Artwork for podcast The Industrial Talk Podcast with Scott MacKenzie
Marcelo Lorenzo with Bluefin ESI
13th May 2024 • The Industrial Talk Podcast with Scott MacKenzie • The Industrial Talk Podcast with Scott MacKenzie
00:00:00 00:18:07

Share Episode


Industrial Talk is onsite at DistribuTECH and talking to Marcelo Lorenzo, Sales Engineer with Bluefin ESI about "The reactive power impact of AI on datacenters".
Marcelo and Scott MacKenzie discussed the growing demand for power in data centers, particularly in Virginia. They emphasized the need for collaboration between utilities and data center operators to find innovative solutions to meet this demand efficiently. They also explored the potential of generative AI to transform various industries.

Action Items

  • [ ] Reach out to Marcelo Lorenzo from Bluefin to discuss potential power quality solutions.
  • [ ] Attendees should check out Bluefin's website ( and connect with Marcelo Lorenzo on LinkedIn to discuss collaboration opportunities.
  • [ ] Schedule Marcelo Lorenzo as a speaker at future Distributed Tech conferences to share updates and solutions.


Power and data centers in Virginia.

  • Marcelo from Bluefin discusses data centers and the intersection of electrification and power needs.
  • Speaker discusses energy consumption of AI-powered search engines, citing example of Google and Bing.

AI's growing demand for power and its impact on the energy market, with potential solutions for utilities and data centers to collaborate and

  • Marcelo discusses Bluefin's niche market in power solutions for data centers, particularly for AI workloads.
  • Scott recalls a moment where chat GPT's popularity suddenly increased, leading to increased energy consumption.
  • Utilities and hyperscalers are adopting AI to address fast-paced energy challenges, requiring collaboration and synergies.
  • AI adoption in energy is accelerating, with moral implications and demand for nimble solutions, as seen in Marcelo's perspective.

AI and generative power quality solutions for data centers.

  • Marcelo highlights the importance of power quality and efficiency, citing transmission line inefficiencies as a major issue.
  • Bluefin uses creative engineering to address these issues, leveraging existing technology to solve customer problems.
  • Scott MacKenzie and Marcelo discuss the rapid changes in the industry, including AI's growing presence.
  • Marcelo highlights the need for drastic changes in the traditional model, citing examples like data centers building their own substations.
  • Marcelo Lorenzo discusses Bluefin, a company using generative AI to solve real-life problems in various industries.
  • Industrial talk podcast network aims to educate, collaborate, and innovate with individuals in the field.
If interested in being on the Industrial Talk show, simply contact us and let's have a quick conversation.
Finally, get your exclusive free access to the Industrial Academy and a series on “Marketing Process Course” for Greater Success in 2024. All links designed for keeping you current in this rapidly changing Industrial Market. Learn! Grow! Enjoy!


Personal LinkedIn: Company LinkedIn: Company Website:



NEOM Hexagon: Palo Alto Networks: Palo Alto Networks Report HERE. Fictiv: Hitachi Vantara: Industrial Marketing Solutions: Industrial Academy: Industrial Dojo: We the 15:


LifterLMS: Get One Month Free for $1 – Active Campaign: Active Campaign Link Social Jukebox:

Industrial Academy (One Month Free Access And One Free License For Future Industrial Leader):

Business Beatitude the Book

Do you desire a more joy-filled, deeply-enduring sense of accomplishment and success? Live your business the way you want to live with the BUSINESS BEATITUDES...The Bridge connecting sacrifice to success. YOU NEED THE BUSINESS BEATITUDES!


Reserve My Copy and My 25% Discount



ai, bluefin, marcelo, industrial, utilities, substation, solve, power, problem, talk, solutions, innovation, point, customers, datacenter, lorenzo, data centers, efficiency, adoption, improving


Welcome to the Industrial Talk Podcast with Scott MacKenzie. Scott is a passionate industry professional dedicated to transferring cutting edge industry focused innovations and trends while highlighting the men and women who keep the world moving. So put on your hard hat, grab your work boots, and let's go.


Alright, once again, thank you very much for joining Industrial Talk thank you for your continued support of the platform that celebrates industry professionals all around the world, you are bold, you are brave, you dare greatly you innovate, you collaborate, you solve problems, and therefore you are truly making the world a better place. We We here at Industrial Talk, appreciate your efforts. All right, we're on site DistribuTech. lot of buzz happening here, Orlando, Florida. And it is a guy again, I say it all the time a collection of problem solvers. That's what they do. And they're trying to make definitely solutions and innovations all around here. That's what they do. All right. In the hot seat. We have a gentleman by the name of Marcelo, the company is Bluefin and we're going to be talking about Marcelo datacenters. All right, let's get cracking


power and data centers.


Having a good conference, or


having a great conference, it's pretty cool, huh? Yeah, a lot of buzz.


