Keeping the cloud and planet online

Show ID (00:04):
This is Power Onward, the podcast that unveils how power technology shapes your everyday life. You are listening to Power Award. What is power in your world On This is Power Onward, the podcast that unveils how power technology shapes your everyday life.

Kim (00:34):
We all know what it's like to ask a quick question online. What's the weather forecast for today? Where's my package? Can you edit this photo? On the surface? These moments seem small, but as you're learning behind every search stream and smart device is a growing demand that's putting pressure on our power grids. We're talking about data centers. These facilities power the AI that's changing how we live, work and play. But as they grow, they face a big challenge. So the question is how do we keep AI online without turning our planet offline? Hi, I am Kim.

Bhargava (01:09):
Hi, I'm Bhargava.

Kim (01:11):
Bhargava. I honestly can't believe how rapidly AI has been evolving and how it continues to evolve. Growing up in the nineties, AI stood for Alan Iversson. He was a point guard in the NBA. Now when you hear ai, you're thinking about all the different ways that data centers could be strained from this. So let's start with this. How does Cummins support data centers all around the world?

Bhargava (01:31):
Welcome, Kim. That's a great place to start, right? Because data centers are really the heartbeat of the digital economy. Every search you make, every show you stream, every AI cushion you ask runs through them. They're not just the warehouse full of servers, right? They are the digital engines that keep our modern lives moving. Welcome to Cummins. I think we make a big contribution here. We make sure that these digital engines never misses a beat with our global expertise and technologies, from advanced generators to batteries and to next gen solutions, we keep these facilities always on. And the best part, every what we deliver comes with confidence, giving customers the reliability they need today and the sustainability they demand tomorrow.

Kim (02:16):
And you mentioned Watts and we measure data center capacity in gigawatts, which is way larger of a unit than I could ever fathom, and it's because we have over 36 plus gigawatts installed globally. So I did some research and I was like, okay, what does one gigawatt look like? And this is what my search came up with. One gigawatt is about the same as 1.3 million horses. And for those car enthusiasts, that's 2000 Corvettes Z oh sixes. So when it comes to 36 gigawatts, that's 72,000 of those Corvettes. What I'm trying to get to is that people may not realize the scale of energy demand required by data centers. So can you help put this into perspective?

Bhargava (02:57):
Most people don't think the scale of energy behind every search, every stream, but when you zoom out, the numbers are staggering. For example, like Hyperscaler, some of these big cloud providers, they consume as much as energy in a year as a hundred of thousands of phones. Instead of seeing this as a challenge, industry sees this as a driver of innovation. Think about it's building a backbone for the digital world, enabling millions of people to stream, to connect and to tap into AI every day. That incredible demand is why it utilities are asking data center operators to bring their own power. It is why it's very critical for some of these data center provides to be innovative in nature. How do they support these challenging energy demand? And Cummins is at the center of the innovation and it's also part of the solution, right? We provide backup power and we also providing the option for customers to onsite power where they ensure that these data center hubs stay reliable, resilient, and increasingly sustainable no matter how fast the digital growth accelerates.

Kim (04:08):
So many of us have tapped into AI for various things, helping us create spreadsheets, putting together meal plans, and even tips on honing our skills as individuals. Why is AI driving such a surge in data center growth?

Bhargava (04:19):
Yeah, I think before we go to ai, I would probably start with cloud computing, right? The last big wave of growth, it was mainly about storage and access, right? You upload photo or you stream a video or you run any app. These workloads are heavy, but they're bursty. And it also depends on how many people are online and the systems and other side center sites or facilities they could scale to match that. When it comes to ai, it's a completely different beast, right? Training a large modal means huge clusters of servers running full tilt 24 by seven for weeks or even months. And once these modals are live, every prompt or every chat bot response or even every AI powered search set thousands of processes to work at once. Unlike streaming and storage, these jobs don't taper off the constant and their Palo, and that's why AI is driving such rapid data growth. And it's not just about how many users are online, it's about the intensity and always on nature of workloads. If you see today, majority of data centers are built for cloud and were not designed for this level of compute power demand. And now if you watch also some of the news that coming out, like a lot of these new facilities are coming in, new architectures have been built in, there's new energy solution every day that's been explored in short cloud scaled with people, but AI scales with machine and machine never sleeps.

Kim (06:00):
That's a very, very great way to position that in season one. I kind of covered the truth behind the cloud, just showcasing that the cloud is not something overhead. However, it's these brick and mortar locations filled with servers that help compute these things. And how you mentioned ai, it just never sleeps. It's like New York, but on steroids and it keeps going and going and the complexities continue to grow and build up. So I always hear with AI, the term grid constraints. So what does that mean for the power grader? What does that mean for everyday people?

