This episode is hosted by Boundaryless Conversation Podcast host Simone Cicero with co-host Stina Heikkila.

The following is a semi-automatically generated transcript that has not been thoroughly revised by the podcast host or by the guest. Please check with us before using any quotations from this transcript. Thank you.

Simone Cicero:
Hello, everyone. So, we are back at the Boundaryless Conversations Podcast. With me, I have my usual co-host, Stina Heikkila.

Stina Heikkila:
Hello, hello.

Simone Cicero:
And two special guests from the Open Compute Project and Foundation, Steve Helvie.

Steve Helvie:
Hello, everyone.

Simone Cicero:
And Archna Haylock.

Archna Haylock:
Hello, everyone.

Simone Cicero:
Great. It’s nice to be in touch again with you. Both Steve and Archna were interviewees in our research that brought us to write the whitepaper that we released in November 202. You can find the OCP case study in chapter four, if I’m not wrong, something like that. It’s great to have you here because, as I was saying, in our preparatory conversation, I always refer to OCP when I need to talk about ecosystem strategies that really enable change and transformation in the industries where they’re created. And no later than a couple of hours ago, I was with a customer with a specific problem of how do you enable services — new service ecosystems on top of a shared infrastructure? And those are problems that come up very often. So, the first question that I would like to ask you, I mean, of course, it’s a bit of framing of what we’re talking about. So, maybe one of you can give a couple of — a kind of introduction to Open Compute Project. And I would like you to, I would say, focus on the four tenets that you express on your website, and the tenets of efficiency, openness, impact, and scale. So, how do these tenets embody, let’s say, the mission of OCP?

Steve Helvie:
Thanks, Simone. The Open Compute Project is an open-source hardware organization. And it was started in 2011, by Facebook, and at that time, Simone, they were outgrowing their infrastructure. They were still small at that time, 10 years ago. And as they started to get bigger, they started to realize they were going to have some problems with their infrastructure around their compute and storage, and even their facility design. So, as they started to decide to build their own data centers and do their own infrastructure, they worked very closely with the manufacturers of say, servers, and network devices. And they took it one step further. So, once they solve these problems for their own infrastructure, they use their community aspects and open-source those designs.

Big companies, working directly with manufacturers to make something special for them is not new — what is new is the open sourcing part of that. So, back in 2011, they took those designs, open sourced them to a community, and formed the Open Compute Project around that. So, we currently have board members around Intel, Microsoft. At the time, that was Goldman Sachs, which was one of the original board members, they have since moved away, and Google has taken that spot. And we also have Rackspace. And there’s one individual, Andy Bechtolsheim, who is the co-founder of Sun Microsystems. So, within those core board members, they started the group and now we’re well over 200 companies, several thousand engineers working on common problems across the data center. This all works around efficiency and impact and openness and scale is around these working projects, say the 25 to 35 different working projects we have, those four tenets that you outline embody everything that we do. So, every contribution that comes through the foundation is measured on those four aspects. And each group may have different definitions around what is scale and how open something maybe. But that really drives everything we do. And Archna, would you have anything to add on the four tenets?

Archna Haylock:
Yeah, just a little bit more to piggyback on. So, originally, yes, we were a hardware-centric organization. And a large percentage of our portfolio is still hardware-centric. But over the past, I would say four or five years, as more and more companies started to join our ecosystem, the primary hyper scalers still lead a lot of the concepts and ideas and innovation. But we also had innovations from companies that were coming in from the telco arena, or from the Edge, or from some other verticals. And that kind of, I think, broadened our scope from just being hardware-centric, to really providing solutions. And those solutions embodied, not just hardware at the core of it, but also embedded software. We’re starting to look at full solutions that can be deployed and readily adopted by these markets because maybe they don’t have the manpower and the skillset in-house to do it. And they want to get something really quickly and be able to deploy something very fast as a proof of concept.

