Non-GAAP net loss for 2023 was $16.6 million and compares to a loss of $14.1 million in 2022. And as with our GAAP results, this was primarily due to increased R&D expenses. The differences between GAAP and non-GAAP operating expenses in all periods presented are primarily due to noncash stock compensation expenses which were approximately $4 million in 2023 and $3.4 million in 2022. Our balance of cash, cash equivalents and short-term investments on December 31, 2023, was $19.5 million compared to $21.2 million at the end of 2023 and $20.4 million at the end of Q3. During the last year, we used $14.6 million of cash in operating activities and we sold approximately 1.8 million shares under our ATM facility at an average price of $7.97 per share.
Of those amounts, approximately 320,000 shares were sold in Q4 at an average price of $7.37. As of December 31, 2023, we had 26.1 million shares outstanding. As Scott mentioned, we met the first 2 milestones under the ST license agreement during Q4, resulting in $550,000 of license revenue. The next grant of license rights to ST will be our distribution license which will enable them to both manufacture and sell products with MST. Those sales will result in royalty payments to Atomera. While we are not able to share the financial details of the ST contract, the terms are consistent with our target model. Under this model, total upfront license fees have a list price of over $3 million, with payments increasing as we grant customers additional rights.
As Scott explained in his remarks, timing of entry into Phase 5 which is when we will recognize revenue on the grant of the distribution license is largely under ST’s control. Accordingly, I’m not providing guidance on the timing for recognizing that revenue. I expect that our Q1 2024 revenue will consist only of ratable recognition of MS TCAD licensing. Moving to our expense guidance; our non-GAAP operating expenses for 2023 were $17.1 million and we expect that in 2024, our non-GAAP OpEx will be in the range of $17 million to $18 million. While this is a wider range than I provided in previous calls, this is due to the uncertain financial impact of moving from TSI semiconductor to a new foundry. Our work with TSI wrapped up in January of this year.
And as a result, our R&D expenses in Q4 declined from prior quarters in 2023. We’re making good progress in talks with possible replacement providers for foundry services but nothing has been finalized yet. While the interruption in foundry work will cause Q1 2024 R&D expense to decline further from Q4, we expect to incur some one-time fees as we transition to a new foundry and we will update our guidance when we have more visibility. With that, I will turn the call back over to Scott for a few summary remarks before we open up the call to questions. Scott?
Scott Bibaud: Thanks, Frank. As you can see, we made great progress towards commercialization this past quarter and in 2023. We are doing everything in our power to get ST to production quickly but we also have an incredibly valuable portfolio of other potential customers who we are working to take into the commercial stage. Our team is confident that it’s only a matter of time before we can announce license deals that will further solidify the potential of that Merus business for the future. Mike, we will now take questions.
A – Mike Bishop: Okay. Thanks, Scott. [Operator Instructions] And right now our first question comes from Richard Shannon of Craig-Hallum. Richard, please go ahead.
Richard Shannon: Thanks, Mike and thanks, Scott and Frank, for getting me on your call here and congratulations on the great success of indels year here. Maybe a couple of questions related to STMicro about the time frame to productization of 1.5 to 2 years here. It sounds like you’re talking about some very interesting work with them. I don’t want to get out ahead of my skis here but there’s any reason to think that, that time frame could be at the lower end of that just because of all the engagement you’ve been having. I don’t want to read too much into it but I just want to get your thoughts there, Scott.
Scott Bibaud: Yes. As I said, we can’t predict the timing of it. They haven’t actually given us a schedule that they’re trying to hit. However, I will say the progress we’ve made in just the last couple of months, we did — since we did the installation is really quick in our experience and they have the possibility of moving very fast. Once MST is installed inside someone’s fab, they can literally crank up the engines and turn out these new wafer experiments very quickly. So, there is some chance, I would say. I would say I’m still holding to the 1.5 to 2 years from our announcement date last May but there’s a chance it could be on the earlier end.
Richard Shannon: Okay. When you say have we gotten the schedule from STMicro, would you expect that? And what kind of — any way you can characterize what that means? Is that a normal process of bringing up any Sengupta process? Or is there any specifics to Atomera that create the deviations from the normal, I guess?
