Now what’s going to happen next is since we can soon expose wafer, every customer are going to come to see us here to get access to the tool we have in our lab and start to expose their own vertical, so that they can decide for themselves exactly how to use the tool. And we expect that this works will lead at some point to, I would say, the next set of decision on High NA. So what’s happened in the lab in the next few months, the work we’ll be doing with all our customers, and I stress again, all our customer, I think is most probably the most exciting milestone to come and this will really help everyone to understand basically what’s next for High NA. But I cannot stress enough on — I will say how happy and excited that we’ve been able to generate those first images.
When you look at the undertaking of a technology like High NA, for us, for our customer, this is a very, very important achievement. And again, the next few months, we’ll build on that.
Francois Bouvignies: Thank you very much.
Operator: Thank you. We will now go to the next question. And your next question comes from the line of C.J. Muse from Cantor Fitzgerald. Please go ahead.
C.J. Muse: Yes, thank you for taking my question. And Peter, big congrats. Pleasure working with you. And maybe if I could direct the last question for you here at ASML, given your long experience. As you look at the strong secular trends led by high performance compute, historically at the leading edge, it was led by Apple and Huawei and then only Apple post the embargo. But now given the performance power cost requirement, how are you seeing kind of the trends on the high-performance compute side moving closer and closer to kind of being the pipe cleaner for the bleeding edge? And how does that impact your thinking of what kind of longer-term growth for leading edge workers will look like?
Peter Wennink: Well, I think it’s two questions, so I answer the second one. You’re basically saying, what will drive leading edge high performance compute, but you’re absolutely right. I mean, when you think about high performance compute, and especially in the context of AI, and I’ve said this many, many times before, AI is driven by massive amounts of data and about also understanding the correlation between those data elements and then overlaying that with smart software. But — and I also believe actually what I’m seeing and what I’m hearing is that IoT in the industrial space will actually be an — will be an area where we will see a lot of AI applications. Well, in order to collect all that data, you need sensors, because you have all kinds of examples, whether it’s the car or whether it’s life science, medical equipment, it’s about sensing.
And that is really the domain of mainstream semiconductors. I don’t like the word mature, like it’s something which is old fashioned that you don’t need it. It is mainstream and it’s critical in the — I would say in the — I would call it amalgamation of mainstream semiconductors and I would call advanced semiconductors. And this is also why you cannot distinguish the growth of one against the other. I mean, you need both. And I think what you will see is that the growth of the industry, and especially high-performance compute will be driven by the value that is going to be created. Yet, cost is a significant issue, but Moore’s Law is an empirical economics. And when you create more value of the transistor, the new transistor, the next generation transistor, then it costs, then you’re going to grow.
And when you think about AI and with some of these examples and especially in the software space where you see productivity, just calculated productivity as voltages of 30% to 50%, then the value of the next generation transition will be huge. Now, if you ask me, Peter, what’s going to be the next killer replication? I’m going to give you the same answer as I gave you for the last 25-years and I have no clue. But what I do know is that when the value of that transistor, next generation transistor, whether it’s — and it’s particularly driven by high performance compute, its energy efficient performance, if we bring that to life and we’ll bring it to life together with our customers because we have High NA and potentially Hyper NA. Then the — I would say, amalgamated consolidated demand for also mainstream semiconductors will go up.
So this is what I believe and this is what I see when I talk to customers and I talk to actually users of the value in the space. And this is why I’m so confident about the long-term future. And it’s all connected, so you cannot distinguish the two.
C.J. Muse: Very helpful. Thank you. And Roger, a follow-up question on your backlog comment around the EUR4 billion plus required to hit the midpoint over the next three quarters. I guess as part of that, if you had to think about the higher end of the range, I think that would add two plus per quarter. And then also, if I look at your backlog today, excluding High NA, it’s still sitting at roughly 18-months. And so obviously I would expect your backlog exiting ‘24 to have tools that will be shipped in ‘26. So is EUR4 billion the right number we should be thinking about for the next three quarters or should be it — should it be significantly higher?
Roger Dassen: So if you’re looking at the midpoint, it is EUR4 billion. And you shouldn’t underestimate what we’ve already in that calculation taken out for the post ‘25 period. And I’m sure you’re all going to do the math, right? Because it’s not all that difficult to do the math. And then you’re probably going to figure out that that’s a pretty healthy number and probably a number that maybe exceeds a little bit what you currently have in your models in terms of High NA and that’s okay. But that’s a calculation that I leave up to you to make, but that has clearly been recognized in our calculation. So we’ve taken out whatever is for ‘25, and our focus of the EUR4 billion really is what pertains to 2025. And to your question, C.J., if you’re looking at the high-end of the range, there’s a EUR5 billion delta last I saw between 35 and 40.