A lot of a lot of us a lot of us. It. I think it's back to sort of pre COVID levels. I'd


say even beyond that. Yeah, right now we're seeing a lot of intersections between electrification of vehicles. And, you know, the space that I occupy, which is more than datacenter space. It's, it's just a race to get power and a race to it's just not enough power. Not enough. Not


enough power, isn't it? I was hearing stats like three times, you're gonna need three times 3x. There it is. That's the that's the mantra. I think it's I think it's conservative. I think it's gonna be more.


Yeah, I mean, some of the stats, even recently, I've been looking at how much energy each AI search takes. So say, for example, now you're on Bing or on Google, you type in your search. All of those platforms are going to be running AI, generated responses. And each one of those responses takes enough energy that could power your cell phone seven times over. Yeah, see? So it's a lot of power.


Yeah. See, I absolutely understand exactly what you're saying. Couple of sort of foundational first, tell us a little bit about Marcelo. Yeah,


so I'm actually I grew up in Canada. We're about Toronto.


All my family's from Hamilton. Oh, no,


no way. But if you're from Hamilton, then I should clarify the Toronto is the GTA is where I grew up in which, but the actual city was Mississauga, which is right


outside. Dallas. Yeah. Burlington right out there.


So my family's still up there. But I moved down to Pittsburgh, I lived there for a little bit. And then now I live in Virginia, which is datacenter alley. So Virginia, where I live is Northern Virginia is 80% of the world traffic for internet goes through Northern Virginia.


Oh, say that STAT again. 80%


of internet traffic for the world goes through Northern Virginia. I did


not know that. Yeah. How does? How does Virginia deal with the load a


ton of data centers?


I didn't know that is


easy. Yeah. I mean, there are other tier one markets that are popping up like, you know, Arizona, you have Ohio and other other towns, Atlanta, maybe North Carolina. But Virginia is still by far the highest concentration of data centers to the point that there's not enough power anymore.


Right, that there isn't that it's only going to get there's going to be greater demand. It just is just because of what we're talking about here at at distribute tech, and the necessity to be able to have that data available as fast as you possibly can imagine, right? And it's not just just little spurts of data. It's it's mega data. Yeah. This is the jump edge. Tell us about Bluefin what what that company is.


So we started as a sales rep for manufacturers. And we were primarily representing General Electric GE. And then we've sort of more rigid, you know, to your point on how you are introducing the show is solutions driven. And so that's the space that I try to occupy most right now in the data center space, which is trying to solve problems that they're facing now with an electrical hat on, right and so, you know, if we talk about power For the data center world, which is the problem that we're trying to solve right now, specifically in the AI space, the power signature for these workloads for these AI workloads is so aggressive, that it creates a lot of reactive power. And so if you want to think about, you know, power is going through the lines, sure, but you have a little bit of reactive and real power, real power is what you can actually use, what powers your computer, what what you can profit from reactive power, you can think of it as unusable, inefficient, right. And so, it just happens that the characteristics of the chips that are powering AI, do create a lot of these disturbances and inefficiencies in the, in the lines. And so it's a prompt for the utilities, and it's appropriate for the datacenter Yeah, yeah. And so that's the sort of niche market that we're trying to, to help solve right now. And, you know, creating solutions that can mitigate and correct that power, power signature.


So clearly, it it's, it's happening, it needs to be more efficient. Did, did you see just all of a sudden, and that's just thought came to mind? Where chat GPT, you know, hit, there was like, a switch? Yeah. And then all of a sudden, that's that. That was that was the tipping point, from my perspective, that sort of just catapulted it did. Did you see sort of that that increase in in energy and all of that associated with that? How do you? How do you see that


it's just, it's happening at such a fast pace, that, you know, utilities are not known for how quickly they move. And they're talking to datacenter, hyperscalers, a, you know, Microsoft, Google, all these guys, that are perhaps used to moving faster than that utilities are. So that gap is a constant pain point that we're trying to address. And it is, you know, you're really trying to solve problems that were not here three, four years ago, right. And so it's got to be an adoption of technology that addresses that issue, that is very new. And convincing these large utilities or large organizations to adopt these technologies, and how to implement those technologies is the key. And so I think that, you know, these fairly new and fast problems that are coming up, are really shifting the mindset of utilities and the mindset of, say, you know, these big hyperscalers, to adopt more of a partnership standpoint. So rather than, say, the data center guy, say, going up to utility and saying something power, come on, give me power, right. There, really, it's not available. So really, they got to work together with the utility, you say, how do we improve transmission? Or how do we improve generation?


But the reality is, is that I don't think that given the complexities that exist out there, and in just the the energy market, just the power, the utilities, what, what's going on here, it requires 100 All it requires collaboration. It there's no other options.


100% Yeah, and that's what we keep seeing that's what we keep seeing, you know, there's it's really looking for synergies or a symbiotic relationship between the two. Right? Utilities also profit from having a constant customer that has consistently load.


