Hi, welcome to the Canaccord Genuity Growth Conference. I'm Kyle Mikson. I cover life science tools and diagnostics for Canaccord. Please welcome me to a fireside chat with PacBio. PacBio offers some of the leading products for long-read sequencing across a range of throughputs. With the company, we have Jim Gibson, CFO. Thanks, Jim, for joining us today. Appreciate it. Great. Good morning, Kyle. Thanks. Just to start, you reported your second quarter results last Wednesday, almost a week ago. Maybe just talk about some of the key factors at play with respect to the financial performance and, of course, the management transition that was announced as well. Sure. All right, Kyle. Again, thanks for inviting us to Canaccord to speak this week. First, we finished the second quarter with $39 million in revenue, up sequentially from the prior quarter. Solid consumable revenue, $20 million, so we were happy with that. Instruments at about $13 million and services at six. Services are down slightly. We finished a big population genetics study over in Asia year-over-year, so that slightly decreased our service revenue. We were happy with our growth in consumables, especially related to clinical. We had 67% growth in our clinical business, which is really an area we're leaning into. So we're excited to see that performance. Also really excited about some of the expansion in our fleets. We had two big deals with two existing customers to expand our fleet with Revio, and then we did a really large deal with a new population genomics initiative that we will be talking about a little more in Q3 that shipped 5 Revio. We are really excited about that deal as well. So continued strong demand for Revio. We believe primarily driven by the SPRQ-Nx launch, which this was also the first quarter of the commercial launch of SPRQ-Nx. We released that in May, so we got great reception from our customers on that as well. Then I think as important as the company continues to lean in heavily to the SPRQ-Nx transition, as well as kind of leaning heavily into clinical, we completed the transition to Mark Van Oene as our new CEO. He stepped into that role. We announced that last week. That is something that has been in the works for a few years. He joined PacBio about 5 years ago, and in that time, he ran R&D. He was in charge of successfully launching the Revio, successfully launching the Vega, as well as successfully launching our SPRQ-Nx chips, as well as our original SPRQ chemistry. He also ran operations, so got intimately involved in driving operations and came from a long history of commercial when he was at Illumina. We are really excited for Mark. He is really focused on kind of expanding our presence in the clinical markets and really sort of continuing to expand the growth we are seeing in EMEA and the rest of the world. Some of those big successes we are seeing in EMEA, we want to replicate in both the U.S. and Asia Pacific. Yeah. Okay. Well, that was a great recap. You also had a little bit of a reduction in force, I believe. Correct. Can we maybe talk about what divisions were impacted and if any additions will be added now that will maybe help optimize the areas that were impacted by this RIF recently? Mm-hmm. A couple things, yeah. One of the things that we've been impacted by this year, and we anticipate somewhat next year, is this dramatic increase in compute prices. One of the things we've committed to the street is to get ourselves cash flow positive around the launch of our next tool. In light of the continued impacts of the compute and memory shortage, as well as slightly slower uptake of SPRQ-Nx than we anticipated, we did a targeted reduction in force, and we went and we reduced marketing. We're now specialized. We're going to focus on the clinical market. We think we have a fantastic presence already in academia and government. The plant and animal are well-known in those areas. What we've decided to do is have a targeted marketing team focused on clinical, and they're going to be joining the commercial team. I think that's one of the things Mark brought to this: he wants to see that as one united team. In doing that, we reduced marketing fairly substantially. We also reduced our spans and layers. We went through the company and took out some of the folks that were focused more on management than execution, and I think one of the things that's important as we move towards continuing our success with SPRQ-Nx and the launch of our next tool, we really want people that are hands-on with the business and our customers, quite honestly. That was what the reduction in force focused on. Part of the reason we did it, as I said initially, 2027 is the area we're really going to start showing, we believe, some growth overall, especially with the launch of our next tool towards the latter half of 2027. One of the things this reduction in force allows us to do is really focus on reducing our comp and benefit expense. We anticipate $15 million to $20 million in decreased expenses related to that. As well as in 2027, we should be on the other side of most of the major spending we have for our new high-throughput box. We believe we're looking at between $30 million and $40 million of decreased spend in 2027, and that helps extend our cash runway. It really is about making sure the company is successful, focused on executing in the areas that we're winning in, and making sure everyone's focused on that. That was why it was the perfect time to do the restructuring with Mark's promotion, because that's what he's really focused on. Got it. Okay. He had input on which teams were impacted and things like