Peter Chapman: So I certainly QML is the €“ what I would describe is the leader, especially maybe in the €“ even in the short-term because we have seen already even on some of our older equipment kind of very interesting results that suggest that and I’ll get a little technical that the number of features that you need to train a ML algorithm is several orders of magnitude less than what you would need for a classical ML solution. So you’ve probably heard some of these large language models take hundreds of millions of features. We haven’t gotten to that scale yet, but for the things that we’ve been doing for €“ for instance for image recognition we’ve shown that we can get those same kind of results but with far fewer numbers of features.
So that’s important because that we’re not sure that what we can get to much larger feature sets in classical hardware. And the other thing that’s really interesting on the ML side is that again in a small €“ small study is that the ML models that get created on the quantum computer seem to capture the signal better than you see in a classical system. So some of the work that we’ve done on the financial side for instance with GE research it seemed to do a better job and some of those are the most interesting data points because they might be an outlier and a Black Swan event, and so a classical ML might be missing that. And quantum seems to capture that relationship when it’s creating it; so all of these things are very promising. If it turns out the ML that you can do on our machine means that it can do a better, ML algorithm, not only could you do better kind of ChatGPTs but better advertising or anywhere ML is being used.
So that’s certainly a large promise. Chemistry also is working well, but ultimately for the true promise of chemistry to do simulation in chemistry, we’re doing small molecules today, but of course, you will need a larger quantum computer to do things like cure cancer and those kinds of things. So those are still a little bit ways away. But even in these small molecules, for instance the work that we were doing with Accenture is interesting that was working on PFAS and ways to be able to break down those forever chemicals. So it’s still very interesting work.
David Williams: Okay, fantastic. And just one more quick one, if I may. Anything from the Garfield versus Boxed decision and I know you filed a petition, but anything new on that or what the impact could be?
Peter Chapman: No, we have no updates today on that particular one.
David Williams: Okay, thank you.
Operator: Our next question comes from Richard Shannon from Craig-Hallum. Please go ahead with your question.
Richard Shannon: Well, hi guys, thanks for taking my questions as well. Maybe I will ask you a question on your goal of getting to an AQ of, I think, 29 for this year up from 25 this year. What is the long poll in the tent in terms of achieving that?
Peter Chapman: Well there are a couple of things here. First, is we have multiple systems which are going for that goal. So, that’s the kind of first part to it is we have the systems up and running and so at this point it’s tuning the systems. You can kind of think of it as a high performance race car, car is assembled, but you still €“ there is always work to do to kind of get the actual performance out of it. So that’s what we’re doing right now as we speak.
Richard Shannon: Okay. Let’s hear, a question on the manufacturing facility here. Especially in light of increasing interest pulling forward this capability for people or for customers that are potentially buying whole systems here. When do we kind of see this first stage of capability up and running here? And what kind of annual throughput in terms of number machines do you expect when you get there? And kind of what’s the capability of the current footprint over the long-term?
