CyberOptics Corporation
Q2 2017 Earnings Call Transcript

Published: 30 Jul 2017

[Operator:]

Good day, and welcome to the CyberOptics' Second Quarter 2017 Earnings Conference Call. Today's conference is being recorded. At this time, I would like to turn the conference over to Dr. Subodh Kulkarni, President and Chief Executive Officer. Please go ahead, sir.

[Subodh Kulkarni:]

Thank you. Good afternoon, and thanks for taking the time to participate in CyberOptics' Second Quarter Earnings Conference Call. Joining me is Jeff Bertelsen, our CFO and Chief Operating Officer, who will review our operating results in some detail following my overview of our recent performance. We will then be available to answer your questions at the conclusion of our remarks. In keeping with Regulation FD, we have made forward-looking statements regarding our outlook in this afternoon's earnings release. These forward-looking statements reflect our outlook for future results, which is subject to a number of risks that are discussed in our Form 10-K for the year ended December 31, 2016, and other filings with the Securities and Exchange Commission. We urge you to review these discussions of risk factors. Turning now to CyberOptics' second quarter performance. Sales for this period totaled $16.4 million, down from $18.6 million in the year-earlier quarter; but up 38% from $11.9 million in Q1 2017. We also reported earnings of $1.1 million or $0.15 per share in the second quarter of 2017 compared to $2 million or $0.29 per share in the year-earlier period. Current quarter earnings marked a significant improvement from the net loss of $214,000 or $0.03 per share in the first quarter of 2017. We are encouraged by the substantial progress of our sales and earnings on the quarterly sequential basis. A significant portion of our sequential growth was generated by sales of 3D MRS-enabled systems and sensors, which rose to $4.7 million, up 107% from the first quarter level. This type of growth reinforces our conviction that CyberOptics is making strong headway in the rapidly growing 3D inspection marketplace. And consistent with this belief, we have developed a robust pipeline of opportunities for large projects involving multiple SQ3000 systems. But it must be emphasized that converting this pipeline into revenue is dependent upon the timing of orders, implementations and customer acceptances. I will now turn to a brief overview of our various product lines. 3D MRS-enabled SQ3000 inspection systems accounted for a significant portion of CyberOptics' second quarter system sales growth, both year-over-year and sequentially. The SQ3000 is also helping us drive higher sales of solder paste inspection systems. The steadily growing acceptance of the SQ3000 is being generated by the competitive advantages of our disruptive 3D MRS technology platform, which is enabling us to capitalize upon strong worldwide demand for high-precision inspection. In this related inspection systems development, we have enabled the SQ3000 now with a measurement capability, incorporating some of the functionality of our CyberGage product and traditional coordinate measuring machines. This version of the SQ3000, which also will retain its AOI capabilities, is expected to open up additional sales opportunities, primarily with semiconductor manufacturers. We also plan to launch a 3D MRS-enabled solder paste inspection system in this year second half. This higher performing system should strengthen CyberOptics' position in the solder paste inspection market. In addition, we are actively pursuing new customers for our MX600 memory module inspection system. As you may recall, we shipped $5.7 million of MX systems in 2016. Since this product entails a lengthy sales cycle, any sales would not be realized until 2018. Turning now to sensors. Sales of 3D-enabled sensors reached a record high in this year's second quarter. The third quarter and full year outlook for these sensors is also promising. Sales generated by CyberOptics' long-term supply agreement with KLA-Tencor are expected to continue growing as our 3D sensors are now standard on KLA-Tencor's back-end semiconductor inspection systems utilizing 3D optical inspection. We also made further progress during the second quarter with our initiative to apply MRS-enabled sensor technology to mid-end and front-end semiconductor inspection. Our MRS technology is now able to inspect cracks and other defects down to 30 microns in wafer dies. We are currently demonstrating this mid-end inspection capability to semiconductor manufacturers and believe significant orders all possible in 2018. Second quarter sales of semiconductor products, primarily, the WaferSense/ReticleSense product lines were particularly strong in comparison to this year's first quarter. In response to demand from semiconductor manufacturers, we are developing additional product offerings for new applications in semiconductor fabs and flat panel display manufacturing. As a result, strong future sales growth is anticipated for the WaferSense/ReticleSense portfolio. Turning now to our CyberGage360 3D scanning system. The selling cycle for this new product remains lengthy as many potential customers are evaluating the impact of this disruptive technology on their product workflows. Sales activity is picking up as an order for two systems was received in July, and we anticipate several more as the quarter progresses. We can also now report that Proto Labs, the world's fastest manufacturer of custom prototypes and on-demand production parts, is a very happy CyberGage customer. Later this year, we will launch a second CyberGage system that offers greater resolution within a smaller working envelope. The current version of CyberGage will continue to be marketed for larger parts. Based upon positive customer feedback, we remain confident that CyberGage should become an important contributor to CyberOptics' long-term sales growth. CyberOptics ended the second quarter with a backlog of $7.3 million, the majority of which is scheduled to ship in the third quarter. Including other anticipated orders, we are forecasting sales of $13 million to $15 million for the third quarter of 2017. Our guidance for this period does not include the pending orders of approximately $3 million of SQ3000 systems, which we expect to receive in the third quarter. Although it's possible that revenue from these pending orders could be recognized in the third quarter, it is more likely these revenues would be recognized in the fourth quarter, reflecting shipment and implementation schedules and the time required to obtain customer acceptances. Our ability to improve year-over-year sales growth is dependent upon timely receipt of customer orders and acceptances for SQ3000 systems. Despite this near-term timing issue, we remain very optimistic about CyberOptics' future given the growing acceptance of our suite of MRS-enabled and WaferSense/ReticleSense products, the outlook for midyear mid-end semiconductor inspection and for gradually growing CyberGage sales. Thank you. Now Jeff Bertelsen will review our second quarter operating results in greater detail.

