Cracking the AI Power Problem: Vertical Power, Industry Alliances & Game-Changing Strategies

Speaker A: In the future, all data centers and AI chips will be underpowered based on current structure. Now change the architecture, free up energy, have an 85% reduction in losses. That's a game changer. So now at this show it was vertical power is a must have to power next generation AI chips. So that's where we start to get traction.

Speaker B: How are you?

Speaker C: Good to see you.

Speaker A: You too as well.

Speaker C: How was the conference?

Speaker A: Shockingly good. Okay, shockingly good. Uh, we had overflow crowd in our little 10 by 20 booth and really uh, stood out as unique competitors flowing through there asking us stuff. And then we also had an industry session. The product is about data, uh, center power, about making it more efficient. It's a big topic right now. And um, uh, we pitched and got the conference to give us an industry session which is less about our product, although it was about our product, more about the trend in the industry. And then we recruited Nvidia and AMD, huge AI chip companies. Uh, and then GlobalFoundries, a uh, fabrication group that does unique fabrication activity. And they signed up to pitch with us. So we were the very last day the show floors closed on Wednesday. Our uh, things at uh, 10:30 on Thursday, overflow crowd overflow like 250 seat room spilling into the hallway. Standing room only. It was awesome. And so now we're trying to figure out what to do about that.

Speaker C: Awesome.

Speaker A: Yeah, yeah, it was good. How did you get the chip companies to play along? We're engaged with them. Um, right now. Um, we've got a technology that is imagine um, a circuit board, a um, motherboard. Right, um, right now the way power goes in the servers in a data center, the power sits right here and that's about a foot, two foot long. Right. Um, the power supply sits here and it delivers power to the AI chip which is right here. Okay. What we do is um, that the physics of that are very challenging. It is the biggest, ugliest loss in a data center. It's not from the power line to the wall of the building to the server. That's okay. It's in the server to the chip where all the loss takes place. And we have a um, we move it from on top here to underneath, directly underneath. It's called lateral power changing to vertical power. And it is a uh, 90, 85% reduction in loss on the board which is the equivalent for ah, like a 500 megawatt data center. It's the equivalent of putting a small nuclear uh, modular reactor worth of power next to that data center. That's the power that we free up. It would really bridge us to 20, uh, 30 plus where all this new energy is coming online. But between now and then it's ugly. And the chip guys realize that the current structure can't power their new next generation stuff. Everyone wants more AI. More AI, but we can't power it, the architecture can't power it. So we've got a change in architecture. So pretty hot. Yeah, pretty hot, pretty psyched.

Speaker C: That's amazing. So you know when you look at something like that where you're literally essentially helping redefine what the architecture looks like, what are the, I mean there's going to be a lot of organizational drag there, right? So demand is one side of the problem, but the change conversation on the other side of that is huge. Like you're going to have industry players that are already there and you're entrenched in this space dragging what you're doing and things like that. Like how do you overcome the inertia there?

Speaker A: That's a really great question. So a year ago, the category that we play in is called vertical power versus lateral power. A year ago the people in the supply chain of our business for data centers, so not just the chip guys, they're not our customers, we're developing to their requirements. It really is everywhere from the big server manufacturers, so the big PC guys to uh, the hyperscalers like an Amazon or someone like that, and even the integrators that are integrating things and building out the data centers for hyperscalers, that's all like this complex supply chain. They're all building new data centers. It takes about two to three years to build a new data center, sorry, five years max. But they roll the infrastructure every two to three years to accommodate the new AI chips. Um, and so they're rolling it and there are all these supply chain guys throughout, uh, the ecosystem are all engaged in building. That's why they want going after new power sources in Washington, because those are long lead capital intensive items. Um, so nobody wants to change the architecture because that is stacks up their supply chain, that is their forecasts, that is new inventory they have to buy and stuff like that. Changes of things. But what's happening with the AI chips is the servers themselves are changing to accommodate more GPUs and CPUs on the motherboard. In the rack itself, the racks are going from horizontal to vertical, um, which means that there's not a lot of space for the power supplies. It almost has to be underneath as I described it to you. So that's the catalyst. We're building to the chip guys requirements and then turning around and going down the food chain and saying this is Nvidia and AMD's requirements, this is what we're building, et cetera. So a year ago no one was thinking vertical power was needed. It's good enough right now. There's some spit and duct tape options we've got. I won't bag on the competitive option but um, that can be loosely classified as vertical power. But it's not underneath on the back side of the board. It's on top. Um, they uh, um, were saying we're building this to their requirements. A year ago no one thought that this was going to be that big an issue. But then all of a sudden, um, with the announcements from Nvidia specifically on the Rubin platform, Rubin hits Q4 this year and then 2027 is the Rubin Ultra and the processing capabilities and the change to the servers that it requires and then the power requirements. Jensen, uh, last year, this time last year at Nvidia said in the future all data centers and AI chips will be underpowered based on current structure. Now change the architecture, free up energy, have a 90, 85% reduction uh, in losses. That's a game changer. So now at this show it was vertical power is a must have to power next generation AI chips. So that's where we start to get traction. And we had been talking to them prior on the requirements. So the ask was yes, we want to educate the market about this. Yes, we want people to get in line. And that's the show. We were the category was at uh, one of the top themes and we were at the center point of that. So it was in the eye of the storm. It was very fun.