Bhargava (06:29):
It's a very great cushion, Kim. I mean it, it's something that we hear every day. If you're in data center space, this topic, you hear that from everyone who is connected to the data centers in reality, right? I mean, you might feel like grid constraints sound abstract, but not every grid in North America or even some parts of the other world are constrained. In fact, there are pockets of unused energy capacity that industries actually exploring and finding ways to put those energy into work. Sometimes they're relocating the projects where the power is more available. In other cases where there's already a lot of clusters of data centers are available in those areas, the grid wasn't built for today's explosion of AI and cloud and nonstop streaming. So capacity can get tight at peak times. The exciting thing is how industries responding. Data centers are stepping up to be part of the solution, not just drawing energy but helping to balance it.

(07:29):
They're investing in new technologies, supporting grid upgrades, bringing their own power online. I mean, you might hear news about small modular reactors, right? These are being explored or even some of these bigger hyperscalers are investing in this primarily because they're trying to build capacity to these grids. So trying to help the grids to have gained more capacity so they can provide power and not just s other technologies being explored. And in my view, I think it's going in the right space and what we see today versus what you're going to see in the next five years, it's going to be completely different. And good thing is right, everyone is pushing to innovate, to use energy smart, and even expand the current grid capacity in a simple way for everyday people. That means your AI tools, your video calls, or your favorite shows that you watch or even your digital transactions stays seamless even as demand keeps growing.

Kim (08:26):
A few things came to mind as I was listening to you. One, there's constant construction in Indiana right now. I feel like there's two things about Indiana, sorry, I'm a little bias homegrown here. One, a lot of highways are actually being reconstructed and consistently being developed because what they projected traffic would be on a daily basis, it's been far exceeded. So now we're seeing this growth, and that reminds me of how the data centers are having to reach for more. The other thing about Indiana is I heard there's a larger data center actually being built in northern Indiana, and you talk about those pockets of untapped energy and untapped power. Is that why we see a lot of data centers being built in the middle of nowhere?

Bhargava (09:06):
Yeah, I think that is a reality, right? I mean, as I said earlier, right? There is actually university study was done recently that actually shown that the energy constraint is not everywhere. It just, if you see some of the pockets like Virginia, maybe Texas, Arizona, they are the way you see these energy grid constraints. But if you go Indiana, you go some of the other parts of the world in North America, you don't see that. And that's why you see these new projects that are coming in. And the primary goal of these projects is to use untapped energy. And again, as I brought earlier, the innovation, it starts from not just bringing new technologies is about how you be smart in using existing waste off being efficient. And that's what this innovation, again, it's telling us how to be smart, how to use our operations better, and then even data centers are staying top of it in promoting this in the right way. And for Cummins, we are part of the solution as well. We are helping customers in areas where the grid is constrained with giving them options, how to use their assets more efficiently, even running longer hours if needed, maybe giving other types of technologies and help them to be more sustainable and even use energy in more appropriate ways.

Kim (10:25):
I'm glad that you brought up sustainability because I see that data centers are always under pressure to be both reliable and sustainable. So I frequently hear you saying that these are two sides of the same challenge. Can you elaborate more on that?

Bhargava (10:37):
Greatly said Kim, right, exactly. You can't separate these two availability and sustainability, right? Availability keeps distill world running, whereas sustainability ensures its keep running for the long haul with clean power for data centers, it's not a choice. The future demands both. That's why the industry is focused on solutions that deliver uptime 24 by seven while cutting the carbon down along the way. So it's not that we don't have solutions today that there are a lot of solutions are available today. I mean, you talk about HVO fuels that cuts carbon footprint about 90% at scope one emissions, and a lot of our products are already approved to use HVO fuel and enable customers to reduce their scope on emissions. And there is other ways that we are looking into bringing more efficiency in engines and help the customers to use the cleaner engines. In addition to that, there are other technologies that are coming in, right? Battery energy storage systems that could provide hours of backup power in grid constraint areas and some of the other technologies like fuel cells, smart module reactors, and even advanced storage systems. All these are designed for the two mindsets, right? One, how do you provide reliability at the same time? How do you provide clean power? That solves a both problem Together, they form the mix that proves every what can be both reliable and lower. Carbon and reliability sustainability are on trade-offs. They are partners moving towards together.

Kim (12:13):
That's a really common misconception. I do wonder if people think that reliability and sustainability are trade-offs. What are some other misconceptions about sustainable power that Cummins is working to prove wrong or already proving wrong?

Bhargava (12:25):
And one of the most myths that I hear is sustainability power means you're giving up something. Sometimes it's variability or sometimes it's cost or even both. At Cummins, we have proven that that's simply not true. Our generators already run on renewable fuels like HVO, cutting carbon up to 90% without compromising any of the performance needs. And our tier four products and natural gas solutions give the customer reliable options with low emissions. And even when the grids isn't available, we are actually looking at new technologies like battery energy storage systems, fuel cells where customers could run, especially in grid constrained areas. Customers could run these units longer hours so that they don't have to rely on the grid. They can use an onsite power to provide support to their demand. Again, the point is with Cummins, sustainability doesn't mean compromise. It means progress and reliability today, cleaner tomorrow and always moving forward.

Kim (13:24):
So you've mentioned tier one, you've mentioned tier four. Can you break down a little bit more of what this means?