[Jeff Bertelsen:]

Thanks, Subodh. Second quarter sales were up 38% sequentially and earnings rebounded to $0.15 per share from the net loss in the first quarter. This growth was generated mainly by sales of higher margin 3D MRS enabled inspection systems and sensors, which rose by more than 100% from this year's first quarter. We believe the competitive advantage of our MRS technology are enabling CyberOptics to make strong headway in a rapidly growing 3D inspection marketplace. This progress bodes well for CyberOptics' future. Now I will briefly review the sales performance of our product lines. Second quarter sales of inspection systems increased 25% from this year's first quarter to $5.8 million; and by 12% from the year earlier period, which included about $0.5 million of revenue for MX600 memory module inspection systems. These increases were driven mainly by sales of 3D MRS enabled SQ3000 AOI systems, which rose 63% from the first quarter level and by 49% year-over-year. Second quarter SQ3000 sales also helped drive higher levels of solder paste inspection systems. As stated in this afternoon's release, we expect to receive orders totaling approximately $3 million, primarily for SQ3000 3D AOI systems. Although it is possible that revenue from these pending orders could be recognized in the third quarter, it is more likely these revenues will be recognized in the fourth quarter reflecting shipment and implementation schedules and the time required to obtain customer acceptances. For this reason, we have excluded these pending orders from our third quarter guidance. We believe third quarter system sales will be consistent with the level posted in the second quarter, but down from last year's third quarter, which included $2.8 million of MX600 revenues. Third quarter SQ3000 sales are forecasted to increase substantially year-over-year even without the delay in SQ orders that are expected to revenue in Q4. Third quarter SQ sales are also forecasted to be up on a sequential basis. Second quarter sales of OEM sensors increased 44% on a quarterly sequential basis to $5.6 million but were down 12% year-over-year. Within this category, sales of MRS enabled sensors increased by over 230% on both the quarterly, sequential and year-over-year basis. Our 3D MRS enabled sensors are now standard on our KLA-Tencor back-end semiconductor inspection systems utilizing 3D optical inspections. Our sensors are also being incorporated into the AOI systems of Nordson YESTECH, another OEM customer. We anticipate that demand for 3D MRS-enabled OEM sensors will remain strong in the third quarter. And while total OEM sensor sales are forecasted to register further growth in the third quarter on a year-over-year basis, we believe they will be down somewhat on a sequential basis. Second quarter sales of semiconductor products, primarily the WaferSense/ReticleSense product lines, rose 36% on a sequential quarterly basis to $3.2 million and were virtually unchanged in comparison to last year's second quarter. Demand for our products is being driven by semiconductor fabs in capital equipment company. In response to demand from semiconductor manufacturers, we are developing additional product offerings for new applications in semiconductor fabs and also flat panel display manufacturing. As a result, we anticipate strong future sales growth for our WaferSense and ReticleSense portfolios. Finally, second quarter sales of general purpose 3D scanning solutions and services rose 76% sequentially to $1.8 million, reflecting improved sales of LDI's traditional products but were down 54% year-over-year due to a large sale of x-ray systems that occurred in last year's second quarter. Consistent with our previous outlook, sales of the recently introduced CyberGage360 3D scanning system were nominal in the second quarter. However, we received an order for two systems in July, and we anticipate several more orders as the quarter progresses. Moving down the P&L. Our second quarter gross margin of 47% was up from 45% in the first quarter and 44% in last year's second quarter. These improvements are primarily due to the increased percentage of higher margin 3D MRS systems, sensors and WaferSense products in our sales mix. We believe our gross margin percentage in the third quarter of 2017 will be consistent with the second quarter level. Total operating expenses in the second quarter were up modestly from the first quarter level but down slightly year-over-year. Stock compensation expense from the second quarter totaled $212,000 and depreciation and amortization totaled $536,000. With respect to the year-over-year decrease in second quarter operating expenses, higher expenses related to sales and marketing initiatives were more than offset by reduced stock compensation, expense and bonus accruals. We believe third quarter operating expenses will approximate the second quarter level. Our income tax provision for the second quarter of 2017 reflects an ordinary effective tax rate and the absence of any significant favorable excess tax benefits. Our income tax provision in the second quarter of 2016 was insignificant because at that time we continued to have a valuation allowance for deferred income taxes. Consistent with last year, actual cash payments for income taxes will be insignificant in 2017 because of our available net operating loss in R&D tax credit carryforwards. Finally, cash and marketable securities totaled $22.7 million at the end of the second quarter compared to $23.2 million at the end of this year's first quarter as cash was used to purchase inventory for anticipated sales in this year's second half. We anticipate that inventories will be reduced as the year progresses. Thank you. I will now turn the call over to the conference call operator, who will poll you for any questions.