So, I think our aperture has expanded from what we originally were — what we started with. We have migrated to, like I said, hardware, software, and system firmware. We are now looking at solutions, we are now going beyond the data center. And as the data center kind of migrates outside, to modular data centers, to the Edge, most recently, to the retail, to the FinTech environment, we’re starting to see the definition of a data center change. And so those applications are now starting to come into OCP. And like Steve mentioned, we model everything that we do; all of our practices, all of our governance, all of our contributions, all of our solutions are really held up to those four pillars.

Are they scalable across the globe? Are they providing efficiency, whether in a process or in their metrics or in their measurement, right? Are they open? And like Steve mentioned, there’s degrees of openness. But the basic premise still stands, are they open? And then last, but not least, are they making an impact? And that impact has to be on a global scale because we’re solving solutions, we’re solving challenges. And those challenges are applicable, as we’re finding out to companies across the globe. Whether it’s cooling, whether it’s data center, whether it’s power, whether it’s speed, performance, you name it. Are we making an impact? And I think we measure ourselves and the industry is starting to measure ourselves on that impact.

Simone Cicero:
Right. It’s much important, as an industry for everything we do, essentially. We’re really living in the age of the internet going everywhere. And the internet is not, let’s say, electrons in the air, it’s computers. And those computers need to be built and need to be set up, and need to be maintained, and so on. So, very, very important mission, I would say.

Archna Haylock:
Yeah. And I think just this last year, in the pandemic, all of us are impacted by it. Right? From our kids to our workspace to everybody, we realized how heavily we relied on the internet, right? So, I think that if anything this last year spending in isolation has really forced us to really stress the boundaries of the internet, in all sorts of ways.

Simone Cicero:
So, I hear you talking about growth, and also growing into new industries, new use cases. And you also run not only the specification process if you want, but you also run a marketplace. So, an active space where trade happens. So, my question for you will be, besides the partnership, the sponsorships, let’s say the memberships, sorry, maybe that’s a more clear team. Besides the membership, how is this organization sustainable? How does it relate with the idea of growth, entrepreneurship, sales? I’m really curious to know that because it’s a foundation, it’s technically a nonprofit organization. So, really, how does it work?

Archna Haylock:
Well, let me take that, and then I’ll definitely turn it over to Steve because this is his world. I think that what we quickly realized is having a bunch of specifications and design files are of no use unless there’s products built on them. They provide the first level, they provide the critical foundation that is needed. So, OCP provides that platform for collaboration. And all of these companies come in from all over the world, they have a challenge that they’re facing. And sometimes they face it in the corporate world. But sometimes they face it in a market space or a vertical. And they take off their corporate hats, and they roll up their sleeves, and they come and they say, “Okay, here’s my challenge.” And the community rallies around that challenge. They come up with either a specification that is done and contributed by a few group of members with feedback from the community. Or they write design practices, or they write guidelines, best practices, whatever it may be.

But for specifications, in particular, they’re of no use to the community unless there’s a product that makes them tangible because other than that, they’re just a concept, right? So, our suppliers and our ecosystem that makes those products based on open specs are absolutely pertinent and very germane to our success and to our impact. And when they make those products, we as a foundation, basically just make sure that degree of openness that Steve was referring to earlier. And based on that degree of openness, we have certification marks. And at this point in time, I’m going to turn it over to Steve, because my world stops there and his world is where it starts.

Steve Helvie:
Thanks, Archna. The marketplace that you referred to, Simone is not — This is not a shopping cart functionality. So, it’s not as if OCP is getting a percentage of those sales. This is simply a marketplace where we connect members who have produced products based off of those specifications. And so one of the bigger eye-opening things to me when joining Open Compute was that how different it is in the software world trying to convince hardware vendors that it’s okay to open-source some of their designs. They get very nervous about giving up IP, and that’s something Archna can talk about around the governance and what we take and what we don’t take within a contribution. But getting those companies to contribute and participate in the marketplace. And you think well, from a customer standpoint, what’s in it for them to look at a community-based product rather than just finding my preferred vendor and going from that direction.