Speaker C: So what I'm hearing is there's not a ton of sort of back pressure here against this and you're at this advantageous position where you're effectively leading a charge that everyone wants solved anyway.

Speaker A: That's the way it's changed. A year ago we could not say that because there was a lot of headwinds. Um, and some of these big companies are very large established companies. Uh, and uh, they're just um, uh, established in their ways of changing things is a material expense for them from manufacturing on up to engineering.

Speaker C: Yeah, skill ups and all that stuff.

Speaker A: Yeah, it's difficult to change these things. So there is natural headwinds. But, but the holy grail is more AI compute. And they know that that's what they're trying to structure. They're trying to make their data centers bigger, but the energy waste is still the same whether you have a nuclear plant by it or not. If you're every one unit that's coming off the power line, only half a unit, something like half a unit gets the chip. Nobody knows that. They're not communicating that they're saying, oh, we're about 90% efficient. But what they realize is, um, on the servers, that's where the loss is. And to get more power, power density on that chip, more power capabilities for that chip, that is the near term and long term play rearchitecture to uh, free up more energy for that chip so that when new energy comes online in 2030 and plus that that energy will be more efficiently delivered, uh, and with more power available to the chip. Does that make sense?

Speaker C: Yeah, absolutely.

Speaker A: So it's like not too late to be early is the saying, or is that the saying? And so it gets challenging because you are early sometimes in a hot category. The market does have to catch up with you. But that's part of where these alliances come in. You know, you get big players. Um, uh, if you're selling a product in the marketplace, um, sometimes getting a major player to sign up with you is a big deal. It's a first mover advantage for them. But also it's a signal to the market that, look, they're paying attention to this company. We should pay attention to, we're late. It creates fear that companies are late and they jump in with you.

Speaker C: Yeah. In the context of where you're headed, um, and for the folks that may be listening to this, that aren't in an engineering sort of centric space, you know, how do you translate that to, you know, other services, other domains, other capacities beyond this situation where clearly you have an engineering advantage at this point. You, you know, yeah, you, you've seen the future, you know where it's going. And so all roads lead to you in, in a lot of the sort of sectors out there, in a lot of the industries, that's much, much harder to ascertain. Do you have a strategy that you kind of go through to get from, to close that gap? Yeah.

Speaker A: And whether you're selling consumer products or consumer, um, technology, uh, consumer electronics, um, the market. So the market includes anybody that's a seller, anybody that's a reseller, anybody that's an integrator or end consumers that are buying on Amazon or your own website. Uh, people go through, whether it's businesses or individuals, they go through four cycles that you have to touch the first one is awareness. If they don't know about, they cannot do the next phases, which is consideration. If they know about you, then they can consider you as an option. So there's certain marketing to get awareness. Certain marketing, marketing to get consideration. I'm an option for you, what problem am I solving? So awareness, consideration. The third step is evaluation. Provide the information so people can make an informed evaluation. Informed assessment. And the last one is, um, so it's consideration. Evaluate. Consideration, sorry, Awareness, Consideration, evaluate and then purchase. What's the incentive to purchase? Make purchase option available. That's also partners, right? Sellers. So whether you're at a Best Buy or whether you're at a Aero Electronics distributor for semiconductors, um, you have to make it available for purchase. So you touch the market in all those four ways. Sometimes if you have a narrow market, you don't have to drive as much awareness because you have a handful and some of the technology stuff that's integrated. Like we're talking about a handful of customers, we're not talking about 5,000. Um, but in the case of consumer products or even business to business products, um, you know, there could be, you know, 100,000, a million folks that could buy the product. And so that cycle then becomes important because you have to make sure all of that, those aspects are covered in your marketing and sales strategy.

