Arman Khalil, Evocative Data Centers: “It Makes Very Little Sense to Develop in California, Unless You Already Have an Existing Location.”
LOS ANGELES, CA – The Southern California data center arena isn’t exactly the easiest to sum up in a sentence or two. It’s got its strengths as well as its weaknesses, but a lot of firms want to be there. So CapRE’s Seventh Annual Southwest Data Center Summit: The Telecom Evolution kicked off with a panel discussion called Southwest Data Center Market 360: Has the Outlook for the Los Angeles Market Changed with Advent of Edge, Micro Data Centers and On-Site Power Generation?
“I’m going to kick this off by talking about the cost of development in the Los Angeles market and ask Arman [Khalil] first about his experience and outlook today on competing in this market and the emerging sub-markets, as well as the obstacles that one finds impeding progress in terms of development,” began Moderator Michael Siteman, Director – Chapter Relations, 7×24. “Then, we’ll move around the panel and get other perspective.”
“Thanks, and good morning everyone,” replied Arman Khalili, CEO of Evocative Data Centers. “I started my career actually in the early 90s in the IT world. Back then it was $1000 to $1200 dollars per square foot to build. And that was at that 35-50 watts per square foot. That was a significant number. You’re looking at almost $10 million dollars to build in these locations. It’s still a lot of money, except now, if we look at it from the residential side, that’s actually a pretty good deal. If you can get a $1000 per square foot house in Silicon Valley, you know, you did well.”
By Khalil’s estimates, it’s the same situation in Los Angeles. “I just looked at, at a show yesterday, a million dollar house or something, a bungalow in Venice, which sold for $7 Million USD! And I mean I would have rented it for a few hundred dollars a month or something, but I guess someone wanted to pay $7 Million USD,” he shared. “From that perspective, it’s pretty cheap and it actually makes very little sense to develop in California, unless you already have an existing location.”
“We tried to develop a site in San Francisco,” shared Khalil. “It was a tiny, 1.7 million square foot site, and we just had so much resistance from the neighbors. So that town has then translated into other locations, like Santa Clara, where people are paying $1.2 or $1.5 million USD on townhomes, and they just don’t want a generator next to them. I get it. I happen to own a small generator in my home but I’m the exception to the rule. From a development standpoint, we still have 22.5 million people living in Southern California – it’s the second largest market. Those eyeballs need to be fed.”
“Regardless of what transpires, or what capacity exists or can be developed, it’s going to be satisfying the current market,” Khalil predicted. “So I feel like most of the hyperscalers are going outside of the California market, with the exception of Santa Clara. Santa Clara still has a few spots left that people are developing but Reno and Las Vegas and Phoenix are all trying to satisfy the California market with 10-12 millisecond latency.”
“We’ve got about seven sites in five markets,” he shared, preparing to conclude his remarks. “We have two sites in Silicon Valley, two in Los Angeles, one in Phoenix, Dallas and Virginia. And we’re expanding. So we’re looking at this strictly from satisfying an audience perspective. And also from a retail perspective, the hyperscale [market] is quite a competitive landscape. The way we look at it, the underwriting does not justify the investment being made.”