And then some really good enterprise wins as they’re looking at really enterprise-level training as well as inference and how to really unlock the full power of the AI ecosystem. And so — and we think the NVIDIA DGX private cloud managed service is a really distinctive offering. And we’re seeing big pipeline build there with NVIDIA on that front. And so — so I do think we have a very broad range of where I think we can offer that around the world and we’ll continue to expand that over time. But again, I think a hugely exciting opportunity. In terms of — you asked about the TAM, I mean as I said at the Analyst Day, I think the TAM is huge. And so I think it’s probably bigger than what we’ve set out there. I think when you look at the possible impacts and kind of what we’re seeing in terms of the early returns on AI, I think you’re going to continue to see a lot of investment flow to that.
And so I think the TAM is probably bigger than what we outlined. I think the key for us is really where can we be distinctively differentiated in that. Yes, I think we’re going to get a piece on the xScale side. But I think the more differentiated position for us over the long term is unlocking the power of the AI ecosystem through this sort of cloud adjacent set of offerings. And on the digital services side, our cloud adjacent storage and Fabric Cloud Router all sort of hitting in that sweet spot of what we think customers are really looking for — control over their enterprise data ability to access AI tools from the hyperscalers who are innovating rapidly in that area, stitch it all together and make it work in a way that makes sense for them.
And so Fabric Cloud Router, fabric, cloud-adjacent storage, all things that really play into that. So we continue to be very optimistic about that. But I would say tempering expectations, I think it’s going to — it’s going to take a little time for that to really fully realize itself in terms of the bookings flow.
Operator: Our next caller is John Atkin with RBC.
Jonathan Atkin: On the churn commentary, is there anything to call out in terms of reading where you saw it or which products? Was it mainly cabinets or cross connects or other?
Charles Meyers: Yes. I would say more on cabinets and power. The cross-connect turn is looking a lot like it has for the last several quarters, John. Growth activity continues to be strong. I think we’re seeing some grooming, particularly in the network service provider segment as their businesses are a bit more challenged and I think they’re really focused on cost reduction. We are seeing some consolidation into higher speeds. So that’s a bit of a bit of a headwind. But I think the more, the more, the elevation was a bit more on the cabinets and power and CapEx side. But it really is related primarily, I think, to people resizing footprints in a way that is — aligns to what their more immediate need is because I think that it’s, we have — I think there was a time there when people were saying, “Hey, I have more than I need but I’m just going to hang on to it.” And I think we — that was the case in ’22.
But in ’23, we’ve seen people a lot more pressured by budget. Part of that, we think, is actually related to — you, guys, asked us a lot of questions when we did the PPI around would that create elasticity. And we haven’t seen what I would consider traditional elasticity of demand but what we have sort of heard coming from our sales teams is a pressure that says, “Hey, I ate up all my budget with the PPI. And so I can’t grow as I expected. And so if I want to do some of the things, I’m looking to do on the AI front, I got to find room.” And so they’ve been more typically contracting footprints. And so that’s really the dynamic we’re seeing. Let me give you a little more color on a couple of areas. One, only a single-digit percentage of our churn is full customer churn.
So almost all — the rest of it is all people moving around resizing footprints, that kind of activity. And quite encouragingly, I said, well, let’s look at those customers and those that are churning and tell me what their other — what their activity level is across the rest of the platform. And quite encouraging for the most part, you’re finding those customers are buying elsewhere in tandem with the optimization work that they’re doing. And so I think that’s really the dynamic there. In terms of, I would say we’ve seen a little more of that in Europe, John. And that’s probably because we were — had a little large footprint population there. So I think we’re seeing it a little heavier there. But again, our guide says sort of assumes that we’re going to continue to see some of this through the first half of this year with attenuation of that in the back half of ’24.
Jonathan Atkin: Got it. And then secondly, I was curious about the xScale initiative and the growth path in the Americas and kind of the puts and takes of pursuing that organically versus inorganically?
Charles Meyers: Yes. I mean I think we’re very focused right now on organic. We would certainly not be — we wouldn’t necessarily not be open to inorganic. I just think it’s a tougher thing in terms of identifying those assets. I think the multiples at which those things are trading are pretty heady, to say the least and plenty of competition for those assets. And so I think we’re primarily focused on organic. And — but again, if the circumstances and conditions change, one, our balance sheet is always ready. And I think we’d be open to that. But I think our focus is probably more so on the organic side.
Operator: And our next call is Michael Rollins with Citi.
Michael Rollins: Just first following up on the point that you’re making about customer optimization. I think in the past, you’ve used the analogy of managing the retail data centers is like a tetra sport [ph] of sitting different pieces and deployments together. As there’s some optimization, can you share your opportunity to resell any space or power capacity that you get back and how that plays into the dynamic for 2024? And then just secondly, just curious if you could unpack the constant currency organic growth range, xPower [ph] of 7% to 8% in terms of what stabilized growth would be — and then within stabilized, how to think about the price ARPU component relative to the volume component.
