there's a need for greater transparency, and really getting these applications in these areas of the business to be able to share and, and have, you know, greater transparency, collaboration, not only for within the business, but like for regulators and so on. So, you know, within that area we just with, within the bank, we've been, we've had a couple of different in private cloud instances that we've used. And so next year, we're looking to move into public. And so there's, with by mid next year, we'll have on prem, Private Cloud Plus public cloud as

Greg Irwin 38:27

your big move. That's a top five financial service platform to be going public cloud says a lot about how far public cloud has come. Steve Weiss, Jacqueline, one question you've got for the group, not just Zac, or Zac, and Jack, but I can take around to others. I wonder about as we go around the group here?

Steve 38:47

Sure. I would be interested to just for like a comparison with what a lot that, you know, was being described is how does that compare to differ from like, just a private cloud, like Pivotal Cloud Foundry or the VMware?

Platform?

Greg Irwin 39:12

Are Straightforward, Zac Yeah, that's, that's a perfect one for you. How is it different from a private cloud? Stan? Stan? Setup?

Zachary Smith 39:23

Yeah so um, you know, from from a private cloud standpoint, both of what you mentioned both say Pivotal, or VMware, like VCF, or whatnot, you know, really what we consider are these I consider to be a software layer of shippable as you're having a private eye as deployment with things like, you know, runtime and orchestration and, you know, further level networking and segmentation and all those kinds of things. The way that we're we're trying to solve is one layer below that, which is providing the infrastructure in an automated consumable way to at least power those capabilities. So we work closely with VMware with their pivotal product or They're, they're Kubernetes or straight hypervisor side. And most customers bring their own opinion related to that, just like they're getting used to about bringing that opinion into into public clouds, for example, running VMC on AWS bringing their VMware workload, and using that as a portability layer across private, you know, hybrid or public cloud deployments, we've seen other people go further and just do straight up Kubernetes, they're able to ship their entire Kubernetes deployment, or work with partners like Red Hat, you know, on OpenStack, or OpenShift. So I think it's, it's, it's actually a really interesting time frankly, for some people, we're a little bit late in moving cloud, actually great, cuz there's this boomerang happening where the software got really good at it as people went to one cloud. And then like your organization, they would multicloud, you bought companies, you had different proprietary services, you wanted to use a different hyperscalers, whatnot, you end up having to normalize that because it's too hard to maintain all the integrations and all the work. So you end up normalizing that around some sort of shippable runtime, like Kubernetes, or VMware, and that that represents an opportunity to have more choice around your infrastructure stack.

Jack Hogan 41:02

And I'll add to that, Steve, I think one of the other big things and we're starting to see the trend in the market move towards this concept of hybrid multi cloud, it's really the essence of saying it's no longer in or model you're not either in a private cloud or a public cloud, you're not in, you know, AWS, or Azure, or your data center, that ability to really, really meet in the middle. And what I'd say is where Pure is really looking at this and the big regulatory elements, the big security elements is where does your data actually sit? Where's your exposure for that? Where is your regulatory compliance for that, and by extracting the concept of being able to put really, in Equinix Metal environment is that center to the hub of all those capabilities, public cloud, private cloud, you know, major, hyper scalar, or a managed service provider, that's a big key to what we see that this solution is ultimately solving. And it's, it's really getting ahead of that market in his accent. It's allowing customer companies to catch up to that without having to completely retool and go all in on any individual layer within that stack, or any individual horizontal play within that which would be you know, public or private or hyper scalar base.

Greg Irwin 42:12

Well, let's, let's keep going. One note here is obviously this is not a one and done. If there's follow up that you want to do, of course, and then one last thought, remember, I want to encourage people to connect across the group. So keep that in mind. Let's keep going. And by the way, I love to see the chat moving. Please drop your questions and keep that side that side channel, though. I'm going to bring I'm going to, you know, again, just random shot I'm going to invite some people in and maybe we can share some stories, Michael the gel over at ETS Lindgren, Michael, I hope I'm pronouncing your name properly. Are you on the line with us? I am here it's actually Fogle.

Greg Irwin 42:56

Alright, sorry. No problem. Give a little intro please.

Michael 43:00

Well, so I ETS, Lindgren, does RF test and measurement equipment. So we build positioning equipment, chamber shielding, you name it in the RF space. So I primarily deal with you guys related to 5g networking. Now part of my interest here and I actually dropped that in the chat every every time we start talking to people talking about cloud and doing other things. A big part of our business that, you know, nobody knows who we are hears about us other than the people who need us is we shield data centers. You know, we protect them from outside interference, physical attacks, essentially, you know, EMP that sort of stuff. So we build shielding and filters and, and yeah, right now the the filter business is going through the roof, because there's a lot of this going on. So that was one of the things I was kind of curious with this is okay. So, you know, we can talk all day about the software side of it and everything. But at the end of the day, those data centers are, you know, us as a company, you know, every time I think about us moving into the cloud, that's the first thing I start thinking is how safe is our data, not just from the hackers, but you know, just in general.

