Lauren Witter, Vice President at Omni, and Scott Sloan, Executive Vice President at Omni, explore the world of AI factories alongside Dave Vellante, Co-Founder and Co-CEO of SiliconANGLE Media in an engaging segment of the CUBE's ongoing series with NYSE Wired. This episode centers on the advanced plumbing and fluid conveyance systems integral to AI factories, an evolving sector in the technology industry.
Omni is a leading player in fluid conveyance systems, essential to the infrastructure of AI factories. In this episode, hosts from the CUBE introduce Witter and Sloan, highlighting their expertise in fluid systems engineering and its transformative impact on data centers. The discussion reveals how Omni's innovations in connection integrity and fluid network engineering enhance data center capabilities, increasing reliability and efficiency in high-tech environments.
Omni's experts emphasize that the integrity of connection systems is vital, drawing parallels to the rigorous standards required across industries. They highlight significant forecasts in data center investments, underscoring the rising demand for robust and innovative plumbing solutions. The conversation also delves into the role of standards and collaborations in the sector, emphasizing collaborative efforts in advancing technological solutions.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI Factories - Data Centers of the Future. If you don’t think you received an email check your
spam folder.
Sign in to AI Factories - Data Centers of the Future.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Register for AI Factories - Data Centers of the Future
Please fill out the information below. You will receive an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for AI Factories - Data Centers of the Future.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI Factories - Data Centers of the Future. If you don’t think you received an email check your
spam folder.
Sign in to AI Factories - Data Centers of the Future.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Sign in to gain access to theCUBE + NYSE Wired: AI Factories - Data Centers of the Future
Please sign in with LinkedIn to continue to theCUBE + NYSE Wired: AI Factories - Data Centers of the Future. Signing in with LinkedIn ensures a professional environment.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
Scott Sloan & Lauren Witter, Omni
In this interview from theCUBE + NYSE Wired: AI Factories – Data Centers of the Future event, Glean co-founder and CEO Arvind Jain joins theCUBE’s John Furrier to unpack what’s really working in enterprise AI today and what comes next. Jain explains why knowledge access remains the first successful AI use case at scale and how Glean’s enterprise search brings AI into everyday work. He details the past year’s lessons with AI agents – from the need for guardrails, security, evaluation and monitoring to democratizing agent building so business owners (not just data scientists) can create production-grade agents.
The conversation dives into Glean’s vision of the enterprise brain powered by an enterprise graph, highlighting the importance of deep context, human workflows and behavior to reduce “noise” and drive outcomes. Jain outlines core building blocks – hundreds of enterprise integrations and a growing actions library – that let agents securely read company knowledge and take actions across systems (e.g., CRM updates, HR tasks, calendar checks). He discusses how organizations are standing up AI Centers of Excellence, prioritizing “top 10–20” agents across functions like engineering, support and sales, and why a horizontal AI data platform that unifies structured and unstructured data – accessed conversationally and stitched together via standards like MCP – sets the foundation for AI factory-scale operations. Looking ahead, Jain says Glean’s upgraded assistant is evolving from reactive tool to proactive companion that anticipates tasks and accelerates productivity.
Vice President of SalesCoolFlow a Divison of Omni Services
Scott Sloan
Executive VPOmni
In this interview from theCUBE + NYSE Wired: AI Factories – Data Centers of the Future at the New York Stock Exchange, Scott Sloan, executive vice president at Omni Services, and Lauren Witter, vice president at Omni, join theCUBE’s Dave Vellante to unpack the “artful plumbing” that keeps AI factories running. They explain Omni’s role in fluid conveyance systems and secondary fluid networks that move coolant from primary loops into racks, rear-door heat exchangers and server blades, and why “connection integrity” between hoses, fittings and alloys has become ...Read more
exploreKeep Exploring
What is Omni Services' line of work and expertise?add
What is the history and significance of Omni Services in relation to plumbing and hydraulic systems?add
What role do certain companies play in the infrastructure of AI technology?add
What is the engineering involved in ensuring connection integrity in hydraulic systems?add
What is the role of fluid distribution in the design and operation of AI factories?add
>> Hi everybody. Welcome back to the New York Stock Exchange. We're here at the Buttonwood Podium overlooking the Options Exchange. This is the theCUBE plus NYSE Wired's ongoing series on AI factories. And we're excited to have two great guests, Scott Sloan's Executive Vice President at Omni Services, and Lauren Witter is the Vice President at Omni. Guys, good to see you again.