Scott Bibaud: I don’t think there’s anything that’s a deviation from the normal in this case. That being said, I am not privy to what they consider a normal development time frame for an analytic product like this. I know that for some of our customers, they may have done a lot of that development work in Phase III. And when they get to Phase IV, they have most of it done. But as people will remember for ST, they did their Phase III work a long time ago, kind of put it on the shelf and now they’re kind of — they’re taking this and they’re doing it for a brand-new process node. So, they do have some work to try to — I mean, I think they’re going to get very good results on their first run but their goal is to spread MST in a lot of places, so they can get good results across more and more components in their designs.
Richard Shannon: Okay, fair enough. And a follow-on question on the topic STMicro here. We’ve talked about this since the day that you announced this back in April of last year but you talked about — I guess I’ll refer to it as a halo effect of driving other companies to want to adopt MST. Maybe you could just characterize any of the discussions that you’re having. And I guess the key for me is do people need to see this in production and actually chips coming out the factory before they pull the trigger? Or do you think this can be an accelerator catalyst before then?
Scott Bibaud: It’s definitely already an accelerator and catalyst. In conversations with customers now, we’re discussing our business model and what we’d like to see from them, they say, yes, I recognize that ST is doing it and then we go on from there. But that’s great because in the past, they could say to us, nobody has ever done this before. So why do you think we should do it? So, it’s really a big game changer.
Richard Shannon: Okay, great to hear there. You talked about this last quarter in terms of — and now you have — I think you used the word license for MS TCAD with the large analog manufacturer. Is this an example of this halo effect? Or is it separate and even started before ST was publicized?
Scott Bibaud: I think it’s definitely an example of the halo effect. Of course, we were going to convince these guys before we announced ST and we actually did work with them. But seeing ST has caused them to take it to the next level. And I hope that the MSTcad is the first step in kind of moving along to doing an actual installation and license. Although we haven’t announced that yet but that’s what we hope is the trajectory.
Richard Shannon: Okay. Well, you kind of co-opted my follow-on question on this topic which is what are those next steps here, both internally that you could characterize plus any public events, so to speak, like the license as you just mentioned.
Scott Bibaud: Yes. I mean if you look at what I talked about for the productization effort for ST, they are starting all of their work with TCAD; An MS TCAD is an add-on to a standard TCAD that adds the MST stuff into it. So, for any of our customers, it would be — in the semiconductor industry, I would say the majority of manufacturers are a — they want to see actual physical results on a piece of silicon before they make a decision. But for some of the players in the industry, they want to run simulations first and then make their decision to move to silicon. And so, I think it’s really good that they’re doing both paths. We’re prepared to support them on both paths and it shows the seriousness. It’s not a small deal to dedicate a few engineers and a few very expensive Synopsys TCAD seats to working exclusively on integrating MST into your process technology. And so we hope that goes someplace is good.
Richard Shannon: Okay. And I’m assuming since you described this as a large analog manufacturer that this is something they would install and use internally as opposed to working with somebody else like perhaps a current engaged Atomera customer. Is that fair?
Scott Bibaud: Yes. I don’t think they would work with an engaged Atomera customer. I think every company in the globe today is even if they’re an IDM; they are probably fab light, so they use some outside foundry as well as inside capacity. So these guys could go in either direction but — yes.
Richard Shannon: Okay. Two quick questions and I’ll get a line here. I’m sure there are a couple of others to want to ask questions here but just related to GD1 here. Sounds like you made some progress in discussions with some business units there but it didn’t detect anything that was imminent and you’re just hopeful. I don’t want to put words in your mouth, so maybe you can just characterize the dynamics there. And do you think you’ve gotten past I know at this point in certain other learning cycles you’ve gone through them with them, I think, over the last two — I think it’s been more than two years, correct?
Scott Bibaud: Yes. I mean we’ve done a lot of stuff with them. We’ve shown them a lot of data. And by the way, we are continuing to do technical work for them to show them even more results of things that can be done with MST. We started in discussions with them on the business side in kind of in the fourth quarter. And no, we haven’t completed that yet. I don’t think that’s that unusual. And obviously, we can’t give details of exactly where we are on those discussions but these are the types of things that take some time. And over the Christmas holidays, we certainly didn’t slow down but every big company has got a lot of things happening at the end of the year with goal setting and reorgs and compensation discussions and everything. And so, I definitely feel like they were a bit distracted at the end of the year. And so in January, we started to really try to get it ramped back up. And hopefully, we’ll make better progress now.
