At Project Production Institute’s ninth annual symposium, Todd R. Zabelle, founder and president of Strategic Project Solutions Inc, and Hunter Newby, owner of Newby Ventures, broke down ways that data centres can tackle supply chain disruptions, and what exactly is causing these issues.
Both Zabelle and Newby outlined the main issue: Supply chain disruption impacts revenue and through that, a company’s share price is affected.
“There’s a lot of consternation in the data centre world about the ability to get supply when it’s needed. It is definitely impacting share price…there’s a lot of pain going on right now,” Zabelle said.
He broke down the types of supply into three categories: Construction materials such as concrete, equipment like generators and items that go into data centres such as CPUs, and lastly, consumables.
In addition, there are three types of disruption when it comes to supply chain issues: force majeure, false demand, and capacity constraints.
An example of force majeure is the conflict in Ukraine. The Russia-Ukraine crisis has had an impact on auto parts, oil, grain, iron products, and more. In addition, a report from Dun & Bradstreet noted that more than 600,000 businesses worldwide rely on Russian and Ukrainian suppliers.
“What we found out was that one of the largest neon production plants in the world is based in Ukraine, and it was shut down as soon as the conflict began. And there was so much concentrated there that all the other sources of it couldn’t make up for it, because they could only increase production by one percentage point…This created a domino effect of problems that took several months to actually get baked in. But when it finally did, it slowed down supply chains,” Newby noted.
Another example of force majeure could be labour action such as port strikes, suggested Zabelle.
When dealing with force majeure, there are a few options. Zabelle said you can buffer with inventory, buffer with capacity or, the most “wise” thing is to combine both options.
“You have to have the ability to model these things and analyze them properly, and simulate to figure out what would happen if we get exposed to one of these, because, inevitably, you’re gonna get exposed to some of the stuff.”
When it comes to false demand, some examples can include strategic or competitive moves from companies and supply chain over-purchasing.
“It becomes a game of ‘win as much as you can win’,” Zabelle said.
He added that another way to look at false demand is the “matching problem.”
“If you have a bunch of generators, that’s probably good, but they’re kind of useless unless you have buildings to put them in, and all the other stuff you need to actually operate a data centre,” he noted. “So how much of the stuff are you going to buy? And then with some of this stuff, it gets outdated pretty quickly because of the technology refresh cycle.”
Zabelle said the general hypothesis is that there’s more supply out there than we know.
“We’ve been out to the majority of the players that supply the data centre industry, on behalf of various data centre companies that have built the production system models, to see what’s going on there. We’re seeing a lot of inventory… Our hypothesis is not something that we dreamed up or challenged. It’s actually because we have the data from looking around out there.”