In The Arena by TechArena - How RackRenew and AMI Are Tackling the AI E-Waste Crisis

Episode Date: July 17, 2026

The AI infrastructure buildout could generate up to 5 million additional tons of e-waste by 2030. In this episode, sponsored by AMI, host Allyson Klein sits down with Patrick Bliemer, Global Sales Manager at RackRenew, to explore how remanufactured OCP hardware and open-source firmware are offering a sustainable, enterprise-ready alternative to new equipment.

Welcome to Tech Arena, featuring authentic discussions between tech's leading innovators and our host, Alison Klein. Now, let's step into the arena. Welcome in the arena. My name's Allison Klein. Today, I am so happy to be joined by Patrick Glemer, Global Sales Manager at Rack Renew. Welcome, Patrick. How are you doing? Thank you, Allison. I'm good. How are you? I'm delightful and very happy to be here today. Rack Renew is one of my favorite stories. in the industry. So we get to explore it a little bit on today's podcast. But why don't you just start, this is the first time having you on the show. Tell us a little bit about you. And for those

listeners who haven't heard about Rack Renew, why don't you go ahead and introduce the company? Yeah, sure. First introducing myself then, Patrick Blimer, I'm a 30-year veteran within the high-tech industry. I spend a lot of my career working for one of the largest semiconductor companies, but also actually I've worked at a D-Tech startup, so pretty much anything in between, but always in sales and management positions related to either data center or PC market. When explaining Rec Renew, I always start off with the parent company with Sims Lifecycle Services, we are subsidiary of them, because that's a company, it's a global company, well known for decommissioning and reverse logistics from hyperscale customers.

And it is obviously that opportunity, having them doing the decommission at scale, that gave us the opportunity to start rec renew and say, is there a better way to reuse the equipment that is be decommissioned through a re-manufacturing process and offering that to the border market. And the unique aspect of it is that it's OCP equipment. And so far, you can find on the market a lot of refurbished IT equipment, but there's really no party. that does remanufactured OCP equipment. So we really thought this was a great opportunity for us to jump into. AI is driving an unprecedented demand for compute power. Can you explain the impact of the embodied carbon savings when a company chooses a rack-renewed solution?

Yeah, sure. I think as a society, we all ought to be worried with the incredible investments that are being made right now in additional capacity or AI. And I just read an article last week where it was suggested that up until 2030, there's going to be an additional 2 to 5 million tons of e-waste hitting the market simply because the expansion acceleration in IAA data centers and IA factories. So obviously, when you think of that and when you think about the clouds,

contrary to popular belief, over 70% of the cloud is made up by iron and plastic. And you can imagine to go and manufacture all these components that takes energy, takes water, and it has a very significant toll on embodied carbon. The opportunity for rec renew, when we do the remanufacturing and bring it this to the market again, that embodied carbon already is being written off. So we can basically offer it either at zero embedded carbon or when there's still going to be some new components that need to be added to the equipment, either because the customer specification or something broke down that we need to replace, then it has very low embodied car.

So this is really the right way forward. And I think as an industry, we all need to be concerned that there's more of a negative perception in the market. People feel that there's too many data centers. People feel that it's consuming their electricity and it's taking their water. So I think overall, as an industry, we need to really look at opportunities that we can prove back that we're actually very responsibly at incapacity and we're finding second life for dispossioned material once it basically is being retired.

Now, we know that unlike traditional vendors who are often hit by component shortages and the media is rife with component shortage issues right now in the data center arena, you took a different approach. You harvest existing hardware. Is speed to market becoming your biggest competitive advantage even more than sustainability? I wouldn't say it's bigger than sustainability. I think there's a number of differentiators that we have. And time to market is definitely one of them. But I wouldn't put one over the other, to be honest.

It's absolutely true that we can deliver products within weeks rather than within months or sometimes even you're hearing the E-Times now that may be somewhere between six and 12 months, so getting to a year plus. So yeah, there is definitely the time to market advantage. But I must say that obviously it's easier when we're working with prospects that already are familiar with OCP or have OCP ready data centers because if they have not, then there may be some alterations required in their infrastructure to go and actually accommodate OCP brands. By and large, we can definitely provide equipment to the market much faster than other parties.