So, the benefit to the customer is that you get the opportunity to do dual sourcing. So, the best operating model of this organization is that many companies like to have multiple sourcing for their products. They also like the idea of one specification and multiple suppliers. So, maybe not some of Product A some of Product B from two totally different designs. In OCP, you can have that one specification, Vendor A, Vendor B, and Vendor C supplying those, so you mitigate your supply chain risk. And the other thing that’s in it for customers is you can influence the roadmap itself. So, you can meet with many vendors, you can share your feedback early on about, “Hey, here’s what I need in my product, rather than having a one on one with a vendor and then going to the other vendor and telling them the same thing and seeing which vendor meets the majority or 80% of your requirements.

So, what’s in it for the hardware vendor — so that was the customer side — what’s in it for the hardware vendor is that you get early access to markets. In many cases, we’re moving faster than normal product cycles, because we can get early access to the customer requirements. So, the second thing that’s in it for the vendors, is that I get a broad set of customers. I can meet with both telcos and gaming companies and retail. And I can get a sense of well, where do I want to take my product roadmap? So, we don’t take any percentage off that marketplace, it’s strictly a connection point within the membership. So, the model is really a membership-driven revenue stream for OCP.

Simone Cicero:
So, basically, this idea that the foundation provides a context where the value chain can connect — the users can connect with the suppliers and ensure that market-drivenness, let’s say, comes through the chain — the supply — the suppliers know what the developer for the customers and so on. The only thing that I was thinking, it seems that it’s really related to this idea of a commodity, because you refer to specification having multiple suppliers. So, essentially competition-componentization. So, my question is, how do you see — if you see this conversation inside the foundation on how to move beyond the commodity space. And you mentioned software, for example. What are the processes you see in motion that go more towards innovating in the industry, maybe besides just making supply chains less risky?

Steve Helvie:
There’s a few areas. Specifically, I’ll highlight a couple and Archna can elaborate on a few of these because she’s intimately involved in the projects. On the innovation side, we have the standard projects that people are working on, say traditional networking devices, but Archna mentioned it early on around Edge. This is an area where we are — the community itself is driving a lot of new designs, the aggregated Edge is all the way out at the Telco base stations. The other one is advanced cooling. So, we have engineers from around the world working on the newest advanced cooling techniques within the data center facility and all of that intellectual horsepower is coming together within the community. And the last one I would mention is around sustainability. And I’ll let Archna — she’s co-leading one of the sustainability workgroups here, and I’ll let her comment on that.

Archna Haylock:
Yeah, to go back to cooling. You know, it was a concept that was brought together by our European counterparts that seemed to be ahead, in some ways on cooling. And they came to us and said, “Hey, we really need some design guidelines for immersion and IT gear specifically in immersion. We need to understand how other people are handling cold plate and/or door heat exchange.” So, what we started to do is have communities that coalesced around all of these three concepts of cooling. In some cases, you had more Asia-centric team members, in some cases, you had some European-centric team members, in some cases, you had a combination of both, because I think immersion is starting to impact both parts of the geography. And it was really interesting to see the kind of things that they’re trying to come up with. They’re trying to come up with material compatibility, they’re trying to come up with fluid dynamics. And these are all things that have been dealt with in other organizations too, like ASHRAE. But now we’re starting to look at it from a hardware perspective, and not just a pure fluid perspective, and the impact of that fluid on hardware. So, there is tremendous amounts of innovation that’s happening, because you have these companies that are coming together and trying to solve the challenge of cooling, right, and the cooling in an environmental sense too, and in a sustainable sense. So, that’s one thing.

The other thing that started to happen as the cooling group went underway, and this is by far, one of our largest growth communities, is we were so centered on the rack and the equipment that we were forgetting the data center, and the impact of cooling on the data center itself, on the facility. And how do we take either a brownfield data center that’s in the middle of migrating and looking at immersion options or cooling options to those that are brand new, and are considering cooling as part of their ecosystem? So, advanced cooling facilities were born out of that, and we had connector guys, and we have power guys and rack guys now looking at that. So, there’s innovation happening now around the facility. And they’re taking all of the learnings that they had in the advanced cooling solutions work stream, and now applying it to the data center facility. So, lots of really cool things happening there.