Scott Sloan
>> Great to be here, Dave.
Dave Vellante
>> From last year we had a panel. Lauren, you participated in the panel with Tim from Dell and Dion from NVIDIA. Remember we were talking about direct liquid cooling. We had an interesting debate-
Scott Sloan
>> ....
Dave Vellante
>> on phase change, which was fascinating. And you guys are in the business. Omni is in the business of fluid conveyance systems, AKA hoses. But it's good marketing to call them fluid conveyances. But Scott, take us through what you guys actually do.
Scott Sloan
>> Well, look, so it was about 50 years ago. My father-in-law founded Omni Services. So it's all about plumbing effectively of hydraulic systems originally. And so as we've grown over the last 50 years, the notion of fluid conveyance and the artery networks that we connect to is becoming increasingly important, hence why we're here today.
Dave Vellante
>> So I like to frame this as the unsung heroes of AI factories. It's kind of what we're talking about here. I remember Dave Hitz one day, he's a founder of NetApp saying, "Look, we're the plumbing." You guys are part of the plumbing infrastructure, is that right?
Scott Sloan
>> Well, we would say maybe the artful plumbers.
Dave Vellante
>> What, like the master plumbers, or?
Scott Sloan
>> In a sense. Because if you think about all of the tech that's involved in an AI factory to deliver that compute, and you work down the stack, if you will, to the picks and shovels that we represent down at the bottom of the stack, that connectivity to bring the compute to life flows through our artery systems. That is a craft. So yeah, master plumbers is a good way to say it.
>> Yeah, absolutely. So I always jokingly say our heritage in OEM hydraulics kind of led us here. The integrity of the connection, which is a fun word for how the hose and the fitting or connector stay together. So that's a big part of what we do. Brilliant companies, as we know, that are building IT infrastructure, building beautiful facilities. And we work hand in hand with them to make sure that the fluid networks and the racks, and everything within, connect.
Dave Vellante
>> So this notion of a connection integrity is interesting to me because I remember the first super computing event I ever went to... And of course I remember taking tours of the Cray factory back in the day. You would see this mesh of hoses and it was kind of scary. Now today, you're seeing these highly engineered systems. They've got these beautiful, I don't know if they're chrome, but connectors that are, they look like they're easy snap on.>> Absolutely.
Dave Vellante
>> So what's the state of that technology today?
Scott Sloan
>> Yeah. So I'd say it's really encouraging. I mean, even since we met last year, the advancement, we always say AI world, it's like 7X. If you've been here for one year, you've been here for seven. So we're seeing a lot of changes in how the importance of how fittings and hoses go together, understanding the different designs. We've also seen a big push on the different alloys, the different tubes that the hoses are put together, big how the media interacts. So I'd say for the 20 or so years that we've been involved, we were never really in the rack. We were that secondary network coming down into what we'd refer to as a chilled door, a rear door heat exchanger. But now that we're coming into the electronics, it's increasingly important.
Dave Vellante
>> So I'll explain to the audience why we're here. And all you have to do is look at data center spending. And so the spending on data centers, and we're talking about the facilities, the power, the cooling, the compute storage, and networking, perpetually, it's been a $200 billion business. We just did a forecast on this. My colleague, David Floyer, put this out. About 200 billion, maybe popped up to 220 billion. In 2024, it jumped up from 220 billion to 350 billion. And it's now growing at about 16% CAGR is the forecast for the next 10 year. And of course, the AI piece of that is growing in the mid 20s. And we're talking about compound annual growth. So there's a massive growth, maybe close to a trillion dollars by the end of the decade. And you compare that to or position that next to the CapEx spending that's going on. Just last quarter, all the big three hyperscalers and Meta doubled down on their CapEx expenditures. So we're talking about north of 400 billion just for the top four hyperscalers that they're going to spend this year, next year, just in a year, next year on CapEx, 400 billion. So the artful plumbing underneath that has to work. So what's the state of that? And are you comfortable with the state of-
Scott Sloan
>> Well, exactly. Because when Lauren talks about connection integrity, it's not a joke. If you think, when I was in school, in college, back in the mid '80s, what was the single defining moment of the decade? And a painful one was the Challenger explosion, like the space shuttle. And that is all about the O-ring. That was out of the-
Dave Vellante
>> That is one little part.