Charles Meyers: Okay. Maybe Keith can jump in here on the second part of that, too and we’ll tag team it a bit. But let me catch your first one first. Yes, you’re absolutely right. We’ve long for many years, talked about our business as a bit of a tetris game [ph] in terms of figuring out how to get optimal returns from our capacity. And I would tell you that I think that given the increasing price environment, given the tendency for our churn to be a bit biased towards the large footprint side of things, we generally see churn as value accretive over time. That doesn’t mean we want it all. It’s — but it’s — sometimes, we need it and we actually may sort of work to get it to happen. And that’s, as I said, that happens sometimes in Singapore in the markets like that.
But there is inherently a trade-off between growth rates and return on invested capital. And what we’re seeing is that even in high-demand markets, there is a vacancy drag. And what I mean by that is the time frame that it takes to really fully replace churn with new revenue and that’s particularly true when you’re replacing a single large foot implementation with a large number of smaller deals, we’re seeing probably a little longer vacancy drag than what we maybe would have thought. And so while I think that the — that kind of positive mark-to-market opportunity exists and improving our business mix has always been central to our ability to deliver increasing MRR per cab and return on investment sort of stabilized asset performance and importantly, FFO per share, it is sometimes a revenue headwind for us.
So we do see that on the churn but I think there are positive aspects to it as well. As to the 7% to 8%, look, if you look at stabilized assets, absent the PPI, they’re in that 5% range. And — and so — but that includes selling interconnection into them. It’s probably not a ton of additional volume growth. They’re operating at reasonably high levels of utilization. So I do think you’re going to see some positive price on mark-to-market. And I think you’re going to see continued interconnection sort of probably more, in our more traditional range. And then the rest of that, that growth is going to have come from the broader footprint, including our non-stabilized assets which are probably growing at a slightly higher rate. Anything to add further on that?
Keith Taylor: Yes. And Michael, let me just add maybe just a few other quick points. We’ve always said that we think stabilized assets can grow 3% to 5% on a sort of constant currency and normalized basis. And this quarter, we’re at that range 9% with the power price increases. What are you going to feel — what you’re seeing this year being 2024, there’s a couple of things. So we’ve neutralized currency, we’ve neutralized for all intents and purposes, the power price decreases. Again, that’s going to have a roughly 30 basis point impact, I mean the power price, it will impact sort of the growth rate a little bit there. And so where we’re really focusing is really the timing. Okay to somebody [ph], so what we’re really focusing on is the timing.
And so as Charles alluded to, had a little bit higher churn as we entered the — exited the year. We have a little bit more higher churn at the front end of the year. And so when you look to the back end of the year, you actually get a much more attractive growth rate than what you start the year at. And so what it blends itself out, basically, you’ve got a 7% to 8% growth rate. But overall, when you sort of the business in and of itself, extremely strong pipeline. We’re taking into consideration what we think is the timing delays in — and although a reasonable book-to-bill interval, we still think that the just how — the speed at which things are converting from the pipeline into a billable event, that’s just taking longer. And then the other thing, I would just say is nonrecurring revenues for all intents and purposes, it’s going to be roughly flat year-over-year.
It’s going to move around quarter-to-quarter, as we’ve talked about. Q4 was very rich, Q1 is still pretty darn good because we closed 2 large assets in xScale space in January. And so we will get — there be some fees that. But I think what’s most important is understanding that the richness of the pipeline, the timing of the year and what we envision that we’ll exit 2024 with, is what gives us the confidence that we can continue to drive the value into the main FFO per share number and give you the growth.
Charles Meyers: Well, Mike, I’d say that the short story on it is that 7% to 8%, I think, is 3% to 5% is the way to think about the stabilized assets. And the balance of that is going to really need to come from the broader portfolio which is probably going to have less mark-to-market juice. I think the 3% to 5% has to come in part from some juice in the mark-to-market opportunity that we have in the stabilized assets. Because those are the ones that are going to be rolling through. You probably have a little less than that in the non-stabilized portfolio because they are newer — in newer contracts with probably less of a gap there. And then we’re just going to have to continue to drive the volume on the gross bookings side.
Operator: Our next caller is David Barden with Bank of America.
David Barden: Two, if I could, just real quick. Charles, we’ve been talking about the hybrid private public cloud infrastructure for the longest time. You brought up a new term that I hadn’t heard before, the private AI. And I wondered if you could maybe elaborate a little bit how that compares, contrasts or doesn’t to our understanding of hybrid private public cloud. What is that, the private AI architecture look like as far as Equinix concerned? And then second, Keith. Last quarter, we talked a lot about cabinet, 8 cab [ph] per cab, consumption and how that’s evolving and the potential to bring a new number to the forefront which would be something like a cabinet equivalent billing number. Could you kind of elaborate a little bit on how that looked like in the fourth quarter and where we are in evolving that disclosure.