Speaker C: And that moves your spend around too significantly. Right. If you need to spend a lot of time in brand awareness, that obviously means capex there and that changes on how much you're willing to then optimize the other downstream effects.

Speaker A: It's true. There's a beautiful thing called opm, other people's money. And this is where alliances come in with your distributors and your sellers. Really important, um, our customer's customer is our customer. Does that make sense? So when I'm selling to a retailer or I'm selling to a distributor, they're selling to others, that's their customer, but it's also my customer. And so providing um, uh, programs that meet their needs. I'll give you an example. Um, in a prior company I worked uh, in the smart home space, right? And we had, there's distributors and there's um, uh, dealers, uh, all across the country. We had 5,000 dealers and we had uh, eight, eight, six or eight various um, kinds of distributors, right? So what, what I did is I took one of these very large dealers and they're the dealers are folks that install the stuff in your home. Think of um, a small version of ADT or a small version of vivint. Um, and they're going around knocking on doors and installing smart home systems and security systems in your, in your, in your neighborhood. Um, they're marketing in the neighborhood. It might be guerrilla marketing. Knock on the door. But the bigger ones will have general marketing budgets that are spending. Um, I went to my distributor and said, I'm already giving you cooperative marketing dollars. I want to focus it on demand generation with this partner in New England. And then I took the program to the partner in New England and said, I've got this money from my distributor that you buy from. I've got this money that I'm poning up, and I'd like you to, to, to. To do no more than 2% of our sales with you, but I'd like you to contribute equally. So I turned $1 of spend for my partner, my, uh, dealer partner, into $3 of a program. And so other people's money can be a force multiplier. And what I did is I structured a program so it met the needs of my distributor, it met the needs of my customer, my dealer. And it helped them drive their business with consumers. And so everybody wins. Um, you have to drive that program. And we tested and we rolled out, and it was effective driving leads and driving, um, business for my partner.

Speaker C: Ah, that's great. The thing that sort of stands out there is again, the depth of industry knowledge you're going to need to have to develop a program. Like that isn't something that you're going to pick up in five minutes of study. You're not like, oh, yeah, let me crash this on the airplane ride over. Right? You have to get to that. Uh, and do you leverage, like, how do you leverage the organizational knowledge you may already. Domain knowledge you may already have against the industry knowledge that may be harder to come by.

Speaker A: Let me see if I can answer this correctly. For you.