Scott Sloan
>> Tiny little rubber O-ring.
Dave Vellante
>> That created that disaster.
Scott Sloan
>> Because of the temperature that was out of spec when they launched in January of 1985, let's make it up, in temperatures in Cape Canaveral that were incredibly low. So it's serious.
Dave Vellante
>> So it deteriorated the O-ring.
Scott Sloan
>> Exactly. Sadly.
Dave Vellante
>> That little part. And sadly, it created a disaster.
Scott Sloan
>> Exactly. I think it's important when we talk about the CapEx, all of the... It's almost a private... I mean, the scale, it's almost like a private version of a war footing in terms of investment that's marshaling to this moment. And I feel like if we're here to talk about the importance of making sure that connection integrity is actually looked upon as a primary axis of performance, and that's what we're trying to do.
Dave Vellante
>> Right. So that O-ring example is a good one. Could you guys bring up the graphic? I want to share with people the value chain, if you will. So this is just one view. There are a number of views on this, but this was a pretty good one that we found. So you can see, of course, everybody talks about the semiconductors and you see NVIDIA and the other ones. By the way, I would have NVIDIA virtually everywhere in here. They're funding and innovating across a lot of these areas. But I'm going to work from the top down to get to you guys. So you've got the semiconductor piece, the high bandwidth memory is really critical now. You've now got CPUs, as well as GPUs. You've got the deal from NVIDIA and Intel, sort of merging those two worlds. And then you've got the compute storage and networking. So Dell, HPE, and others, Broadcom's doing networking. Of course, NVIDIA, Cisco, Arista. And then the storage folks, that's a key player. I'd say sort of compute and networking are sort of 1 and 1A and then storage maybe a little bit behind that. You got the operators like the colos over there and the middle left. And then data centers, these are companies that are building data centers. So the big hyperscalers, Crusoe we've had on here, CoreWeave of course. xAI, OpenAI now in that business. And to the right of that, you can see Crusoe, Lambda, we've had Nebius on and CoreWeave and a number of others. Energy is the big gate everybody's talking about. And then you got the industrial equipment, Schneider Electric, Vertiv, their stock's going crazy, Siemens. And then the cooling piece, and that's where you guys come in. Thank you guys. Things like secondary fluid networks. So can you explain what is a secondary fluid network?
Scott Sloan
>> Absolutely. So it's interesting. We operate, I love your finish on the cooling side there.
Dave Vellante
>> Yeah. Bring that back one more time and while Lauren's explaining secondary fluid networks, which would be kind of in that lower right, although there's a primary fluid network, which is piping in lots of, well, you need water, land, and energy to build a data center. So please take it from there.
Scott Sloan
>> Absolutely. So you're seeing these hyperscalers, we know who they are. You've got the logos up there. So they're bringing fluid right into these data centers. And in a lot of the AI factories, we have a luxury. They're new. We can design them however we want. And I'll tell you, that's a lot more fun than some of the greenfields we've had to work with for another day. But after that, where does the fluid go right within that AI factory? So our role is how do we bring it from that secondary fluid network that distributes down a particular sector or distributes it through a particular row of racks and then into these various cooling industrial equipment companies. And the exciting challenge, I think-
Dave Vellante
>> Thank you guys for that....
Scott Sloan
>> you talk about master plumbing is it's not uncommon to see a hyperscaler have a standard. But then within that particular AI factory, you'll see Schneider or you'll see Vertiv all in there together. So how we bring that together, and again, you're trying to standardize, but at the same time, we're moving so fast, we have to use what we have. So we're always adapting and it's exciting.
Dave Vellante
>> So the high volume of water comes in, and then that's where you and your ecosystem pick it up, and then that cools the rack actually down to the direct liquid cooling .
Scott Sloan
>> It can go into the server blade itself. So from CDU to blade up from through the roof, chillers on the roof, all the way through, that's primary. And then as it goes to sets of racks, that's where the secondary fluid network kicks in.