that, right? Absolutely, yeah. Great. Okay. On the GPU usage and the memory usage and all that, what can you do to mitigate that headwind, I guess? Mm-hmm. So I think right now what we're doing to mitigate it is we actually went out in the markets and bought inventory ahead for the rest of the year. So we secured our supply for the rest of the year. Now, one of the areas we've not historically focused on is something that the engineers call the datapath, which is how much memory and GPU you actually optimize for. Historically, it's been more of a commodity, so we haven't spent a lot of R&D time finding the best way to maximize that. So what the R&D team now has prioritized, quite honestly, is how to optimize the GPU and less memory in order to keep the performance we want, but not overkill on the amount of memory and compute we supply. I think that is an area we are really focused on as we exit this year and enter next year on the R&D side. There is a big focus on that. We are also working with some of our partners on how they analyze and run the data once it is coming off the box to try to make that incredibly more efficient as well. We are hoping that in 2027 we will make some significant improvements to our demand for memory and compute that goes in our boxes. Are you keeping all this in mind for the next instrument, the ultra-high throughput box, as well? Maybe that will have a more efficient memory usage or something, storage. Absolutely. Conveniently, we had already started with that premise with that box; now we are trying to push it backwards into our existing boxes. Okay. That was part of the strategy on the next box, was how do you get more efficient with compute? How do you actually give the customers the choice of how much compute they apply? That is one of the things we are really looking deeply at, and one of the things, the design parameters we put in for the next box. We are pushing that back, and hopefully, like I said, we will see some results on that in early 2027. Got it. Okay. Going back to the quarter and the consumables growth that you were referencing. It was impacted quite a bit by the transition to the SPRQ-Nx. That's a lower-priced consumables offering. It was launching, taking you seven, the Revio, in May. As a result, customers weren't building inventory of the historical Revio consumables. It kind of had this impact on the consumables revenue per year, or the per box, which is the pull-through. That was $202,000 or something. per box in the quarter compared to $220,000- $230,000 is what I think you were kind of guiding to, maybe we were getting towards. Maybe, again, walk through. Maybe I didn't explain it well enough. Walk through it and then also how that will shape up going forward as the transition works its way through. Sure. One of the things, if people aren't familiar, is that we launched our new SPRQ-Nx chemistry in May officially. That provides three uses for the chip. It's also about 35% less on the ASP. One of the things we did anticipate, but we didn't know how much, is that two things would impact our launch. One would be if some of our largest service providers had to use up their inventory. On the positive side, about a third of our customers have already converted their software to the SPRQ-Nx software, enabling multi-use. A lot of these bigger customers did start buying SPRQ-Nx, but one of the things we weren't anticipating was they're fully running through and eating all of their inventory on-hand right now. We did see a slight lull in Q2 related to while they chewed their inventory. The reason we know that's happening is we can see usage, so we see them still continuing to use at a high rate, but they're not refilling, and that's in anticipation of SPRQ-Nx. That was something we did anticipate. Not as much as we thought when the first launch happened, and we expect that most of those folks should be through that by the second half of 2026, the tail end of 2026. The other thing is, obviously, this is 35% less than the standard ASP. We knew as customers started to onboard it and validate their new workflows, there would be a small lull while they brought more samples on. One of the things we are looking at is the utilization we need to get back to revenue parity, if you will, is you need each Revio to run about 10-15 more samples a month. That is kind of it to get back to where they were. With this 35% decrease in cost, we really expect price elasticity will drive those more samples, especially at our larger service providers. Part of the reason we ended up adjusting our revenue down slightly was anticipating it is going to take a little longer to people to burn through their inventory. Two, to validate the new SPRQ multi-use workflows, three uses, as well as start getting ramped up kind of in the tail end of 2026. Okay. Well, it has been a few months. Maybe you have seen some promising kind of developments on the multi-usage side. Any takeaways that give you confidence that the elasticity is there, that volume will pick up with these multi-use models? Yeah. We just launched at the tail end of May, so we do not have enough data yet to comment about utilization or sample volume. That is what we expect. That is what our customers tell us. We do not have enough order volume yet to actually conclude that, but hopefully, as we come into Q3, we will have some good indications of how that is going. Okay. This will be helpful for the larger scale projects, like you mentioned some in your call last week, I guess. You mentioned, I think, five placements maybe in the second quarter or in the third quarter, but as a result of this large-scale project. What