Speaker B: Um,

Speaker A: a lot of this stuff is experience, and a lot of this stuff is watching what doesn't work. And, uh, one example was before I did that program, I looked at my cooperative, uh, marketing budget with my distributor. It was 2% matching. And we were not driving at all. We were just giving them our 2% and not asking for anything. And we were getting messages on hold. All right? And we were paying big bucks for messages on hold when folks would call in. That has a low return. It really is selling to who already is buying, and it's not really right. It was a lot of money, too. A lot of money. And so it's understanding what doesn't work. And Then trying to figure out is how can I think differently, how can I put, put a program together? Um, it is out of the box and it's non standard sometimes, but it's okay to be nonstandard and try things. The other thing that was I dropped the barrier on risk. I said I'm going to use some round numbers. I want us all to spend a million dollars in the back half of the year on this program. That's $3 million total. But I don't want you to spend a million dollars up front each. Okay. What I want to do is I want to go Washington D.C. and let's do the D.C. market. That will be $100,000 for each of us. Okay. And it's a high demand market, but it really can be indicative of a rollout throughout the eastern seaboard, um, Virginia and places like that. Uh, um, and it's a saturated market, more affordable. So we tested, we tested in one city, city metro region and it worked. And also keep in mind the program does not need to um, deliver massive profits. What it needs to do is it needs to be break even for the money you're generating. I'm getting the money back. Then you can scale that all day long. You can scale it right up to that million dollars for each of you because it's a direct return on the budget. It's don't spend it or spend it and drive top line. So then when your top line is being driven, then you can modify and tweak to deliver your bottom line. Cross sells with high margin products and stuff like that. I'm getting a little inside baseball on this. But um, but the point here is this try test because some of these things have never been done. Others are only done by a few, a handful and not done by, by, by the mainstream. And that's okay as long as you can have all the stakeholders can get their piece of value from it. What do they care about? Right? And make sure that the program does that. So it is a testing thing and sometimes it doesn't work for sure. That's why you test. That's why you test. I've had um, products, uh, with $10 million consumer electronics products, wireless speakers, $10 million in investment to build, you know, millions of dollars and then a million dollar test marketing campaign. Best Buy loved it. We drove a million dollars in one month of sales and then we shot the product because to really make it successful beyond December, it would have taken $50 million and the return would be five years out and it wouldn't have Worked. So we tested it and then decided that it was best to take that $50 million and invest it elsewhere as example. And so you've got to test it. There's three things you do test, test, test. With stuff like this, I feel like I'm in real estate. Right? Location, location, location. But that's it though, right? It's like you gotta test and you gotta think out of the box and make sure that stakeholders value is delivered in the process that gets them to bite. And then look at your KPIs, look at the metrics that you expect on a minimum investment. If I'm already spending a million dollars in marketing with my distributor anyway and it's not giving a return, then let's go make sure that the distributor gets the return on that dollars that they've spent. They get increased revenue. But also I get things I don't get by the current spend, so I get something new. Even the test market that you don't continue still can define value in learnings. And that's the whole thing.

Speaker C: Yeah. When um, when you started in this

Speaker A: position about eight years ago, is this

Speaker C: the strategy you took or were there

Speaker A: initial steps that you did beforehand as part of marketing? So um, uh, the company uh, eight years ago had some breakthrough technologies. So understanding what they were and how to productize them was a major first step. And uh, um, and so engaging with targeted uh, customers uh, on a early basis. Not everybody, you know, I could get into the biggest customers on the planet but they also move slowly and they can bleed out a small company. So we worked with forward leaning customers who were interested in new technology and we said this is what we do, this is what we think it can do. And um, what do you think? So I'm really big on for anything you do is um, a sell design, build structure which means um, you convince customers to come on board with you to get you the feedback so that what you build is what they want. It sounds simple, stupid, but so many times if you build it, they will come as a model and that's hard. So if you can define requirements, this is what we've done all the whole time at this company and others as well. And okay, to be honest with you, um, I stole that line, uh, I stole that line from um, Finest Connor who launched Shoegart uh, hard drives and then Seagate hard drives and then Conner Peripherals. And I was involved with them when they were the fastest growing Fortune 500 company ever in American history. And Finest Conner's thing was I build what my customers Want. I'm not trying to build a better mousetrap unless they want a better mousetrap and they have to tell me what they consider better and I'll go build it. So sell it first, then design it and then you have a preset market. It's very powerful and very lower risk when you do it this way. And this is what we've done, you have identified market opportunities, you engage with customers to say, what are the problems you have? And let me give you some ideas on how to solve it. Whether it's a consumer or whether it's a, ah, you know, a business buyer, it's the same thing. What problem are you solving? How is it better than the five others that are in the marketplace to solve that problem?

Speaker C: You know, it's interesting too because I think, you know, what I'm hearing in what you're saying is that there's a ton of uh, this strategy that really is hyper effective in the product space where you have multiple distributors, you've got a supply chain, you've got the traditional like I am selling objects. On the service business side though, that conversation changes significantly.

Speaker B: Right?

Speaker C: There's no downstream distributor necessarily. It's a little bit harder in terms of the mechanics. Have you been in a position where that was something that you've been responsible for and how did your approach change?

Speaker A: Yeah, service is interesting too because service can be um, let's use the software, uh, as a service or subscription services, which is the holy grail for folks because they make so much money, they're productized services.