Dave Vellante
>> And CDU is chiller distribution network, like PDU is power distribution network. CDU, there's a chiller version of that.>> Correct.
Dave Vellante
>> Okay. And then->> And what's interesting too, just on that is years back, we were maybe extracting hot air out of a rack and now we are chilling power distribution units. We're chilling Chips, we're in the blade, we're everywhere. And that's where, back to that O-ring analogy, the media is touching a lot of different places throughout that system. So we can't just be worried about cycling from a fluid network into a door and out. We're touching high value NVIDIA electronics.
Dave Vellante
>> We know that, well, today AI factories, pretty much everyone is a custom built system. But to scale, you've got to have standards. What's the state of standards in secondary fluid networks?>> Yeah, I'd say it's advancing well, it's a balance. You've got the hyperscalers that kind of have standards within themselves, which is helpful.
Dave Vellante
>> Each hyperscaler.>> Yeah, each hyperscaler has a standard. I presume each one is the best.
Dave Vellante
>> Yes, sir.>> And then within it, it's very hard to standardize. The fluid networks, again, when you think of the biggest challenges, how do we go from that fluid network into that rack? Those racks could be very different. The servers within could be very different. Dell is a very common one. So we're standardizing as fast as we can.
Dave Vellante
>> Or where you can.
Scott Sloan
>> Yeah, where we can. And I think the big areas we're seeing it, which are very encouraging, is cleanliness, testing, and packaging, how we get it to these data centers quickly.
Dave Vellante
>> So as you know, Meta and Microsoft, and there may have been some others, but those were two of the big players in OCP that standardized on a lot of the hardware and they open sourced their Facebook, the way they built data centers. Are the hyperscalers working in the industry working together to create these standards, or is it like stovepipe standards?
Scott Sloan
>> Look, I think it's a nascent is what we would say. Honestly, the whole AI factory concept and the modularity and the standardization around it is at a nascent stage. So I think what Lauren's articulating is all the way through in the various elements, that standardization challenge is an ongoing one. And so our job when the need for speed and scale in that environment is to be as adaptive as we can be and try to shape where we can, those standards as best we can.
Dave Vellante
>> Okay. So what is CoolFlow? Is that a brand? Is that a separate company?
Scott Sloan
>> CoolFlow is a liquid cooling division of our company.
Dave Vellante
>> Okay.
Scott Sloan
>> Which is extremely relevant obviously to these conversations. But CoolFlow also, of course, is a product portfolio that's curated specifically for what we're talking about in those secondary fluid networks. So that's part of our product portfolio approach, and that would feature some of the best manufacturers of componentry like Danfoss, for example, and a variety of best in breed suppliers. And we integrate that together and bring that to bear in the data center.
Dave Vellante
>> So we have made a big leap from that Cray computer example that I gave you with the mess of hoses to where we are today. Where do you see this going? Are you optimistic that we are in a good place now or do you have consternation that there could be another O-ring like disaster that where we're supporting all these critical infrastructure? Go ahead.>> Yeah. I mean, I think we're very optimistic. I think what this industry gets right that I'm not sure a lot of others do is collaboration. We've never experienced, we operate in five or six other verticals, but never experienced an industry that is open, is collaborative. In some of our other markets, you would never see some of the big server companies collaborating for the greater good. So we're very optimistic, but I'd say we're still very cautious because there's a lot of importance that is going into how this goes together.
Scott Sloan
>> Well, I think just to underscore the importance of connection of integrity and fluid conveyance, Lauren, you should talk about that incident where we sent some very important hose assemblies, where did they end up in the AI factory in that particular day?
Scott Sloan
>> Oh yeah. These particular ones, they ended up in a situation where there was some contamination in them.
Scott Sloan
>> Well, what I'm saying is the point really being they ended up back in the backshed of the data.
Scott Sloan
>> And that's where the contamination began.
Scott Sloan
>> So my point is-
Dave Vellante
>> Oh, with the weather.
Scott Sloan
>> Exactly. So my point is when you put something like hose assemblies, something as vital as providing the liquid cooling arteries out in the back, that underscores the importance of, hey, we got to tighten up.