is the pipeline for those projects like globally, I guess? How important will that be as a theme for 2027, for example? Well, I think one of the things, if we go back in history a little bit, at ASHG last year, we launched SPRQ-Nx and multi-use, and we started having more conversations than we ever had before with a lot of the population genetics studies and larger clinical customers, quite honestly, because the price point was much closer to where short-read was. We believed it would drive larger studies and more volume to us, so we have had more conversations than we have ever had. We talked in Q1 how we won the GeneDx deal, which was the company's largest project to date, 100,000 samples. That was enabled by the SPRQ-Nx chemistry. Then we just announced in Q2 another big population genetics study that we have not commented on the volume yet, but another significant size population genetics study enabled by SPRQ-Nx. They are the group that we shipped a number of Revios to. You expect the ramp on that take four to six months, so in 2027, we will start seeing that contribute meaningfully to revenue, and that is kind of the typical timeframe before you start seeing strong contributions. Both GeneDx as well as this new population genetics study we penned should be contributing meaningful revenue in 2027. We are not really anticipating a lot of that in 2026 for those two studies. Overall, like I said, there are more deals than we have ever had. We are quoting deals left and right with SPRQ-Nx as well as Revio, and hopefully we will be talking about more of those deals over the next couple of quarters. How large are these deals from an annual revenue perspective relative to your guidance, for example? Well, let's see. We're not giving any guidance into 2027 yet, but each of these would be a material deal, a material amount. Yes. to our revenue base, each single deal. These are substantial. Historically, when PacBio has announced these size deals, it's one 5,000- 10,000 sample deals for a year. These are much larger than this. We're getting 50%, 60%, 100% of these deal sizes now, as opposed to before when we got 5% of a deal size where people just wanted to try long-read. Now they just want a long read. Okay. That's one of the big shifts we're seeing as well. Yeah. Are these clinical-focused deals primarily, or kind of research? GeneDx's research, with ultimately the goal to drive clinical discovery. This other deal is PopGen with the goal of becoming clinical. Okay. So. I think GeneDx was a competitive process, I believe. Mm-hmm. Yep. What's happened, like going head-to-head versus either other long-read or kind of legacy short-read or legacy molecular sensing technologies, I guess? Mm-hmm. It was a competitive deal. It was interesting because it was a deal primarily driven from the AI angle. The way they considered the deal was slightly different than you might see with some of the more standard clinical or some of the researchers we've talked with historically. What they wanted was depth of data and quality of coverage and repeatability. It was a competitive deal. They did look at the different technologies. That's the reason we won. They liked our data depth. They liked the amount of coverage we had. They loved the reproducibility of it, and that's why we won the GeneDx deal. That is playing in many of the deal discussions we're having now. Many of these groups are coming to us because of those qualities of our data, especially as they're looking at building brand-new data sets. What they don't want to do is build incomplete data sets, then have to go back later and enrich them. A lot of these people's goal now is to start with the fully enriched data set. So depth, multi-omic level, as well as completeness of data. That's one of the shifts we're seeing in the discussions, is people are thinking more about, with the tools that are coming out today, do I want the best data I can have when I run these studies, or do I want to potentially have to go back two or three years later if I'm using short-read and enrich those data sets? I think a lot of them are now saying, because of the price point of SPRQ-Nx, let's start with what we consider the best data now. That's one of the things we're seeing that's a dramatic difference in some of the conversations of two, three years ago. Yeah. Will you talk more about that as being like a tailwind, a growth driver kind of going forward? Because it sounds like others in the tools space are talking about these biology-based foundation models that are powered by the data from some of these vendors. So. Yeah. One of the primary goals of Mark taking over and leaning in as CEO is to really push hard on the differentiation of our data quality and the depth of our data, especially in supporting research like this. Now, one of the benefits we have, because of, honestly, where we are located and the people we associate with, are some of the biggest AI labs are our next-door neighbors. So there is lots of conversations about what is the best type of data, whether it is proprietary in-house data or publicly sourced data, to run these models on. And we believe, as some of the publications are also saying, that it is our data type and data quality. So it is an area we are really starting to lean into. It is something that I know everyone is talking about, and I think, quite honestly, the proof is going to be in what the discoveries are made with the data set. One of the other things we are excited about is GeneDx just released their EDEN model on Anthropic in June. Now our data is just starting to be popular with that. They are really happy with our data