Greg Irwin 44:13

Yeah. Okay. Well, let's do let's do it guys. What do you do to secure physical infrastructure?

Jack Hogan 44:21

So well, first of all, starting from the storage and data point of view, and so the way that Pure is ultimately architected to be able to have data data placed in NAND chips on non commercial SSTs, these are actually Pure actually manufactures our own our own D. FM's our own fabric modules that ultimately allow that to use the lowest layer of how data is stored and manage that with the software that sits on top of that. So ultimately, the way that our arrays are architected are to prevent against the manipulation of data through something like an EMP or any sort of external interference. And so that's really one of the core areas that, you know, moving to a silicon-based media has allowed Pure to really take over take the market and become the market leader in the All Flash space. But it also allows for different layers of data protection at the physical layer. And so that where the data is placed in the in the different cells within the NAND and we're now using IQ LLC, and its latest iteration, we can get a little bit deeper in terms of how the the physics of that work, but ultimately, the data itself is going to be protected at the at that physical layer, because it's it's really how it's spread across the NAND and and not capable of really being impacted by that. I think, you know, the, you know, from the physical data center point of view, I look to Zac

Zachary Smith 45:46

Yeah, well, I mean, you got Jack's coming from, from the tree tops, I'm gonna come from the bottom right. So we have a lot of experience in 1998, building, kind of robust data center facility. So Michael, I'm sure you know, about that, and how some of the most critical workloads, both government and, you know, about 20% of our revenue comes from financial services, where we run vast majority of the financial transactions that go through, you know, that we rely upon in the world. And related, so we have a pretty strong and robust physical security regime, as well as availability, you know, really don't spare our expenses related to to our datacenter footprints. And then as it pertains to the physical operations, of course, you have a few things, right, you have the logical side of the network. Luckily, we have 3000 networks to choose from, at Equinix. And so part of that is, whether you're using our global IP network or our fabric, to reach between locations, you can also reach your cloud on ramps, your different, you know, 1000s of NSPS, private connections, we actually call it the industrial internet, the internet where you connect privately with people in programmatic ways. And so keeping more and more of your sensitive data or related stuff in a high speed, localized interconnected fabric, is a really awesome alternative to I'm going to say, telco MPLS, or public Internet, right, like, this is a really awesome capability that a lot of our customers do by just removing huge portions of their traffic flow from the internet, you can be a physical cross connect our fabric. And then from, from a system standpoint, you know, our customers leverage all sorts of software, like we said, above the stack, right to, you know, introduce zero trust into their workload, you know, programmatically control that, frankly, one of the biggest things as people use our our programmatic access to infrastructure to change, like, get into the cattle versus pets scenarios, actually delete their infrastructure on a regular basis, one of the best things to do is go through our full firmware automation, our full drive wipes everything else and recreate their systems. And so that's part of the actually the security strategy that a lot of customers move to, because of other issues that they have related to systems breach or drift or configuration management or whatnot. So there's a variety of strategies, but our job is there at the physical layer, and then making sure that we have the right interconnections to, you know, move your traffic of the places that that you need it. Excellent. All right,

Greg Irwin 48:05

let's keep going here. Guys, we got 10 minutes left. Let's finish strong. I know everyone probably has something coming up at the top of the hour, and everyone will run for it. And that's fine. But we got 10 minutes together. So let's make really good use of this time. I want to bring in Darren Young. Darrin, I hope you're with us, you and I get to talk on Autodesk all the time, I'd love to talk a little bit about infrastructure from your perspective, because I don't know how the rest of this group to what extent they understand CAD CAD software, and just the huge amount of data that you move for your designers and engineers to run them. But I'm really curious about your perspective on how you run that from a third party data center versus on prem and how you're thinking about it.

Darren 48:51

Oh, I don't know that I could add a whole lot to to some of your other speakers there are a lot of our stuff tends to go through the cloud solutions implemented by our software providers. You know, in general, just our data sets are so large, especially when you get into things like laser scanning and 3d parametric models of building and civil infrastructure. You know, the application needs to be co located the same location as the data, there's just there's too much to pull across. So you know, one of my bigger challenges is a lot of people in our space are not very sophisticated. They're they're, they're busy going after revenue selling product, even a company as large as Autodesk we have no service level agreements with them. I guarantee where they're hosting their data, they have service level agreements with their data center partners, but we don't have any there's no there's really no backup strategy extended to the customers. That's one of the bigger challenges I think we have just from an end-user standpoint