Bhargava (13:30):
Yeah, I think tier four is the EPA terminology for emissions. Here. What it means is you need to produce your pollutants that are coming from the in engine out at a lower level to make sure that the environment remains clean. I diesel engines are getting cleaner. I mean you look at 20 years back, 30 years back compared to engines that are being manufactured today are very clean. And there is solutions that enable these engines to meet some of these E Ps standards. And I also mentioned about the best battery energy systems. Again, the purpose of best is some of the locations where grids are constrained, which we talked about that earlier. When customer is not able to get the power from the grid during peak conditions, maybe during summer, everybody using everybody's streaming, everybody is doing AI bot or used AI bot, whatever it is, right?

(14:25):
When everything happens at peak, the grid would say that I can't provide the power at this junction and data centers, we need to figure out your load and that's where it call has onsite power and best place a major role in it, right? It could provide four hours of power in a day and helps customers to remove their load from the grid and they can run on their own. So these are the two technologies that are shaping and even in addition to that, natural gas plays even role purely because they are accessible. Some of the sites already have access to the natural gas lines, so customers can run these and even completely avoid reliability on a grid. They may even run this at a prime rating and run these natural gas at a longer run hours. So what it brings is customers are using these technologies to enable them to reduce their emissions or to stay reliable and also improve some of the efficiencies, what they're looking for.

Kim (15:22):
And just to reiterate, best battery energy storage systems is one of our premier product offerings right now. It's been very high in popularity. We've got case studies. We also have an episode called Beyond Your Average Household Battery. So please do check that out if you're curious to learn a little bit more about what this means. So we've done a lot of work towards solving some of these misconceptions, but why does the type of work that we do matter to someone streaming a show or even this podcast or like you said, simply asking an AI chat bot questions.

Bhargava (15:50):
Yeah, I mean that's really the heart of it, right? I mean, why does it matter to people out in the industry? Because almost everything we do runs through these data centers. When you stream your I to show or jump onto a video call like this, what we're doing here, or even talking to a family on a video chat or even asking AI, Hey, travel might trip for next summer, all this is possible with the data centers. They're making it possible of all of these searches, all of these videos and even the AI language models have been efficient in delivering some of these different set of needs. And most people don't think about the power behind these, but we do and so do our customers. And Cummins is helping those data centers to stay relatable in those everyday moments while also giving them the tools to lower carbon and operate more sustainable. So the next time your AI responds instantly or you show never buffers, right? There's a whole ecosystem working behind the scenes and Cummins is proud to make sure that it's done in a way that is respecting the planet.

Kim (16:58):
Reliability is such a key aspect of it because I can't, you said buffer and I can't remember the last time one of my shows had to buffer. And it's when those minor inconveniences happen where AI is taking a bit too long to respond that I finally think like, oh, there must be something wrong. But again, majority of the time, uptime is great and we don't have to worry about those things because Cummins and data centers are equipping us with everyday conveniences. Before we wrap, knowing that reliability and sustainability must go hand in hand, I'd love for you to leave us with one last thing. What's that bold vision that Cummins is chasing to make this reality?

Bhargava (17:34):
Very simple, right? Our bold vision is this new and zero a roadmap to eliminate emissions over time. And it's a commitment that we share with our data center customers everywhere. What it means in a practice is twofold. We are investing in future with technologies like advanced batteries, fuel cells, or even hybrid systems that will reshape how the digital infrastructure is powered. And at the same time, we're giving customers practical tools they could use today by HQ already, generators that cut carbon by 90%, tier four diesel for toughest regulations and natural gas systems that lower emissions while keeping reliability strong. The big picture, we want to make sure that the growth of AI cloud and the digital economies powered in a way that's not only reliable today are sustainable and future ready for tomorrow.

Kim (18:27):
Thank you so much for letting us download some insights and also reminding us that keeping AI online also means keeping our planet online too.

Bhargava (18:36):
Yeah, that's right. That's the right way too. And I mean, always remember, right? I think I really like this quote. I've read it online, right? Cloud means it depends on how many people are using AI means it scales as a mission, scales, and mission never sleeps. So that's a good way to relate AI whenever you have a doubt about AI.

Kim (18:56):
Thank you, Bhargava.

Bhargava (18:58):
Thank you.

Kim (18:59):
Alright, so what's the big idea for this episode? Data centers are powering the AI that shapes our daily lives, but they're also pushing grids to the limit and fueling demand for lower carbon solutions. Companies like Cummins are meeting the challenge with reliable power that also reduces emissions from advanced gen sets to battery energy stored systems, and beyond. The takeaway, reliability and sustainability aren't opposites. They're partners in the future of AI, depends on solutions that deliver both power onward.

Show ID (19:39):
Thank you for listening to Power Onward. Your support means the world to us. If you enjoyed this episode, please don't forget to subscribe, rate and review on your favorite podcast platform. If you want to dig deeper on what we covered today, check out the show notes for additional links or go to cummins.com/podcast. Until next time, power onward. We are from Cummins, the company that's been innovating toward the future for more than a hundred years. We're no strangers to rapid change, global shifts and economic uncertainty. We want to be your constant during the energy transition whenever, wherever, forever. Power Onward.