Now, I know that this takes some work to get this hardware ready for enterprise deployment. And I guess one question that I have for you is you recently discussed a collaboration with AMI. How has that been instrumental in liberating this hardware from proprietary configurations found in large-scale, hyper-scaler data centers, and making it enterprise ready? Yeah, I think the challenge for us was that getting that hardware remanufactured, that took a bit work in terms of processes, we got that done. But just the hardware alone isn't sufficient, because there is this element of all of these components. They have a BMC chip based on Open BMC, and you need that for system management. And when we do the decommissioning from the large

hypers, even it starts with Open BMC that they use. They have added additional features encode into that chip, which makes it more proprietary. And they do not want that chip then to become available again for second use. So what Sims does is actually we grind up the BNC chips entirely. It goes to dust, literally dust. And without these chips, yeah, the hardware might be remanufactured and again, as new or better than new, but you can still not actually integrate that equipment without the BNC. So we were looking for a partner that could actually help us with that. And we found a great collaboration, a strategic partner in AMI was one of the leading names in this space, and we started to work with them to basically

develop a new open standards, of course, BMT chip, which is going to be more of like generic product, but also when we have customers that require specific or bespoke feature sets and changes to the BMC, they can actually also do that for us. So it goes both ways. They can develop the standard product and they can do the customization. And that capability is absolutely, what we needed to complete the remanufacturing loop from hardware and software to go and bring it to the market. Now, security is often the biggest concern with refurbished hardware, and that's understandable. How did secure boot technologies integrate into your manufacturing process to guarantee a trusted

foundation for customers? Yeah, good one. That's all related again to the firmware. And for many companies, I would say, of course, for all companies, security is the number one concern. So this also is something that we had to go work with a partner like AMI. But at the same time, you need to also work with the OEM vendors, the original equipment manufacturers.

And we do not have all those relationships, but AMI does. So it is a combination in working with AMI and the OEM equipment manufacturers to go and get to the updates required in the firmware to make them again up to date for today's use. And you can see that, depending on how. old the equipment is that gets dispositioned, some of it actually may be already beyond an end of life or end of service kind of cycle. That could be a challenge because if the OEM doesn't provide the updates anymore, let's say at a CPU level, then you can make still changes in the firmware, but you're not going to address actually that there's no security update anymore

at the CPU specific levels. It is a little bit of like it depends how all the hardware is, but if it's still serviceable with a partnership AMI, we can actually get that all in the stack and we get to that route of trust or trust foundation, get that established and basically pass any sort of requirement that a potential customer may have. Now, in 2026, why is having a transparent, open source aligned firmware stack

no longer just a nice to have, but a critical requirement for a scalable data center? I think especially OSIC, everything is open standards, open BMC. So all of the bonus that we have in our portfolio have a BMC chip. And there is no other way around it. For companies, for data centers to run their infrastructure efficiently, but also in terms of energy use, in terms of monitoring,

in terms of preventive maintenance, really the BMC is the critical piece. And it doesn't only work, of course, when the systems are powered on, which in the data center is pretty much all the time. But even when it goes offline, you have this out-of-band support that you can provide through the BMC. For us, it is really a critical element. And I would even say, if you cannot offer it, you don't have a proposition in the market anymore.

Because data centers and equipment is standardized on Open BMC at large, especially in the OCP community. As you're delivering high-density solutions like your 11.5 petabyte racks, what are the challenging to having it seamlessly integrated into existing infrastructures that might exist within your client environments. Yeah, I start smiling because it's a bit of a cheap shot, but I think the biggest challenge right now is actually sourcing the media at affordable prices because we can happily sell the hardware, but finding the drives at a reasonable cost these days to go and get to 11.5

petabytes of storage is more of a challenge. But it's a different topic. But yeah, that density, yeah, I think, It's again all the matter that those systems also have one or two BMC chips in them. They will communicate either with the head node or with the internal controller. For seamless integration, you also need to understand all the other settings and configurations in the data center. How is it being done from an architectural point of view?

What type of software data management kind of software are they running? How is the networking provisioned? So it's not just like, I don't think honestly that a seamless, integration exists. There's always going to be adaptations that need to be made, but provided that the customers or the prospects are willing to work with us and they give us the opportunity to go and do an evaluation test or something like that, we can actually work through those together with them

and make sure that we adapt the hardware and the software in such a way that they can integrate it as easy as possible. But seamless is, I think, potentially an over-proness. So there's always going to be some additional work required. Now, I know that you've had a front seat experience on deploying standardized open BMC across these meta subracks. Can you tell a little bit about that? In your goal of ensuring that a customer gets the same experience they would expect from a traditional OEM? I think the best complement probably we got was for one of our very large prospects,

that they were well impressed, that a small, call it a startup, as Rec Renew has a collaboration with a party like AMI. That has given us not just the confidence internally that we're working with the right party, but also it gives you that credibility in the market that you're working with one of the global leaders in this space. For us, this has been a massive step forward as well, and it helped us to go and find the missing link on the BMC, the open BMC side, on the firmware side. I think there's opportunity in future also in terms of BIOS, but I think it was especially important that customers may not expect reg renew given its size

and that we're relatively short in business, right? It's getting to two years now since we started that we actually have partnerships like that. So it helped us a lot to go and further our conversations and evaluations with those accounts because we actually have that partnership. Now, some would argue that the complexity of managing this open ecosystem is driving up TCO instead of lowering it, What's your response to that? I think it's a moment in time. Over the long run, open standards will always drive faster innovation and will always lead to lower cost.

