Cto Insights Ep 1 The Infrastructure Crunch

TITLE: CTO Insights Ep. 1 | the infrastructure crunch | Data#3 CHANNEL: Data#3 Limited DATE: 2026-06-02 ---TRANSCRIPT--- [MUSIC PLAYING] I’m Graham Robinson, Chief Technology Officer here at Data3. And this afternoon, it’s my immense pleasure to sit down with Carl Salder, Chief Technology Officer, Cisco ANZ. Thanks for having me, Graham. Carl, why don’t we start with the topic that’s really dominating every conversation that I’m having right now, which is the DRAMP supply chain issues. Given the supply constraints, organizations think or rethink their investment strategies, given the current challenges. It basically comes down to a matter of priorities with the shortage on DR memory. That’s going to result in with the total supply and demand, the prices are going to be going up. It then is a matter of figuring out where the most urgency is in the business that needs that initial investment based on the products you’re buying and the memory that’s kind of embedded in that is where you’re going to get your biggest bang for your buck. It essentially comes down to priority enough. Spoken to a lot of company executives who are going through that reassessment, trying to figure out where they’re going to be applying those investments. Typically, in the past, you might be buying some gear for x dollars. You might be now paying 3x or x5x for the same gear. And as a result of that, that’s going to stretch budgets. You’re not going to be able to buy as much as what you want. You have to then be very particular about where you’re going to be applying that investment dollars and where the business needs it the most. So it’s coming down and working with the business stakeholders to figure out which part of the business is going to need that investment the most and then directing the money initially there. So for those who haven’t had a chance to really get deep into the challenges that the industry is facing right now, what’s causing the challenges and which parts of the industry is it affecting? It’s affecting many, many people. But ultimately, it boils down to AI, good old AI. So AI has some very prescriptive performance and latency requirements with the explosion of AI across the board. There’s a demand on one aspect of memory we call high bandwidth memory. And when you look at the manufacturers who are making high bandwidth memory, they have manufacturing lines which producing x amount of output. And with the high demand on high bandwidth memory, understand it’s also potentially going to generate a bit more revenue and more profit for them. They have the opportunity to figure out how they’re going to be applying their manufacturing processes to. Is it the standard memory, the DDI memory, or the high bandwidth memory? And because of that high demand on the HBM and the fact that they’re going to get more money, they’re starting to shift their manufacturing lines to produce more HBM, high bandwidth memory, at the expense of the more generic DDI memory that we use. So DDI memory is used in everything. It’s used in a lot of the computer appliances, the laptops, the phones, the servers, network devices, fridges, freezers, microwaves, all that is using it. Because the supply is being constrained on that, because they’re producing more HBM memory, which still, by the way, they can’t meet the demands of what’s being asked. So they’ve literally gone from a situation where they had like a shortage of HBM and DDI was OK to now having shortage on both. And as they start to scale up manufacturing facilities, there’s going to be a lag in time while that shortage is going to be in plan. So we’ve seen an impact, brutally impact, some of our customers, particularly in the compute space. And then the conversation quickly with the customers goes to, well, what do I do? And so, yes, there is still some availability of RAM. And there are projects that are moving forward. But some customers are looking at cloud. So we went cloud. We came back on a pram. Is the answer, go cloud again. And if it is, my concern is, what happens if there’s not enough cloud? It’s— you’re kind of caught between a rock and a hard place here because everyone is constrained, even the cloud providers, with that move and demand on cloud resources, that’s also going to force them to bring in more compute. They’re going to be up against the same challenges and struggles. So I think it’s a really interesting time for the industry right now. Everyone’s trying to figure out the best path forward. Cisco is trying to figure out what our plan is. I’m sure all the vendors out there, the cloud providers, are also trying to figure out the same thing. But ultimately, as I said, it comes back to that demand that AI has generated that’s been the trigger that’s resulted in this. I think we have to see how things pan out. I know that many companies have very stringent contracts around supply of memory. But the challenge is, what if the supply of that memory is not enough to meet the demands? And that is the gap that I think everyone’s trying to figure out, how they never go around. So for the foreseeable future, I believe customers need to stay very close to their vendors to understand the later than what’s happening with that vendor. Because literally, even for us, things change on a week by week basis. We have plans. And based on availability, based on supply, the opportunity for what we can do in terms of delivery is forever changing. So I think it’s important that customers stay close to their vendors to keep track of that, especially if they have deadlines of time in place that they need to deliver to support business expansion.