[Operator:]

Thank you. [Operator Instructions] And we'll take our first question from Ash Birla with Dougherty & Company.

[Ash Birla:]

Subodh and Jeff, I wanted to talk to you guys about you guided $16 million to $19 million. You did $16.4 million. So clearly, at least on the high end of your guidance, you were $2.6 million short, but obviously, forecasting does change. And you had guided full year, I mean, kind of like a growth rate. So if you do the math, you should have done anywhere between $17 million to $20 million in the third quarter to meet, kind of like to progress towards your full year growth rate. So $2.5 million here, roughly $2.5 million in the third quarter. So what kind of numbers do you have to pull in for the fourth quarter to actually do the full year growth rate? And is that even possible now?

[Jeff Bertelsen:]

Yes. Good question, Ash. With the pending $3 million order that we expect to receive in Q3, and as of today any ways, we expect that to book in Q4, not Q3. We're setting ourselves up for a good fourth quarter. However, given our outlook for Q3, just doing the math, we would have to do a very large Q4 in order to post growth in 2017. But right now, I think it would be difficult for us to say definitively that we are going to pull growth in 2017, just given the magnitude of the Q4 that would be required. It does look like we're positioning for a very good Q4 but I don't know that we can see our way to having a big enough Q4 to have growth for the full year.

[Ash Birla:]

Okay. But why was it -- can you guys share, what changed between your first quarter earnings and the second quarter in 3 months? The delta -- it looks like it's not that big of a delta, it's $3 million, $4 million delta, what changed? Was it SQ3000? Was it CyberGage360? Was it that big -- large OEM? I know you had taken out large OEM from your previous expectations for full year. So is there something that happened that changed from first and second quarter [indiscernible]?

[Jeff Bertelsen:]

I think what we've seen -- and so if we go back to April, when we announced Q4 and...

[Subodh Kulkarni:]

Q1.

[Jeff Bertelsen:]

Q1 rather, sorry. And we talked about our pipeline of SQ3000 opportunities and having a robust pipeline of SQ3 opportunities, which is that -- we still have, it's definitely do -- still there. I think what we've seen, though, is some of those orders coming to fruition. It's taken longer than maybe what we had originally anticipated, even for example, looking at the order that we're talking about today for $3 million, I think it's taken a little longer for that order to come to fruition here. And so I think that's probably -- that the difference is we're just seeing some push out in the time lines. We haven't lost anything. But I would say, in fairness, it's probably just taken us a little bit longer than what we had anticipated to have some of these orders get booked and recorded as revenue.