If you start to go and look within that window, you may find that there are certain hurdles that need to be cleared, that there are certain technological challenges that need to be addressed that need to be addressed before you get to where you want to get to. But I still firmly believe that this is the right path. And we can already see, like also with the OCP community, there's so much collaboration, there's so much contribution to this that ultimately it's not going to go and lead to a higher TCO. I'm pretty unconvinced about that. But we need a bit of time to clear all the hurdles. Sustainability isn't about reusing hardware. It's about efficient operation.

How do the optimations at the firmware level help wrap renew hardware meet the strict energy efficiency targets required by modern European or even global data centers. Yeah. Yeah, the firmware is kind of the brains of the system and what the capabilities that provide, of course, in terms of workload management, load balancing, real-time energy adaption, all these things that basically the firmware enables, they help to make the system as efficient as possible. Also, from a compliance point of view, the firmware can,

actually do the audits and reduce all the data that is required to see that all those items are being checked off in line with whatever policy or requirements has been stipulated. So we believe that, again, the firmware is key. And having open firmware on our devices is basically going to go enable us to go and get, not just into small customers, also into the larger cloud service providers and data and their customers, for example. Patrick, one of the things I wanted to ask you about, the circular economy advantages of remanufactured hardware is our own question. But the hyperscale of hardware that's coming through decommission

and into a rack-renew opportunity is already one or two generations behind in terms of power efficiency per watt. How is that impacting sustainability and how do you work with customers to understand that broader landscape? No, and it's true. And some even is N-minus three, so three generations behind. And that is because the innovation cycles are accelerating and because of that, refreshes are taking place much quicker. So the effective use of equipment when it's leading edge gets shorter

so larger volumes need to get decommissioned faster. That is all true. We also say, ask yourself the question, what workloads do you have that really require leading edge? And of course, we all know that when it comes to IA, agentic IA, other like you need to be a leading edge. But there's more traditional workloads still running within the same organizations, within the same data centers that do not need to run on the latest and greatest.

And that's, I think, where the sweet spot is for Redmond. We can actually make that infrastructure available at much lower cost to your earlier question faster than you get it ordered new. And those workloads are perfectly, if it's not time critical, latest, critical, you can actually run it perfectly. And we have great examples in the space of virtual private servers, Kubernetes, email servers. There's a lot of use cases still that do not need to go on very expensive cutting edge technology. And that's where we're shooting for. And if everybody would ask that question and comes up with maybe two or three workloads in their company,

that could be subject to remanufactured hardware, we'd love to take that challenge on and go into an evaluation to go and see if, okay, let's go and try that out and see if we can make that work. And so far, we'd be pretty successful. Now, as we look ahead, we talk about a circular cloud. What is your vision for the next three to five years and how that's going to be playing out? Yeah. So, as I said before, the sheer volume of equipment that will get decommissioned over the next couple of years is going to be huge.

So, you know, we are, of course, very privileged the fact that's, since Lifecycle Services is such an established name in hyperscale decommissioning. So they will play a big role to go and do that as responsible as possible. And it will provide us with more and more opportunity to keep adding new, newer generations of equipment to our portfolio and also move up in terms of capability. We now, we offer compute nodes, compute servers. We have storage solutions now as well added. and we're looking to go and introduce AI GPU servers in the foreseeable future as well.

And we're going to be able to continue to go and add to that portfolio because we know more and more equipment is going to be dispossioned. And it would be a crime, honestly, to see that go to the waste aisle, so to say, as there's still much life left in that equipment. Patrick, it's been wonderful talking to you. You provided some great insight, not just in what Rack and Renew is delivering to the market, but how certain things like firmware are unlocking the potential for re-manufactured hardware. Fantastic insights today. Thank you so much for being on the program. Where can folks reach out to you and the team to learn more about Rackerynew solutions and how you're engaging in the market? Our website, www.orgrenue.com, obviously, and we have a LinkedIn page too. All right. Thank you so much for being on the program with us today. It's been a real pleasure.

All right. Thanks, Alison. Subscribe and engage at our website, Techorina.aI. All content is copyright by Tech Arena.