Now sustainability is one of our newer initiatives that we established for 2021. We’re still trying to figure out what sustainability means to the OCP ecosystem because every company has their own view of sustainability. And some people talk about recycling, and some people talk about secondhand gear, some people are talking about heat reuse, right? So, when you say sustainability, it’s such a big word and it encompasses so much that we’re still trying to distill it. It’s breaking up into two different work streams. One is defining metrics for life cycle assessment, and then the second one is around actual — the lifecycle of the systems and the components. So, they’re getting input from companies like Facebook, Google, Microsoft, they’re getting their inputs, and now they’re starting to coalesce around these two work streams because those two work streams seem to have the most relevance to our community, right? I think at the end of the day, they have to decide what that impact will be to the rest of the community, and then they’ll socialize it. So, they’ve been kind of socializing it a little bit in the server group. They’re starting to socialize it in the storage group and trying to learn from them and seeing what storage vendors think about sustainability.

Enterprise, everyone’s talking about enterprise right now. Right? It’s the new buzzword. Before it was the Edge and the mist in the cloud and now it’s enterprise. And so what we’re starting to see is, as I was mentioning before, as data center centric equipment now starting to be deployed in these micro data centers in the campus environment, in branch environments, whether it’s for retail, or FinTech, or automotive or transport. You’ve got cars that are mini data centers, you’ve got trains that have data centers on the side of the tracks. So, it all just depends on what one calls a data center now. It could be a container sitting in a parking lot, and that’s now considered a data center. Right? Those are the kind of questions that’s happening around the Edge. We’re looking at how server requirements that were maybe used in the data center could now be deployed towards the enterprise. What connectivity solutions? Is it 5G at the campus level? Is it something else, right, from an access point perspective? I think that’s being explored at OCP. And then they’re also looking at storage in the enterprise. What does that mean?

So, innovation at OCP is happening in so many different areas. We mentioned cooling, we mentioned Edge, enterprise is the next big thing, right. So, we’re talking about all sorts of enterprise-related innovations. I think that, if anything, it’s nice to see a diversity of companies, and a diversity of geographical scale. We have different communities organically coming up in countries like Taiwan and Japan and China. And they’re kind of starting to now think OCP and innovation too, locally. So, I think that it’s happening on a rapid scale, and it’s bringing us together.

Simone Cicero:
Yeah. That’s what I was thinking when I was listening to you, this idea that it’s kind of an accelerator of the industry. Like we get together to deal with our infrastructure so we can accelerate our infrastructural developments, and the other parts of the ecosystem can think about services and so on. So, it’s a kind of amplifier of the speed of innovation in the ecosystem when something so shared exists. Because when you were talking about that, I captured also that when a new innovation comes up, it comes up already in open source, which is such a powerful idea that something very new, oh, it’s already in the commons. So, you don’t have the IP, and the things that normally would stop the innovation to happen. It’s already shared, it’s already open.

Archna Haylock:
Yeah, I think that that’s kind of I think the beauty of OCP is that we don’t create projects, they organically come up from our community. Our community comes forward with a need or a challenge. And then I think as the concept or those ideas kind of percolate amongst our community, they naturally develop into working groups and working people. And companies start to rally around it, and then it takes off. So, we provide a collaboration platform, we provide tools to collaborate, and then we bring people to the table, and then just, we don’t want to stand in that way, in the way of their creativity. We just kind of let them run. We do ask that they follow our mission, that they do make it relevant to our community, right. And so each project does have a charter that they go by. And at the end of the day, the four tenets and the charter is really what helps us navigate and move things forward.

Steve Helvie:
We have some really bright members within the Open Compute Project. And I think, on the community side, and I’ll quote, one of our members now who’s leading one of our projects, Ron Minich. And he was speaking about his particular project and the power of open source and the acceleration that you had mentioned is he just said that no one person is smarter than the rest of the world. And with that model, that’s why he believes that open source will always win. And I was listening to one of your other podcasts on the gentleman from Zappos. Now, he wasn’t necessarily speaking about open source organizations, but he was talking about a network of connections internally, which is essentially what this is, it’s a network of connections. And there’s many connections that happen within Open Compute, that really may not have anything to do with OCP. But it helps accelerate other lines of business that these companies have. Because of course, there’s not one company that does only OCP, they have multiple lines of business. So, it’s about moving the entire industry forward, and then slowly nudging everybody over to open source as fast as we can.