[Ash Birla:]

Okay. The $4.7 million was really nice to see, the 3D MRS-driven revenues. That's a pretty high -- sounds like almost like a 90%, 100% growth rate. Can you give us a sense of what you would do for Q3 '17 in terms of 3D MRS? Is it going to be sequentially up?

[Jeff Bertelsen:]

Yes. So we would expect, definitely, would expect in Q3 '17, 3D MRS revenues to be up. That would be SQ3000 and CyberGage would be the big contributors there. I mean, we already have pocketed a couple of orders for CyberGage. We expect to get some more CyberGage orders and also are expecting good quarters on the SQ3000 and also on the sensor side. So we would expect them to be sequentially higher.

[Ash Birla:]

The 3D MRS driven?

[Jeff Bertelsen:]

Right. Correct.

[Subodh Kulkarni:]

I'll expand on it further, Ash. As you know this quarter, the numbers tend to move around a lot because of the timing of orders and acceptances. So I try to look at things on a more annualized basis. And when I look at on an annualized basis, we believe that 3D MRS-related revenues are going to see significant growth from 2016. We are expecting obviously a strong second half in 3D MRS, driven by SQ product. But on an annualized basis, we are still looking at maybe not quite close to doubling from last year, but pretty close to that kind of a range. So we are going to see very nice healthy growth in 3D MRS this year.

[Ash Birla:]

Right. I mean yes, if I just plot what you're doing, if you do 5 million or 4.5 million, 4.7 million or $5 million, that's a 190% growth rate in 3D MRS year-over-year. But that sounds like...

[Subodh Kulkarni:]

Again, don't take just 1 quarter numbers. Try to do it on annualized. I think we will see very high double-digit growth rate, I mean, approaching 100% in that range, but not quite -- maybe not quite triple digit. [indiscernible] substantial growth. Yes.

[Ash Birla:]

Right. So my question is then if 3D MRS tech-related is doing well. WaferSense clearly is doing well, growing at double digit 18%, 19%, 20%. So then let's look at the non-MRS. What happened there? It sound like your Q3, you will drop to almost anywhere between $5.5 million down from 8.5 million in Q2 to 5.5 million to 7.5 million, 8 million, right?

[Jeff Bertelsen:]

You're comparing Q2 to Q3?

[Ash Birla:]

In terms of legacy. I mean, not legacy in the 2D business. Even if you compare it to -- you cannot compare it to year-over-year because in Q3 '16, you did 11 million.

[Jeff Bertelsen:]

So I mean, if we're looking at Q2 '17 to Q3 '17, and where we're going to be down, is that your question, I think?

[Ash Birla:]

Right.

[Jeff Bertelsen:]

So in Q2, we had a good Q2 '17 in the WaferSense area, about $3.2 million. We did get 1 large order there from a new fab. So that will be down probably in the area of $500,000, $600,000. And then in the legacy kind of 2D OEM sensor category, we expect to see some reductions there too. We're going to have good 3D sales but some reduction in the 2D LaserAlign area. So those would be the two big areas where we'd be down and then also a little bit of decline in the solder paste and 2D AOI products.

[Ash Birla:]

Okay. So that means your gross margin is going to be higher in Q3 versus 2Q, correct?

[Subodh Kulkarni:]

Yes.

[Jeff Bertelsen:]

I had commented. We thought they'd be about the same. We're expecting them to be about the same in Q3 and Q2.

[Subodh Kulkarni:]

About 47%.

[Ash Birla:]

Just one last one. I'll get back in the queue. Proto Lab, clearly, put a press release out, they also mentioned in them their earnings call. And you talked about this metrology lab and it's kind of like a unique thing for them. And clearly, they're buying CyberGage360 for that. And then they will be launching it in Europe and in U.K., I think. Or they'd probably launched there or about to launch it this year and get there some production flowing in Q1 2018. And then they have plans for Japan. So my question is that how many units you feel that a company like their size would need? And how many labs can they do? And that is just one customer and you talked about a second customer. So how is that traction going with the hundreds, I believe, that you had? So that's kind of like a two-part question.