Speaker C: Effectively though, it's just kind of a hack, right? It's not quite the same thing.

Speaker A: Yeah, yeah, yeah. It's like, you know, I was in the um, security business, so uh, alarm monitoring, right. That's the ADT business. They make so much money. Profit, profit. So I wrote this article a while back, says, thanks for the cash. Don't call me again or ever. Because once they get people signed up, they don't want them to call them because then they're saying, why am I paying $49 a month again? And the alarm never goes off so they don't want to talk to them. I completely don't agree with that because I believe that every commerce event, when you have such a thing as a service subscription, every commerce event after acquisition resets the clock for when they get, uh, is this valuable or not? So I believe in a lifetime customer life cycle approach and creating a product mix, services and products that can service people in a life cycle. So I'll Give you an example. With security and alarm monitoring, many people get the security system. Uh, when you have young, um, children, that's a life cycle point. When there's a busy life and you want the family to feel secure at home. Right. Makes sense. So guess what happens in 10 years? Uh, so right now profitability occurs after three years. That's where they break even, start making money. That's why they don't want to talk to you. It can be sooner too. But then after that they are printing money. There's very little operating expenses. It's a very high margin. So they want to keep it going. So they don't want to upset the apple card. However, after a certain period of time, people understand the alarm is going off more. For my teenagers, not uh, my little kids anymore. It's been ten years, it's been eight years uh, from my teenagers than from burglars. Why do I have this system? I should just cancel it. The kids can take care of themselves. Very real thing. What if you then add an automation layer on top? Now I can control my garage door with this thing. When you sell that system back in after acquisition, it resets the value clock on a service, uh, and you start from scratch again. I love my garage door opener and I don't mind the security expense cause I paid one and done to add the control thing and I'm still paying the $49 a month for the security service and enabling my garage door to be opened. What happens the next year? I get a thermostat. What happens the next year? I get some m lighting. And so each year you're resetting that value prop. Let's go through the life cycle again. Teenagers are gone, kids out. I'm now 68, 70 years old and my wife uh, has some issues. So now what I'm getting in my is I've fallen and I can't get up the Life Alert pendant. So there's a Life Alert you could add that subscription service to at the end. So thinking through not a three year customer cycle, but a 30 year customer cycle, it's important. What else do you. The best thing to sell is to an existing customer. I'm sure you guys have all heard that. And so in the services business, understanding that life cycle, understanding the adjacency, selling opportunities that have affinity, they're adjacent to the core solution, but they have affinity. Um, uh, I worked for a while in a company called Sonic Solutions. And what they did is they did DVD authoring. They were in the Windows operating system to make a burn a DVD. Uh, and to author a DVD. We were in 95% of all Hollywood movies. DVDs that you got at Blockbuster. Blockbuster was a store that sold DVDs back in the day. If you're number. I'm just kidding. Uh, you know, they disappeared quickly. But, um, the company was very good. And what we realized is that, um, um, there was huge adjacency to DVDVD playback and home video authoring. What is it? It's music and photos. Some people just like photos, some people just like the movies, home movies. Some people like music, some people like two. But everybody does not want to feel caught short because they have some affinity with all three. So providing the photo people with the ability to add movies to their stuff is a big adjacency. So that's an example how you can evolve your relationship and broaden it. And um, whether it's a service or a product sale, that was a great experience. We end up creating these huge suites that had everything in it. Because if you want to buy photos or you want to buy music, but you can get everything in a suite and not be caught short. Whether you do movies in the future or not doesn't matter. Shocking amounts of people would trade up to the most expensive one to not be caught short. They want the complete solution. And so that the concept, whether it's a service of looking at what is that affinity and adjacency that you can actually sell to continue your relationship. That's huge. That's a big deal.

Speaker C: Yeah, I dig it. Go ahead. Go ahead, Brian. So, I mean, uh, I'm listening through what you're talking about here, and every one of the service examples that you've provided has been effectively what is a productized service? Um, it's not, you know, medical services necessarily, or, you know, professional services or consulting or IT enablement as a concept. Right. You're really talking about, uh, productizing the service down to that level to then get your increased market pen and all that other stuff to essentially get greater wallet share. I think that makes real good sense. It's hard to get past the traditional service mindset when you're a plumbing company or a, uh, roofer or whatever. How do you translate, you know, these concepts, which are, you know, profound and super well developed in a product space for those folks that, that are just so used to like, no, no, I turned this wrench or I, you know, type these forms out.