Greg Irwin 49:50

Hey Derek, are your service and storage then I mean, I know your Pacific Northwest maybe Seattle. Are Are your servers and storage sitting in In the same facility as your designers, or have you gone to a managed service offer, what's what's the basic structure? You know,

Darren 50:10

most of the data is stored either not our desks, wherever your Autodesk uses for their clouds or Microsoft's clouds in the Seattle area, Microsoft is king. So that's tends to, you know, anything non CAD centric, tends to go to Microsoft direction, everything else goes to wherever Autodesk puts it. And then, you know, there's a variety of other third parties there. In general, though, you know, my, my, my business, though, actually, we do build data centers for people. So we build the infrastructure honor, I would, I'd love the opportunity to actually kind of do where we could hope post some of our own stuff, but like I said, the software required to us we don't get to make some of those choices. Unfortunately.

Greg Irwin 50:52

There's not big enough. Good good to speak you. We speak with you interesting. Little I'm going to do a quick sprint, bring some put some people on the spot here. Got them got them dossier got. Are you on the line with us? I saw you were Oh, yeah. Our documents solutions. That's why I wanted to get you involved. Are you with us cotton? No. Ah, my mic pilot. Mike, I saw you a little bit earlier. I haven't looked here in a moment. Mike, are you with us? No. Okay.

Zachary Smith 51:25

This is this is like second grade, I used to raise my hand first. Otherwise, the teacher would call on me when I was a junior.

Greg Irwin 51:30

Do it. Let's try Ron Allen. Same thing. I'm thinking, Ron, you and I talk Autodesk. What do you tell us a little bit about infrastructure?

Ron 51:40

Yeah, so right now we're looking at shifting to cloud, like a box cloud infrastructure. And we've also got some box flex machines in our server rooms. And biggest issue that we have is we've got that gap over the web. And I think probably what would be the best solution there is if we were to go through and get into a core infrastructure where everything's in the same location. So everything's data, Jason, that's the biggest issue that we're having right now.

Greg Irwin 52:08

Got it? Where where are you? Where's, where's the data?

Ron 52:13

So I think that the the box cloud is outside of Virginia, it's over on the Eastern Seaboard. And then we have our offices that are scattered between California, Denver, Illinois, and Virginia as well. So it's just a matter of we're kind of in that that awkward in between stage and trying to find the infrastructure to support Revit, in particular, because it's such a chatty application. Yeah. And the latency is just killing us, you know. And so being able to have that have those layers up above the actual data's data storage so that all of our office can communicate effectively. With some of these model sizes. Typically, we're looking at model sizes of 70 to about 400. Meg's, and then in more extreme cases, eight or 900. Meg's. And because it's a database application, again, it is so chatty, and just causes so many issues. Yeah.

Zachary Smith 53:11

Well, definitely modernizing around network, we call it there's a great paper we just put out, it's our annual report on on the state of interconnection, and or an interconnection or an architecture really like massively increasing throughput and lowering latency, they kind of go together. And so having that network effect of concentrated partners, especially with I'm sure work from home offices, people food, storing things in different clouds, definitely a big part of what we do here at Equinix is having that network aggregation to dramatically lower those costs and increase those speeds. So I can point you towards are some of the data we've got, especially in that Northern Virginia market, which is probably one of the most interconnected markets in the world.

Ron 53:52

Yeah, I appreciate that. Thank

Jack Hogan 53:52

you. And Ron, I'll say that, you can start looking for some announcements that come from our site about where we're actually taking some of this low latency interconnectivity into some of the larger clouds to then be able to access high performance computing, workloads and environments. So that ability to use those interconnections. Zac talked about architecture around that is something that we're really seeing very large scale HPC, high performance computing companies really see some great success.

Ron 54:24

Excellent. Yeah, I'm really excited about what I've been hearing so far. So

Zachary Smith 54:28

what are you guys working in? mL? You haven't early AI applications, whether in model, I mean, obviously a little bit on this way, right. Similar similar problem HPC. Like, or any of you guys having initiatives that your company is around that.

Ron 54:43

We're starting to work a little bit with generative design, which incorporates you talked about machine learning, correct? That's correct. Yeah. Yeah. So we're starting to work a little bit of that. The applications have some backends built to them. That comes from Autodesk and then basically, it's a sort of touch They off to their cloud to go through and do the iterations for the really big projects. And some of the local computing processing.

Ron 55:07

For smaller projects. We use it with within our fin crime and fraud detection areas. That's probably one of the more advanced applications within Currently, we have some others that are in the works, building out the use cases, and really trying to understand how, you know what the real ROI on will will be. But the certainly fraud detection, the whole fin crime area, that's been for two years, we've been building that out with AI NML.

Ron 55:41

And we're already using that your Trend Micro and in the machine learning aspect for our managed detection and response service.