[Subodh Kulkarni:]

Okay. And both are good questions, Ash. And I'll try to answer them without quite quantifying or giving out details that are pertinent to their company. So as you know, they have 5 or 6 facilities worldwide and they plan to expand. What we get excited about is that they have incorporated CyberGage as an integral component of their manufacturing workflow. So now they are adding value to their customers by giving them in-process scans and final scans, which only CyberGage can do because of its speed and other abilities. Conventional CMM wasn't able to keep up with their speed. So this is exactly the value proposition of CyberGage. So they definitely plan to buy multiple units per facility of CyberGage. They are starting with their metrology lab but their plan is to have, certainly on the metrology lab, but also on the production floor. So we are talking multiple units or facility multiplied by their facilities and whatever expansion they do. So we are pretty excited about how it's all working out and how it will play out over the next few years. And that is a typical customer and that's what exactly what we are doing with customer number 2 and customer number 3 right now, maybe start the conversation, it takes several months for them to understand and appreciate the value proposition, incorporate it and then they start -- once they realize that this is not exactly a conventional CMM that should sit in the metrology lab, they see that this is actually a production floor environment equipment, where it can sit next to manufacturing, and it's a -- robust enough to handle all the manufacturing demands. They see how they can add value to their customers in that. It's an exciting product. We really feel very optimistic about where this whole thing will go long term. But as we have stated and as you have seen, it does take time to get one into the customer's facility, make them understand and appreciate the value proposition and then let them put their plan together.

[Operator:]

And we'll now take our next question from Jaeson Schmidt with Lake Street Capital Markets.

[Jaeson Schmidt:]

I just kind of want to follow up on some of those previous questions. I know on the last earnings call, you talked about having 20 large projects in the pipeline equating to about $40 million, which you expected to recognize mainly this year. Just curious if any of your -- if anything's changed from that standpoint. And if it has, is it mainly due to just timing of orders or share ships? Just additional color there would be helpful.

[Jeff Bertelsen:]

Yes. So we did talk about a $40 million pipeline, about 20 customers. Right now, we would quantify the pipeline about $31 million in 20 customers. Probably, the only thing that's really changed right now, we did have 1 customer where we won the eval, but they essentially decide to defer buying any 3D AOI. Haven't had any competitive losses per se. I think as we talked about, converting these opportunities to revenue has taken longer than what we would thought. But I think we'll -- with this $3 million order which is part of that pipeline, hopefully, we're going to start seeing it kick in here soon.

[Jaeson Schmidt:]

And then I know you guided for OpEx to remain fairly flat in Q3. How should we think about that in Q4? And then even looking at 2018, given these opportunities, will you continue to need to invest heavily?

[Jeff Bertelsen:]

I think relative to Q4, right now, I wouldn't see any significant overall change in OpEx. I mean, there would be some puts and takes. Just naturally in the fourth quarter, you see some expenses start to drop off, but we have had higher level of sales and marketing spend. So on balance, relative to Q4, I don't expect any significant type changes. I guess going forward, there -- as we get out further, there may be some handful of selective ads in mainly in the sales and marketing area again, but I don't think there's going to be any huge investments that will significantly change the equation.

[Jaeson Schmidt:]

And then the last one for me, just shifting to CyberGage. I know it seems like you have a ton of customer leads and some of these being really solid engagements. What has been the major pushback from customers as far as moving forward with actual POs?

[Subodh Kulkarni:]

The number one pushback usually is new technology compared to what they have been used to for 30, 40 years. And that always takes business -- the natural skepticism and reluctance to accept the new technology in a quality control area. So that continues to be the number one pushback. But now that we have had multiple engagements, some other limitations are coming out. One is size. Some customers want ability to go larger than what CyberGage can currently handle. The third pushback has come in the area of resolution. Right now CyberGage has, what we call, roughly about a 50-micron resolution, not accuracy, accuracy is in the 7-micron range, but the resolution of looking at fine features is about 50 micron. And that's where in my comments, I said that we are investing in a new CyberGage, which we'll do smaller parts but at a much higher resolution. Basically, what we are doing is we are taking out SQ sensor which has -- which is a design for PCB inspection as you know. And that is a higher resolution sensor, and we are putting it in the CyberGage small factor right now. And that will enable us to measure smaller parts than the current CyberGage but at a much higher resolution. And that will open up the market for some other sophisticated manufacturers that care about resolution. So at a high level there's been skepticism over new technology and just willingness to accept the new technology then size and resolution.