Archna Haylock:
We’re not the only community out there, right. There are so many wonderful communities out there like Open Networking Foundation, or TIP, Telecom Infra Project. There’s the Linux Foundation. So, there’s a lot of companies that are kind of working towards the same open-source goal. And they’re standards bodies out there too, right, ASHRAE is one, IEEE is another one. So, we’re kind of finding ourselves in this unique quorum where we’re not a standards body, but we don’t also want to reinvent the wheel. Right? So, if someone else is working on something, we want to be complimentary. And we want to be able to enable them and then, in turn, enable us. Right? So, it’s a really nice, I think, co-op-petition/friendly common goals, right, that we try to work together. And I think it makes it easier for everybody. And at the end the suppliers win, the consumer wins, the adopter wins, right? Anybody that supports that from software to hardware vendors, it’s a win-win for all.

Stina Heikkila:
Incredible, because this sounds like to anyone who wants to start an innovation community, this is just the dream come true. Everything is sort of happening in a self-organized manner. And you take sort of a backseat, and you just enable all this innovation to happen. So, I’m curious to know, what might be some challenges that you faced and like in the relationship between the organization on the one hand — you don’t want to over intervene — but are there instances like when that relationship — are there any challenges that you face, that you have overcome in that process, and any other elements that could be interesting to share in that relationship between the community and the organization?

Archna Haylock:
I can speak from my perspective, and then I’m sure Steve will have his perspective, too. I think that we are still growing. We are still learning. There are things that we had to adapt to. Licensing is a good example. IP, as Steve mentioned, people are very protective of their IP. And OCP is one of the few organizations out there that doesn’t own any IP. So, when someone contributes a specification to OCP, they are making it open to the public. They’re not assigning any IP over to OCP. They’re just making it open to the public. And they’re saying, “Hey, I’m going to put this IP out there, if you use my spec, as it’s written, I will not assert patents. I will allow you to have it royalty-free, you have copyrights royalty-free, and the patent is open to you royalty-free. And it’s a different model than some of our sister organizations out there that work under a RAND license or other licensing. And it’s really hard to get hardware vendors to wrap their heads around contributing something openly without any patents. So, that is a challenge that we constantly come up against. Once they understand it, then it’s open, and you’ll see a lot of contributions come from them. But until they do, it’s a long process to convince their legal teams that this is for the good.

It also complicates things when you have multiple companies making a contribution on one spec. So, our legal cycles sometimes take a long time, just to kind of get everybody to the table that is contributing. And in the interest of transparency, it’s never easy to get very large organizations to come together and agree to contribute. But we’ve managed to do that with companies like Facebook, like Intel, like Microsoft, IBM that have huge patent portfolios, and still have contributed. Or have found a way to contribute and lead. So, I think that that’s one of the challenges that we have. The other challenge is really how not to grow too quickly. It’s very easy to take in all of these projects and want to do everything for everyone.

And the challenge that we face is we’re a foundation of 10 people if you can believe it. There are 10 people that are part of the foundation. But having said that, we are truly standing on the shoulders of giants. I mean, we have 15 people on our strategic IC, our incubation committee, they are the 13 companies that represent thought leadership, right, and they’re helping us take it forward. Our board members, which Steve mentioned, are helping us take things forward. And then 200 plus members, thousands of engineers that are helping us. And so all of our projects are run completely by volunteers.

Simone Cicero:
Yes, that’s a lot of contributions happening. Do you have an idea of how much is the GMV that was mobilized, let’s say through Open Compute?