Dave Vellante
>> So somebody saw hoses in the box and they said, "Oh, this must go outside."
Scott Sloan
>> It must go outside, it's a garden hose. And that's the point where-
Dave Vellante
>> Well, yeah. And there's your O-ring example where the-
Scott Sloan
>> So there's an education. There's an education around like this, we appreciate the opportunity to talk on behalf of SFN, the importance of it, and driving the standards and performance and execution and scale of that. And that's what we're trying to do in our own individual company. And that's what the challenge is for the industry.
Dave Vellante
>> Well, I think too, I use NVIDIA as an example a lot because they're so ubiquitous in this industry and they've got this value chain, there's this value chain that we showed earlier. And they may be touching major portions, of course, of that value chain, but each layer of that value chain has its own supply chain that it has to worry about that you're describing. And own set of engineering standards, engineering best practices, standards, capabilities that have to be developed.
Scott Sloan
>> You're so right. And if you think about it, as far as where I view it, those companies who take an integrator type of approach and perspective win. That's what, for me, Jensen and what NVIDIA have accomplished is not GPU. It's the network and the systemization of what you're talking about. And the Dell, for example, they're looking end to end solutions. I think that's what, in our secondary fluid network world, that approach, that mindset of systemized, integrated channel partnering has to elevate to meet the standards that Jensen and the industry are in.
Dave Vellante
>> Well, the reason we talk about unsung too is because it's like the dirty jobs.
Scott Sloan
>> We like getting dirty. That's right.
Dave Vellante
>> It's the center and the line or-
Scott Sloan
>> Picks and shovels....
Dave Vellante
>> I don't know. The artery, you talked about the arteries before. You're not the brain surgeon.
Scott Sloan
>> Brain surgeons, they're the quarterbacks on the ... They get paid the most. Heart surgeons, plastic surgeons.
Dave Vellante
>> Plastic surgeons are up there on that.
Scott Sloan
>> Number three.
Dave Vellante
>> They're in Hollywood.
Scott Sloan
>> They're up there. Only in America. And then vascular surgeons would be below. So we were looking for a little respect for the secondary fluid network.
Dave Vellante
>> Where do you guys want to be in, I don't want to go up five years, it's almost too far. Where do you see this industry, these AI factories in two to three years, let's say?
Scott Sloan
>> I'm going to swing it that way.
Dave Vellante
>> Yeah, absolutely. I mean, I think to Scott's point, you're just seeing even as recent as when we last met at Supercompute, everybody was operating in their own little quarter. The blade guys were concerned about the blade and then the rack guys and the header guys. I think in the next two years, we're going to continue to see that end-to-end collaboration. We've seen it through acquisition. Schneider, for example, but that's how we're going to scale this appropriately. You think of it as a funnel, Dave. The secondary fluid network is six inch, eight inch piping, down to two inch hoses, interact manifolds into tiny little tubes in the blade. And the more we communicate and align, the better we're going to be. So very bullish on where we're going to be probably a year from now.
Scott Sloan
>> I think there's many companies who've shown us the way towards that type of collaboration. This industry, David, for us, is a breath of fresh air in how to do the business of business more effectively.
Dave Vellante
>> Yeah, you were a 50 year old company. Back then, there was water cooled mainframes.
Scott Sloan
>> You remember that?
Dave Vellante
>> IBM. I do remember that as I've seen virtually every wave in the tech industry. But well, it's ironic because when you... IBM was everything in this business and mainframe was the leading product. And you couldn't sit through an IBM presentation without them taking you through the thermal conduction machine that was the fluid network. And now it's come back.
Scott Sloan
>> The irony is it has come back full circle.
Dave Vellante
>> It really has. And so anyway, guys, really appreciate this.
Scott Sloan
>> Really appreciate the time, Dave.
Dave Vellante
>> .
Dave Vellante
>> Thanks for coming on and helping educate the audience on the importance of the artery here and the unsung heroes of AI factories.
Scott Sloan
>> We appreciate the time.
Dave Vellante
>> This is the AI factory series, NYSE Wired plus theCUBE. I'm Dave Vellante with John Furrier. Thanks very much for watching. We'll be right back right after this short break.
Dave Vellante