Dell Technologies is up a staggering 260% this year, touching record highs. The company does report quarterly earnings this week at a time of sustained demand for AI-DIN data center servers. Wall Street is projecting revenue of over $45 billion while looking closely at how Dell executes each against its $51 billion AI server backlog and also manages complainant supply constraints. Well, joining us to separate the hype from enterprise reality is Don Gentile, analyst at HyperFrame Research and a former executive at IBM and HP. Don, great to have you here. Thank you so much for joining me.
Thank you, Remy.
Well, we are officially rounding out earnings season and all eyes are on Dell as we head into the rest of this week. So what metrics does Dell actually need to hit in order to justify their valuation?
Well, the good news is we're already past NVIDIA. So a lot of the demand that's already been established in the marketplace, everybody knows that that's coming. So Dell doesn't have to say that there's demand. What we're looking for is them handling the shipping against that server backlog. Right. So they said that last quarter they had a 51 billion dollar backlog and they shipped $16 billion of AI servers against that. So their ability to meet that demand and then also be able to manage their margins in doing so are the prime metrics that we're looking for.
So Don, you mentioned metrics as well as backlogs. So when it comes to shortages, tell us what the reality is.
Yeah, so we have memory shortages. We have storage shortages. Those constraints are not going to go away anytime soon. The good news for Dell is that they have been in the supply chain business for a really long time. They are decades of expertise. They have really strong partnerships. They understand how to bring components to marketplace and meet the market demand as it's happening. So when you talk to Dell, maybe they've been talking about it for 20 years, but people are finally paying attention to supply chain constraints now.
Yeah. And as an executive who has been at HPE as well as IBM, I do want to get your perspective on margins. So tell us what you think is going to happen and what we should be focused on.
Well, the components are becoming very expensive, right, inside of these AI servers. So they are razor-thin margins. When you look at the ISG group last quarter, they were 10.5% margins. That is not good enough, obviously, for the market. We want something between 18 and 20%. And so other parts of the portfolio have to pick up the margin difference in this case. Storage, which is an area that I cover very closely, has excellent margins. Now, NVIDIA has 75% margins. It's a massive difference. When you're talking 18 to 20 percent, the mix now of when Dell is shipping these AI factories, which is integrated storage and servers and power and cooling and software and services on top of that, that's where they're able to extract the better margins than just the AI servers themselves.
Yeah, and artificial intelligence is something that all of us are paying attention to, especially here on Wall Street. So when it comes to Dell's core business, as well as AI PCs, give us your outlook.
Well, they told us last quarter that their traditional servers are also up 92%. So there are times when an AI workload does not need an AI server. You can run it through a traditional server. And so that's important. They also have the commercial PC business, which is a massive, you know, basically a third of the company is also commercial PCs, right? So that's an important part of the mix. It keeps their footprint in the enterprise as well. AI PCs are a very small part of that today, but I think that that'll be a growing percentage over time.
And while I have you here and while we're talking about infrastructure, I do want to get your perspective on the ROI timeline. What do you think is realistic?
Well, so we do primary research at HyperFrame. We talked just last quarter with 520 infrastructure and operations executives around the world, and they told us that only 23% of them are seeing production ROI from their AI projects today. So that's not good enough, right? We have to see better ROI numbers, those hopefully that will go up. We refresh that every six months. And so every time we go out and talk to the customers, we're going to need to see that number going up over time.
And we're here at a time when we're looking ahead to Dell's earnings, but we have heard from the hyperscalers as well as many of the companies in this ecosystem. So what do you think we need to keep in mind as we weigh Metrix's valuation when it comes to a lot of the key players in this space?
Well, one of the interesting things is that for Dell, for example, regardless of whether customers do their compute on-prem or whether they do it in the cloud, Dell winds up winning either way because they are supplying full rack-scale infrastructure to enterprise customers as well as the hyperscalers, Neo clouds. They shipped an NVL72 rack to CoreWeave last month. It was a big deal. And so they're able to put together these fully integrated rack systems that helps customers to deploy quickly and get their AI systems up and running.
Well, Don, it was a pleasure having you join us here. Thank you so much for your perspective as well as all of your insights today.
Thank you, Remi.
Thank you.