sets. So we believe we will be contributing meaningfully over time to some of those discoveries. And they will be one of the first models that show how you can use our data to unlock those pieces of information. All right. One last one on this topic. Any additional ways to reduce friction to integrate further into those models like EDEN and so forth? Just because that seems like the future. Is there a way to kind of align better, maybe? Well, I think ultimately, when we talk about our next release, our high-throughput box, it really is to unlock a couple of things. It is to unlock even better price parity with short-read, much higher volumes, so you can actually build these training data sets, and more importantly, actually potentially diagnose on the other side of this. And three is designing the tools and the variant callers so they can work almost seamlessly with these models. That is really one of the company's focuses, and that is an area Mark's really excited about, and one of the reasons why the time was right for him to step into the CEO role. It is an area he is extremely focused on. It is an area he has pointed himself at and pointing the company at as we move forward. It is an area I have a lot of historical bandwidth with, both working for Apple and GoDaddy. I led some of their largest data teams, so it is one of the areas I am most excited about with our company, as well as kind of being in that foundational data set that for a lot of us in biotech is going to be incredibly important over the next decade. Okay. Yeah, definitely located in the right area. Yeah. Your next-door neighbors are working a lot of these companies, so you have some interesting conversations over beers on the barbecues. So, yeah. Yeah. Going back to the clinical consumables progress that you've had, so grew year-over-year 67%. The consumables, that's been growing 100%. It's been obviously very robust growth. The base is not enormous, right? So maybe just talk about the mix today that clinical represents of all of consumables. I believe it was like 15-ish percent most recently. Yeah. Mid-teens. Yep. Okay, so still mid-teens. That is coming from Europe, non-U.S., right? I think that the key bottleneck maybe in the U.S. is just the clinical labs there want higher throughput. That is why you are launching this box in the relative near term Yeah to address that. What does the future of the PacBio clinical business look like, I guess, kind of relatively soon? Sure. I'll start with EMEA and talk about why the success is happening there and how we think we can replicate that. In EMEA, they've been very forward-leaning in using the Revio and now SPRQ-Nx, especially with rare disease. I think one of the big publications that came out in The New England Journal of Medicine highlighted that, how we've always believed that you could use whole-genome sequencing to consolidate tests, and ultimately, hopefully, have a better clinical outcome. The New England Journal of Medicine just published a research that confirmed that. That was led out of EMEA, led out of Europe. The reason we're doing well there is, quite honestly, a lot of the single-payer systems, as well as some of the smaller hospitals, the Revio has the perfect throughput for some of these localized hospitals. Their reimbursement environment is favorable towards whole-genome sequencing, so we're seeing a lot of success there, which, once again, we had over 50% growth in EMEA year-over-year. It really is the group we want to copy. How do we bring that into the U.S.? I think there's a couple things we're working in the U.S., is one is, it's much larger centralized testing in the U.S., obviously. They want higher throughput, number one. Number two is they want a favorable reimbursement environment. Obviously there's a reimbursement environment that already exists in the U.S., and I know a number of our customers are working to get whole-genome sequencing reimbursed at the same rates. The other thing is, as they validate and bring these tests online. We have a number of what we call commercial customers in the U.S. that are moving to whole-genome sequencing, and they are some of these big testing labs that are probably presenting at the conference this week. As they bring these LDTs out of the R&D phase and start ramping them internally, they do have a long process, and especially with something like SPRQ-Nx, they're going to take a little while to ramp up. But we are seeing growth and interest in those areas. Two of the expansions we had in our Revio fleet were those types of commercial providers in Q2. They really are leaning into whole-genome sequencing. Okay. Since we haven't touched on Vega at all, does Vega help this dynamic at all? Maybe like transitioning certain labs, whether research or clinical to Revio over time, or any kind of impasse in the model, really? Yeah, so Vega was created to do one thing, which was really expose people to long-read sequencing at a price point that they could afford. It was the perfect box last year when academics and government in the U.S. contracted. We had a lot of sales there. We are still selling a lot of Vegas in academic and government and around the world. We have been impacted by the A&G continuing to be slow, so Vega hasn't been performing as we had hoped this year. It is still selling well. We had strong sales this quarter. Our ASPs normalized. We did a promotion last quarter. They are back to normal, and we are still selling as many boxes. 