Yeah, that's a consistent load them I just it up as, as time goes on. Yes. Right. But the reality is, is that it, you know, yeah,


yeah. But I mean, if you look at the speed up, which adoption has taken over to your point, it's fascinating. I mean, you look at industrialization, it took, you know, several decades, perhaps, then you have the computer to maybe several years, but now with AI, it's like someone's in like, the month, several months to have large adoption and everywhere now I have it on my phone. You know, it's great. Yeah,


that that begets all the other the, the, the moral implications associated with AI and the good, the bad, the ugly and, and, and as AI begins to demand more and more and more, we just, just gotta be nimble. What do you see it going Marcelo? What, how do you how do you deal with that? How do you what solutions do you provide that that help us deal with these problems?


So again, the biggest thing for such a large power demand is going to be power quality. So that's probably tilt, tilt power quality addresses efficiency, you could think of it that way. Yeah. And so there are tons of inefficiencies in the transmission lines, for example. So if you could fix an efficiency in the transmission lines, by improving power quality, you will be able to get, you can think of being able to get more juice out of the squeeze. Right. And so, that is the name of the game. And that is, the more the most immediate results that you could have the fastest results that you could have is the improvements in efficiency. It's really, it takes a long time to build a power plant. You know, all these projects take a long time. But looking at efficiency and improving that is probably one of the quickest things you can do.


How does Bluefin fit into that? How? How do you fit into that?


o Bluefin was founded back in:


Fascinating, I, I, it's way above my paygrade, quite frankly, just because I look at what's taking place in the market today. The dynamic, the innovation, the the, the acceleration of the, it seems like there's a disconnect between these, like, people are talking about, you know, net 05. But, but but that's a heck of a target and people are trying to pursue that. But then we have all of this 3x requirements, it's it's only a software or a innovation solution, because I don't see how, and and how these participants, yes, collaborate, yes, do all of that. That's all important. But to be able to, you can't do it without collaborating.


Like you've got to have that complex model has to change. The model model has to change dramatically. Yeah. And you know, the life cycle, for example, high voltage substations that power these data centers, right. A lot of the times it would take 567 years to get one of these substations installed and commissioned and add powering. But when they pull, they need it in one year. You know, they need the substation in here. So a lot of times you have the customers in this case, datacenters going to the utility saying no, I'll build the substation, I'll pay for it. Just give me the power. Right, and I'll pay for it. Right. So again, it's, I hate you. I hate using a cliche word, but it's disruptive is it goes against the norm. Right? And so if you want to, like you said acceleration is so fast that you have to change that you have to change. How,


how do you? I mean, how long have you been at this? A couple of years, couple of years


and specifically dedicated to this. Yeah.


And have you seen so much just drastic changes that have happened as a result of this? Huge, huge?


I mean, AI wasn't that Yeah, the thing two years ago, right, it was very nascent. And now it's, it's in everybody's mind. Yeah. It's in everybody's mind and it's not going away. It's not going away.




That's very interesting. It's very interesting. And, you know, one of the fascinating things for me as well is that it's something that people can it's palpable, right? People can see every day how much it is improving, right. And it's solving real life problems. Like for example, I think two days ago, I heard this conversation between some executive at in Hollywood, who completely shut off their production and said, We're not going to spend any more money because now we have an AI and machine that could create all these environments for us. digitally, so, you know, it's gonna impact so many different industries.


How do people get a hold of you?


website, we have Lavinia Seidel, calm.


Bluth would say that a glute Bluefin like the fish. Are you out on LinkedIn? Yeah. What's your last name? Lorenzo Lorenzo. That's Marcelo Lorenzo. Just want to make sure. Well, you are absolutely spectacular. Big time. All right, listeners, we're gonna have all the contact information for Marcelo out on Industrial Talk. So if you're not reach out to him, because it is a collaboration game. You can't do it on your own. You need to collaborate with individuals that are true Sherpas in what they do. Hey, go. Industrial Sherpa. All right. Very good. Thank you very much for joining. We're going to wrap it up on the other side. Stay tuned, we will be right back.


You're listening to the Industrial Talk Podcast Network.


All right. His name is Marcelo Lorenzo, the company is Bluefin make a note of that. And the event was DistribuTech. Also put that on your calendar for one T five, to really 25 case. It's an amazing. All right. Once again, we're building a platform here at Industrial Talk. The Industrial Talk platform is about education. It's about collaboration, you can collaborate with any of the individuals on Industrial Talk and of course, innovation. It's happening as you can tell by Marcelo this conversation around generative AI and how to solve those challenges. It's going to be impacting everybody, every buddy. All right. Get engaged, be a part of Industrial Talk, just go out to Industrial click on connect with now it'd be me, of course. And then let's see how we can work together to help amplify your voice to to communicate what you do. Be bold, be brave, dare greatly hang out with the Marcelo and you're gonna change the world have a great conversation shortly.



More from YouTube