Steve Helvie:
About the general market size itself, the annual forecasts, we have an analyst that tracks the open compute revenue. And it’s being forecasted now to be close to $12 billion by 2023. Now, the open networking piece alone, so just the networking device is supposed to be right around 1.35 billion. And I’ve been asking some people if they think that’s a conservative forecast or an aggressive forecast. They think it’s going to be relatively conservative to hit that 11 to 12 billion primarily because of the proliferation of devices and the expansion of what a data center actually is, and all of the other types of components that will be open-sourced in the next two years.

Archna Haylock:
Mind you, that 12 billion does not include our board member companies.

Steve Helvie:
Yeah, that’s a good point. Yeah. So, that’s minus — that analyst takes out the numbers from Facebook, Microsoft, etc. — [crosstalk]

Simone Cicero:
Because they self-bill them.

Steve Helvie:
Yeah, it would skew the numbers slightly.

Archna Haylock:
Those are the challenges that I faced, Stina, just from a community perspective — managing the volunteers and the different companies that participate and their corporate cultures, the legal, and growing really fast. But Steve, I don’t know, if you have a different set of challenges that you’re looking at.

Steve Helvie:
The challenge is it’s about expectations when a member joins one of these communities. We’ve run into members that say, “Well, I’m going to join this, and then I’m going to instantly make business contacts with the top-tiered members, and I’m going to start selling them all of my products quite quickly. And so it’s setting expectations and making sure that they understand the type of open source organization that we are. Because there are multiple types of open source organizations all structured differently. And I recommend to the listeners out there, if you get a chance, Mozilla put out a document called the archetypes of open source. And they go through all the different kinds of open source business models. And that’s a good read. It’s not very long, it’s maybe 30–40 pages or something like that. But they describe, say, a multi-vendor approach or a mass-market approach. And so understanding what type of organization you’re getting involved in, to understand the governance piece specifically, is a big, big obstacle with a lot of these customers and members and clients and end-users. So, it’s a lot of education on our part to make sure that those expectations are set well. Because otherwise, they’ll be eight or nine months or even a year into this organization and not realize how to get the most value out of it.

Stina Heikkila:
I’ve worked a lot in the public policy space. And it also sounds a little bit like a government’s dream to be able to generate this level of impact that you’re doing. Have you had any exchanges or collaboration with governments who are working with you or supporting you in some way? Or are you staying away from that?

Steve Helvie:
I can speak to a few of the European organizations that we’re working with. One is a circular economy data center research project being funded by the EU but also based in East London, I believe, is the university that’s driving this. So, they were measuring kind of the embodied energy that goes into manufactured IT. And so we’re donating some hardware there so they can see the difference in the embodied energy between an open-source server and a more proprietary or closed source server. And that research then will be published across Europe. And that’s just in the early stages. So, that’s in phase one of that project.

There’s another, which I’m sure some of your listeners may have heard about, which is Gaia-X. Gaia-X is the big European initiative for regional cloud, and to ensure proper federation, proper policies, and governance around data used within the European Union. And that is a massive, massive undertaking. And we are members of that as well, along with an organization called Open UK. Now, Open UK is an organization that’s started to drive open data, open software, and open hardware across the UK both at the public sector level, and across education and academia, as well and providing a foundation for all open source in the UK. So, Gaia-X is a massive organization. And then we have a lot of members that are members of that. And then we are, OCP itself is a member of that. And we’re also a member of, say the sustainable data center Alliance, which has a government or public policy slant. And that’s based out of the Netherlands as well. So, we’re quite active in the public sector initiatives in Europe.

Archna Haylock:
In the US, not so much. I think we’re more active on the academic side. And therefore, in the laboratories, some of the laboratories that are funded by the government, but not really directly part of the government. I think it comes really difficult sometimes for a government entity to start thinking open, because everything that they do is so proprietary. But they do work with a lot of the academic institutions. And so, we do have like Berkeley Labs, and we do have underwriter laboratories. We have Lawrence Livermore labs that come in and our events and participate in some of our projects, especially around high-performance computing. And Software-Defined storage, and things like that. So, they’re very interested in what’s going to happen in the future. We do a future technology symposium concurrently with our global summit. And that’s really to address future technologies in academics and in laboratories. So, IBM Research participated in that, Facebook and their research wing participate in that. So, we get a lot of students and a lot of academics and sometimes government officials that come in and want to present and talk about their R&D work.