Speaker A: It's, it's, it's really true. It's really true.

Speaker C: Um,

Speaker A: and that's where the Art of marketing comes in where you have to convince them that there is opportunity that's untapped. Many people will say, I'll give you, I'll give you an example. I work for a um, ah, ah, durable medical equipment healthcare company, ah, selling um, CPAP machines and oxygen concentrators and various accessories. Um, and so we identified uh, early on that um, that CPAP customer was between the ages of 50 and 80. So what does that mean? Um, the first section, 50 and 60 had aging parents. And we realized that there was an opportunity to sell. I'm going back to, I've fallen, I can't get up a personal emergency response system. We realized that um, these folks that had CPAP machines were of an age demographic, that there was a high propensity for them to know somebody who needed such tighter care. Or the second half of that demographic who were buying the CPAP machine were directly in that space. They had oxygen issues, right? And so they themselves could buy it or for a spouse. So we realized that there was a very, very significant affinity. Now this is a product that we were selling but the service is the monitoring. And so we identified and so convincing, um, the folks that said why do we want to sell that? We're not in that business. Why do we want to sell it? That was internally. Why would we want to sell this? And so we had to go through the demographic fit. We had to go through here it uh, sell them another mask. Or you can sell them a service that is five times as profitable. Okay. And doesn't cannibalize your current business to sell them an additional mask. Because that's, that business is you replenish the masks and stuff like that. So that's an example where you had to sell internally. We also had other partners in other categories outside of healthcare, um, where you had to convince them, look, I'll give you, unfortunately it's the same category. I gotta mix my playbook up. But in the smart home space, um, the biggest problem they have is they don't wanna roll a truck. A truck. The security people, I use ADT as the example and then the smaller players, they don't wanna roll a truck, they wanna roll it once and then print the money thereafter and get the profitability and stuff like that. Um, and uh, we're in the business, we sell security. It was a lot to get em to sell some of the automation stuff but they recognized they could get more. Uh, they did that. What they didn't feel is that they could sell anything else other than this smart Home stuff. So we did the analysis. Those young families were identified that they all have aging parents and they could all sell. Um, I've fallen. I can't get up. Why no truck roll. So why would they want to do this? No truck roll. Highly digital oriented marketing. So they could easily execute it without a lot of sunk expense and literally they could cross sell it to um, the mom in the house is a very large influencer for security as well as they're the ones making decisions for in laws and parents that are older and so targeting the mom that just got the security system said we have this other service. We convinced the dealers that you don't have to roll a truck. You ship it, mail it and then the automation folks can just literally set it up for your mom on the phone, for grandma on the phone. So no truck roll, no expense and it doesn't cannibalize their current subscription. That's their biggest fear. Like I said, thanks for the cash. Don't call me again or ever because they don't want to upset that. This doesn't upset it. It doesn't touch that first subscription. It's a separate bill with a separate location and a separate person. They bid hard. They bid hard. We huge cross sell capabilities. When we started educating these channels, this was at a smart home company I was uh, at. So understanding as I uh, went to the beginning, we talked about other people's money. Understanding the stakeholders, perception of value. What are we trying to drive here? So a uh, distributor is going to have a different value than a dealer or integrator than a retailer, et cetera. What are we trying to drive and making sure that you make a case internally. The value. Here's the return we can get from this. And then going through those stakeholders and saying what do they care about most? Not what do I care about, what do you care about? And coming up with a product or a program that services those people's needs. That's everything.