Zachary Smith 55:53

I'm gonna bring, it kind of brings together a lot of the challenges, which is compute network and storage all together. And the reason why I talk about ml is because it's just like the confluence of all the challenges. So certainly interested, as you guys do that, we'll be announcing some stuff with Nvidia, pretty robustly at GDC next week, but making it easier to adopt those kinds of things. And I'm sure, Jack and I would be happy to happy to have more learnings from you all. But where you all are, aren't ml.

Ron 56:25

II and this is an honor, I'm just curious to know if like anyone is doing anything on quantum computing, any use cases on quantum computing? I mean, I'm seeing like more and more information online like or some organization experimenting use cases on that. Just Just curious.

Zachary Smith 56:46

Yeah, I mean, the short answer is no. From from our where we sit, it doesn't mean it's not happening. But you know, we see that as a pretty pretty leading edge where we are spending their time right now is in liquid cooling. So much higher, higher-density solutions, with the chips, kind of moving even just the regular normal processors and accelerators moving to, you know, I'm going to call it the two kilowatt per one UBox range. And also the challenges we have around sustainability. We've seen a big movement related to heat capture, better cooling, moving to 30 to 40 kilowatt standard racks are the ability to do that. And operating gear, kind of without the water waste within the data center, because a lot of places around the world are starting to regulate that. You know, the evaporative cooling is extremely wasteful from a water resource standpoint. So that's kind of where we see people going right now. And computing generations really wards just like hotter, and more specialized, but not yet towards Quantum. It's not directly related, but just to kind of just add a use case. And here we've got contracts with the Army, Navy, Air Force and FBI around quantum computing. And it's mostly around deep data research and running variety of different inquiries for certain select scenarios. But there there is some interest picking up in that space, just in general who wanted to talk about use cases. But I just wanted to throw that in there. Because we're top of the hour and I wanted to say something.

Greg Irwin 58:19

And just to pile on, we run forums on quantum, the quantum use cases and with the federal agencies and global agencies that are running those tests to try and solve the problems. I think I agree 100% with Zac, it's still really, really early and pretty limited in terms of what you can achieve with Quantum. But I'd welcome anybody here to one of those forums to talk about the commercial use cases. Let's let's do a wrap up here. We've got just another minute here, Zac. I'm going to go Zac first and then jack to take us home. Guys, any closing comments for people to think about before we before we call it a day?

Zachary Smith 59:01

Well, I was like, wait, wait, don't tell me. Right. So what are we going to discuss next week? You know, the headline, right? So I think that the two topics that are on my head that I think are super interesting. One is the overall shift of OEMs Pure leading the the stage with both software and hardware technology to an as a service model. That is going to be what we talk about in 2022. These are massive shifts in business models, moving to operated experiences going to be a lot of confusion, a lot of fun, a lot of learnings, a lot of opportunity. So I think that's going to be really great. And the second one I've already mentioned it a few times it's in every top RFP we see is sustainability. The number one not like very deep down below in our industry has a lot of work to do there got to do it together. Things like exposing, you know, carbon usage and related energy profiles out to the workload and customers and managing that from a regulatory and public relations standpoint, I think is going to become top of mind to those

Ron 59:54

22 Thank you Zac.

Jack Hogan 59:57

Jay, no, listen to the exact kid on a The sustainability element nail what Pierre has already done inside the data center by reducing datacenter footprint, power draw and being able to become a key partner in that initiative is a big area where Pure and Equinix are really leaning in, I think you'll see a lot of that be become front and center. As we look at global regulations, really clamping down in this area, it's an area that we all have to be very mindful of because it does have a direct impact on the health of the planet. So again, is that as we move more and more and more into as a service consumption, and start to address really a lot of the legacy challenges of how do you allow those legacy applications to enter this world of cloud, but do it in an intelligent way rather than refactoring? I think that's a big area that we're seeing a lot of customer interest in, and a big area that addresses a lot of the key topics we talked about here. Yes, the skills gap is very real. The ability to actually take advantage of modernize technologies that requires it requires an easy roadmap there. And I think that's what a big piece of what this partnership really is bringing to market and introducing completely new ways to reimagine that cloud journey and reimagine what it means to be able to be responsive and adaptive to your business needs through technology and love it.

Greg Irwin 1:01:17

I mean, it's pretty simple. We talked about, hey, what are the drivers? What are the catalysts? Well, if you can't, if you can't get the product yourself, if you don't have the team capacity to actually run it, and you can save money in the process. I think those are pretty compelling, pretty compelling drivers. So I love I love what you guys are doing. Thank you, both of you for letting us host you and thanks, everybody for taking time and joining. Of course, we'd be happy to connect you with either of these organizations or anyone across the group and I look forward to speaking with everybody on a future session. Thanks, everybody, and have a great day. Thanks, Greg. Thank you. Thank you.