Simone Cicero:
So, the question that I have in mind now is really about the process that starts from one organization wanting to do something like that, like Facebook, in that case, in your case, in another space, not in this space, not becoming part of OCP. I’m talking about doing this initiative, facilitating such an initiative in another industry. And when this conversation comes up with my customers and our adopters, and our open-source adopters, in general, it’s always about how do you deal with competition? How do you deal with the competition also with the ecosystem, not just with your actual competitors? Because maybe you want to standardize a market where you already are because you want to bring more services on top or something like that. And my question is really about how does the process of progressively decentralize the organization from the members that create it, and then hand it over to the community within time. So, how does this process work? What should an organization be ready to let go when starting such initiative, and then creating a shared governance and enabling a community?

Archna Haylock:
That’s a really good question, Simone. I think that for an organization like Facebook, I think it was very easy for them to let go of OCP once they had confidence that OCP could run itself. They had confidence in the leadership that they appointed, and also in the community to take their inputs, and then they would just be the guiding factor. Right. So, if the mission, I think, is strong enough, and pointed off, in some ways, right? I think you’re going to have a community that rallies around it. Then it’s up to the organization to put the right staff in place and the right structure in place. Mind you, it has to be flexible enough to evolve. And I think that that’s what Facebook and our board members have allowed us to do is to evolve from the data center, the traditional hyper scaler data center to evolve into other markets, like Telco and Edge and now retail, and enterprise. There has to be some level of flexibility there. But they’re very active. I mean, Facebook is one of our most active members. And we’re constantly relying on them to come back and help us because they’re the ones that are also helping us identify some of these new markets.

Steve Helvie:
I think what worked well in some organizations is they just look at this and they say, look, is this a differentiator for us? Is this something that really differentiates us from our competition? And if it’s not, then let’s get the community to work on it. And let’s focus on the value-added piece of the puzzle that makes us so good. And I go back to the OCP example. But hardware is not Facebook’s differentiator. And then the other hyper scalers quickly realized that, hey, this is not our differentiator, either. It’s the value-added services on top. The second thing that we’ve seen happen is that when a company just stagnates on growth, they just simply cannot keep up or they cannot have the resourcing financially to keep a project or keep a service going. And open sourcing provides them that opportunity to go get additional engineering resources.

So, all the time on GitHub, people are throwing out software projects to say, all right, let’s throw it out there and see how much interest we get out of it. In the hardware space, it’s very, very different. But having a company sit down and say, “Well, what do I give up? What do I keep? When do I hand it over?” All of those things play into the fact of is it a differentiator? Do I have the resources to maintain it, even if I want to keep it, and at the end of the day, I can probably do a lot better in a community setting than I can buy myself. And you’ve seen that with other things like Kubernetes. Google controls Kubernetes, but it’s a massive open-source organization.

Simone Cicero:
Right. I was resonating your conversation with the experience that Google had with Android, for example. So, it seems like not being interested in that part of the business, it’s a good way to start an ecosystemic alliance, let’s say. Even if Android was tempted a couple of times to make smartphones, but they, I think they had been very, I would say hesitant, let’s say. The interesting question will be for a — And what I get from this conversation, for a company that actually is already maybe selling parts of this infrastructure, so it’s immediately involved within the market, it may be to deliberately choose to go for open source and really accept that that’s a commodity, you need to let it go, you need to develop in the open and try to create this ecosystem of other players that can speed up the development without owning your piece of the industry, but more decided to let it go and then focus on services and other innovations. So, I think that’s interesting. Only if you have any — yeah, that’s it — some ideas on that.