80% of those customers buying Vega are new to long read, new to PacBio, so it was built to introduce new customers. I think one of the things we are excited about with Vega is we do believe it has a space for whether it is targeted genetic testing or other things. We think it does absolutely have a space for that, especially in other areas of the world. We are now launching SPRQ-Nx chemistry for Vega. That is happening in August. That is not multi-use, but what it does bring is increased output, lower DNA input. So we believe it is going to increase the use cases for Vega dramatically. We have not forecasted an uptick for that, but we do believe that will be one of the tailwinds that drives Vega into the future for us as we continue to penetrate. The other thing we are excited about is a lot of public health labs are starting to buy Vegas now too because of what it can see. So that is a new area we have just started focusing in and still see resilience there, even with some of the cutbacks. Okay. All right, great. Then in terms of the guidance, so you cut that by maybe like $10 billion or so. Midpoint now, it is $160. It used to be $170. In the quarter, I think it was relatively in line. You didn't exactly miss by a ton. But going forward, we have this transition going on. The ONG environment is uncertain, like you said. What are some factors that made you guys revise the guidance and make you feel comfortable now with this new range? Mm-hmm. Sure. I think it was two things fundamentally. Number one was the SPRQ-Nx transition. I think we're trying to be a little more conservative in our expectations about how long it's going to take some of our biggest customers to chew through their inventory. Number two, as they validate these new workflows and actually confirm that the third use actually meets specs. We're seeing, and have factored that in. That's really what drove kind of the majority as we decreased the revenue. Number two was Vega, the realization that historically we would see a pretty big uptick in demand, especially in ONG, in the second half of the year. We're not necessarily planning or factoring that into our thinking anymore. Now, it could happen, but it's not in our guide anymore. The combination of the SPRQ-Nx conversion being slightly slower than we'd anticipated, as well as some of the Vega not having the demand uptick in the second half of the year that we were hoping we would see when we came into the year. That's why we brought the guidance down. Now, again, we still think this is eminently doable for us. We still believe we're having strong demand for Revio. We're still seeing strong demand of Vega placements, and we still believe there'll be a dramatic increase in SPRQ-Nx purchases in the second half of the year. We just don't know how much right now. Okay. Obviously that affects maybe cash burn. You have $10 million less of revenue at least, and then you have the severance stuff going on as well. However, in a year from now, let's say, or even like end of next year, you'll have this potentially $40 million run rate expense savings due to the RIF recently as well as the end of the R&D process for the ultra-high throughput, which will, I guess, be announced like late next year. Then you'll launch that and you'll have a big 2028 most likely as a result of the ramp of that or hopefully it's an impact from that. I can see the top line being much better in 2028 as a result of that new product, and then you have these expenses finally being taken out. Maybe just talk about the progression in the next few years, the medium term in the cash burn. Your runway is at least a few years from now, so maybe talk about how it gets to breakeven over time. Yeah. I think there are three things we have to do to get breakeven, and it is kind of business school 101. A, we have got to successfully launch our next box, and that has got to be a portfolio launch, not a replacement launch, and we believe that is what it is going to be. Two, we have to ultimately find a way around the compute and DRAM shortage. Now, I think there is capacity coming online. We are not anticipating that, so we are going to be designing our way out of that to start contributing margin again. Three is SPRQ-Nx has to be successful. We have to convert the majority of our customers to that because that drops a lot of gross margin, or that drops a lot of profit to our bottom line as we convert people to multi-use. So if those three things happen, we feel very comfortable with the fact that we will be cash flow positive in 2028. Now we have a lot of execution we have got to do to get there, but we believe with the publications, the customers, and the stuff we have not explored yet on the datapath, that we are going to make it. So that is why we, A, did the restructuring when we did it to give us a little more runway. B, give us a little more time around to deal with the compute issues as the rest of the world deals with that as well. And C, chart a course to cash flow positive in 2028 for the company. All right. Adding a new instrument in 2027, 2028 will not help gross margin. But how do you think about medium to longer term gross margins that will get you to that breakeven point? Yeah, we have to be knocking on the door of 50%, which we think is very realistic once we get through compute, as well as the heavy product mix to consumables. We think that is a very realistic gross margin for us to achieve. Beyond that, there's obviously like a higher ceiling, I guess, then. Yes. Oh, yeah. We are going to see that from there. What we're planning is knocking on the door of 50% cash flow positive. That's what we need to achieve. All right. Awesome. I think with that, we can leave it there. Thanks, Jim. Appreciate the time. Yeah. Appreciate your time. Awesome. Thanks. Thanks, everyone.