Speaker C: How do you create that type of

Speaker A: ltv Here, here being. Sorry. Oh Amber. Semi sorry. Oh, okay. A little bit different technology integration, handful of customers, very big. Um, but the value we create is we've identified with these chip guys, um, the problem in the marketplace. We also have some folks internally that are very sophisticated and smart on the engineering side who understand the space, who understand where the problems are. Uh, and when the head of Nvidia comes out last year and says we will be underpowered with current infrastructure, that's a message, that's an opportunity. And it's the hottest space in technology. Okay, let me go down a couple paths here. Power used to be a commodity. It's just power for a hyperscaler. For someone who runs a data center, it's the cost of doing business. I have to pay for power. It's baked into everything. However, when the goal becomes I need to drive more AI computing, more AI compute, but I can't get enough power to that chip because it needs so much now. And not only do I not have it on the grid, but I'm leaking like a leaky bucket. I got a great graphic that I did with my CEO on a leaky bucket graphic because it's like you're pouring water in but it's leaking out like a shot with machine gun. It's leaking out all over the place. That's what it's like. And so, and so power um, is no longer a commodity. It's a premium. It's a premium. Uh, we've got some data that shows that we're technically demand is outpacing supply right now. That's why you've got energy prices spiking in data center heavy areas. Virginia right now nationally we use about 4% uh, of our grid energy for data centers. It's slated to go to 12 according to the Department of Energy, um, by between 28 and 2030. Some um, expectations are that it will go closer to 20% by 2030. Virginia, the state of Virginia is already at 25% of its grid capacity is used by data centers. That's huge. That's huge. And it's not sustainable. Um, and so power is no longer a commodity. So from an Amber's perspective, um, our teams identified early that a RE architecture was going to be required, validated that with the original source of requirements, the chip guys. Right, the AI chip guys, and said yeah, I mean we're going to be at 40 times the processing speed. Um, there's something called Moore's Law which is um, how um, capabilities scale over time uh, with size. Uh, and that's what's happening right now is that Moore's law is trafficking um, uh, to power. As processing increases, power requirements increase. And you can't put more water in a leaky bucket. It still leaks. It still leaks. So this is where we identified that uh, driving the requirements on the roadmap from the chip guys, that was the way to flow back into the market and say there's a different way that this can be done. This was last week was a coming out party for us in that sense. Um, and we leveraged a lot in commercial manufacturing of our first uh, chips. And we'll be in sampling at the back half of this year, uh, with key customers. And so using that milestone to say, um, look, we're real. This is not slideware. We are real. We just raised $30 million in funding, our product's in manufacturing. Here's the strategy. I mean we had folks coming up that just said because of how we presented it again, I've got a very good technical team. Um, and uh, we presented it in a ways that our competitors literally said, uh, game over. I'm going to have to revise my resume because they're working on competitive platforms. That was fun to see though, I got to tell you. You don't want anybody to be out of work, you know, um, but that was fun to see them going, oh my gosh, game over. Uh, they came back five and six times asking questions about us, huge amounts of people on the competitive side. So, but educating the market of again, going back to the theme of, of um, what is perceived as value. So right now a hyperscaler, um, has a lot of things that they want to do. But the holy grail across the landscape is driving AI. You've seen the announcements from Washington, um, D.C. they want to be competitive against China. That is more AI computing faster, sooner, et cetera. And so that's the holy grail. We've identified an ability to get more AI compute sooner without that modular nuclear reactor or the wind farms that they're going to build and the more fossil fuels they're going to burn. Um, we get that Sooner by this re architecture. And again we were early in communicating this 18 months ago and the downstream markets were like, we're fine. Power's a commodity, we're fine. But as the roadmap started unfolding, what we said, because we knew that this would take place, we knew that the capacity was, uh, processing capabilities was going like this hockey sticking. And we knew that that power is going to have to draft that power availability, not necessarily new source immediately. And so we knew that was a RE architecture. So that's what my team drove. And then we communicated that the marketplace with drumbeats. You know, we belong to um, industry organizations like uh, Global Semiconductor alliance, um, and we've been involved in various things in the industry with like the Morgan Stanley Conference on Semiconductors, tiny little startup, um, with exposure. That's significant. It's significant. And without saying it out loud, what we're doing for the industry is we're building a perception of who are these guys? And where do they come from? Even though we've been around for a bit, that's what we're building here. I'll give you an example of this. Um, uh, we've got some folks that we're aligned with in Washington D.C. that are helping us navigate the funding, uh, landscape with the federal government, Department of Energy, Department of Commerce, um, and they have. I'm not sure if you know this, but the US Government has a lot of money. They have shocking amounts of money to spend. I knew this intellectually, but now practically I see it. It's amazing. Um, so this team has us engaged and there was a very large energy and AI summit, uh, last July in Pennsylvania, driven by Senator McCormick, supported by the Governor Shapiro, and supported by Senator Fetterman. Both Democrats and Republicans all aligned. It's really interesting. Um, and President, uh, Trump and his cabinet attended, and my CEO attended and others I know from comments and activities that took place with. Why the heck is he here? Well, because we have something that is very disruptive and we've gotten people's attention and people are now pulling us to them. And so that's the, um, uh, the thing that we have done. We create that pull. Are we pushing into customers or are they pulling us to them? We'll push it first for sure. But then all of a sudden, if you create that value identification that we talked about, problem solving, all of a sudden you find that they'll pull you to them. And that is where acceleration comes from. So that's what we've been doing the past two years in this category specifically, is going from push to pull. They're pulling us to them. It's, uh, it's hard work, fun, and can't get discouraged. But it's communicating that message out there that, look, this is disruptive. You know, try to take the high road. Um, yeah, yeah.