Archna Haylock:
Yeah, I was going to say cooling is a good example, I think people have been doing cooling in some form or fashion for the last 30 years, right. But they haven’t been able to rally around the basic concept of cooling. Everyone’s been doing it proprietarily because they think that that’s where their IP is at. And I find that it hasn’t gotten enough traction. But when you get a bunch of people that are trying to solve the same problem at the table, they can go off and do it independently. But if there’s some baseline established amongst them, that, okay, here’s the baseline that we’re going to operate off of, and then differentiate up at the top. I think that that’s where the power of open sourcing really comes in. I think we’re never going to solve the cooling problem all the way up the stack, and all the way in the rack, and all the way from end-to-end. But we might be able to at least get some common understanding. And if we do that, we’re ahead of the game already. Right? I feel like if we can get some industry experts to come in and say, “okay, we’re all going to rally around these materials, or we’re going to all rally about the way that we’re going to use the heat generated from the data centers, or this is the way we’re going to do immersion”. There’s going to be differentiation no matter what, but I think we need to have some idea of what the premise is.

Simone Cicero:
Super. I mean, that was a crazy conversation, really, I think. I’m going to refer to this a thousand times. And I would like to ask you just to close this conversation by highlighting something that you believe is really important to share with our community knowing that there may be companies interested in pushing same approaches or, in general, people that are genuinely interested in how these evolutions are changing industries. So, your final remark maybe, and maybe one of you can also add a couple of bits on where to find the most interesting news around OCP.

Steve Helvie:
Yeah, thank you, Simone and Stina, for the opportunity to have a discussion like this, really. Anytime we get a chance to share our story we learn something too, just by engaging people like yourself. So, I think the key takeaway is open source is extremely exciting. There’s not one company out there doing the new type of architecture that’s not relying so heavily on open source. And the progression from open source software to open source hardware is a natural fit. Software runs everything. You can commoditize and disaggregate the software from the hardware. And the other thing that I would encourage people to do, is to be thinking about the circular economy. So, when you’re buying a particular server, you need to make sure and you think only from the design phase all the way through to the end of life. And the ability for somebody to take that server, and then get additional use out of it is the new way forward.

So, this sustainability, circular economy, reuse. Europe is already putting regulations in place around embodied energy, the efficiencies that IT needs to run, the reusability, the repairability; all of these regulations are already happening in Europe. So, if you’re a company out there, and you’re starting to think about new ways of doing architecture, new ways of doing infrastructure, you should be planning for sustainability upfront to meet these requirements. And you may say, “Ah, it’s not a big deal. I’ll wait for my company to really care about that stuff.” If you can be at the forefront of this, you are going to be leaps ahead of everybody, competitive wise, because you will have already baked these processes in when you move to an open infrastructure.

Archna Haylock:
Yeah, I was going to echo Steve’s remarks here. Thank you so much Simone and Stina, for having us here and giving us an opportunity to tell you a little bit about OCP. You can find us at www.OpenCompute.org. Everything on our website is open. We have a contribution database that has specs that you can download. You don’t need to be a member, you don’t need to have access or passwords or anything, it’s there for you. You can look at every single one of our projects. Each project has a mailing list and a Wiki, you can subscribe, free of charge. You can listen to past recordings, you can participate in the calls. We encourage everyone to listen and join. And if you find value, and you’re ready to contribute, then become a member.

So, not the other way around. Don’t become a member first and then decide what you want to do. We really want you to come in, see how the community works, get your feet wet, explore marketplace, explore how we do things, understand the licensing. And when you’re ready to contribute, become a member. And Steve and I will be here to guide you. It’s only 10 of us, but we respond fairly fast. And take a look. We’re a global community. So, if you happen to be in other parts of the world if you happen to be in mainland China, if you happen to be in Japan, if you happen to be in Korea, or Europe, they’re starting their own communities, and that’s also on our website. So, please do check us out. And if you have any questions, reach out to Steve or myself.

Steve Helvie:
That’s lurk and learn I think is — lurk and learn, yeah!

Archna Haylock:
Yeah, yeah.

Simone Cicero:
Right. Right. Thanks so much. It was such an inspiring conversation. I’m sure our listeners will be excited about this.

Stina Heikkila:
Thank you so much. I am stunned by the project.

Simone Cicero:
Super. You have two new fans. Actually, I was a fan already. So, thank you very much. Thank you very much. And to our listeners, catch up soon.