Speaker C: I mean, the benefits there, when you have a, when you have a better mousetrap are, you know, at that point you're, you're not, it's not a conversation anymore. It's just like, game over. We've, we've got the better mousetrap.

Speaker A: Right, but the problem is though, is that, uh, just because it is better does not mean, does not mean that you can win. The technology specifically is littered with the, forgive me, the corpses of companies that assume to this assume just because I'm better. I think the analogy. I'm going to paraphrase, and it's a terrible paraphrase, but I'm doing it anyway. So Steve Jobs said, I want to Build the best product. And he did and he was successful by doing that. Others have built the best product and failed. But Steve Jobs wanted to build the best product. What I read was, um, Bill Gates wanted to sell the most of the products. And so they had two different philosophies. Good, now is better than great, too late, I want it to be perfect. And both won. But in between there's companies that, you know, can't compete because they don't have enough, you know, you know, yeah, wasteland. Yeah. So that's. Where is that Goldilocks of product quality and making sure you execute in the market. Give the people they need to adopt, uh, or sell. Or adopt and sell or adopt and purchase. You know what I mean? Whether it's a consumer or business partners, make sure that value is identified for them. Of, uh, why us? Why us? Why us? Why what we're saying matters and why do we matter in that dialogue? And sometimes it's a multi year education process that takes place, but it's the third party endorsements that jump on board again. Um, last week we had Nvidia and AMD present with us at this uh, Applied Power Electronics Expo. And they were presenting their own stuff too, but they presented with us de facto. That was, that was, yeah, you're saying

Speaker C: seated at the table with the big boys, right?

Speaker A: Yeah. And so that gets people looking, that gets people to look at the company and go, huh, something's happening here. I have to pay attention. And then after a while, um, when we talked about first mover partners, some companies are big and are slow, some companies are big and forward leaning and aggressive. Uh, some are moderate size but very aggressive. And so picking those right initial partners is everything. Because once, um, you come up with those alliances and start to present those in the marketplace like we did last week without, everything's in our NDA, very difficult unless you're shipping something with somebody to be able to communicate the detail. And so. But the fact that they're on board with you tells anybody that's not in a dialogue with you, I am late.

Speaker C: It goes back to that old adage, right though this is important where you're smart when you pick your friends. Right? Pick the right friends.

Speaker A: That's right.

Speaker C: And I think that's a huge kind of competitive, uh, advantage as you go through this. Rob, I want to take a moment and thank you for being on the show today. The insight you've got is absolutely super, uh, valuable. And for folks uh, that are looking for uh, that perspective and your rich sort of background in that everything through the supply chain management all the way through to partner selection. Uh, you've provided at a masterclass here today and I want to thank you for being on the show.

Speaker A: I appreciate the opportunity. Both of you. This has been a lot of fun.

Speaker B: Thanks for listening to the Masters in Marketing Agency podcast. I hope got a ton of value out of this episode. And before we go, I just want to thank our sponsors. DevNoodle DevNoodle provides marketing agencies with the ability to offer their clients unlimited website design, build and management services with fixed monthly plans. If website design, development and maintenance is holding your agency back back from growing, please reach out to us@devnoodle.com where we make websites easy. Easy for you and easy for your clients. Devnoodle. Com.