[Operator:]

And we'll now take our next question from Greg Palm with Craig-Hallum Capital Group.

[Greg Palm:]

Wanted to first follow up on a couple of questions on the guidance. I guess, Jeff, first, you mentioned some softness in maybe the legacy 2D. Just curious if you think the growth and adoption of 3D AOI is starting to cannibalize some of that? Or is there something else going on, either customer specific or competition, et cetera?

[Jeff Bertelsen:]

Yes. I don't think there is any cannibalization. I mean, that historically, if you go and you look at our 2D OEM sensor customers and some of the other duty AOI products and so forth, they've had a pretty cyclical history. And so I just think we're just seeing some of the natural peaks and valleys that we see from quarter-to-quarter with those products. We are not releasing any cannibal -- cannibalization. We're not losing any customers there. It's just kind of the natural ebb and flow of those revenues.

[Greg Palm:]

Is that specifically related to Q3? I mean, do you see that point out in Q4 and beyond? Or is that more just a sort of Q3 timing issue?

[Jeff Bertelsen:]

I mean, I think that that's just a Q3 timing issue if we go back and look. I mean, these revenues tend to kind of roller coaster up and down a little bit. I mean, we had a pretty good jump there from Q1 to Q2. And now it's going to slow up a little bit in Q3, but it'll pick back up again in the future. I'm not really worried about that.

[Greg Palm:]

Okay. In the 3D AOI market, do you think the overall industry growth rate is still where it was kind of previously? And just wondering if you can update us on the competitive landscape there. Any other color there would be helpful.

[Subodh Kulkarni:]

Sure. I'll take that, Greg. So overall, we still believe the market is growing very rapidly. And consistent with what we have said before which was, last year the global market was about $100 million and this year we think the global market is in the $150 million neighborhood. That continues and we clearly, as we discussed earlier, growing very fast. We are growing faster than the market definitely right now, which means we are gaining share because of the competitive MRS differentiation. So all that is working out great and one of the comments I made which had important impact long term on our ability to grow faster than the market in 3D AOI, this measurement capability that we have incorporated in our SQ system. So essentially, we have incorporated some of the CyberGage functions since we have learned a lot in how to make CyberGage happen in SQ now. So essentially, SQ is not just a AOI system, but it has some aspects of CMM system that we can do measurement. And that certainly has opened interesting segment that I don't think gets normally captured in the 3D AOI market. So I mentioned in my comments, particularly the -- that semiconductor industry seems to be very interested in an AOI machine that can also do measurement because they have some unique applications where they need to measure things in addition to inspecting them. So we feel very good about the overall 3D AOI market globally, how it is doing and certainly how we are position in it to grow faster than the market right now.

[Greg Palm:]

Got it. And in terms of the 3 million in SQ systems that you've talked about, can some of that we recognized in Q3? Or is it an all or none type of recognition?

[Jeff Bertelsen:]

Well you know it's certainly could be possible that some of it could be recognized in Q3, and it isn't necessarily an all or nothing proposition. So it's possible, based on what we know about that particular customer today, we didn't feel comfortable putting it in our guidance. Part of the reason why we did want to highlight it though in our release was if something does change and all of a sudden we were to get significant revenues from that particular order, we didn't want to -- on the flip side, catch people by surprise either. But as of today, it's not in our outlook because we expect it to be in Q4, but it is possible, the implementation schedule could accelerate and some of it any ways could come into Q3.

[Gregory Palm:]

And so just to confirm that is a multiunit order from a single customer. And if you don't get it in Q3, is it almost certain that you will get the order recognized in Q4? Or is there any chance that [indiscernible] even in the next year?

[Jeff Bertelsen:]

We don't see that today. I mean, as we said in today's release, we expect to get this order in the second half, and our anticipation is it's going to revenue in Q4.

[Gregory Palm:]

Got it. Okay. I wanted to shift gears to the MRS opportunity in the front and mid-end. I know previously, you were talking kind of two to four years for mid and front-end, respectively, in terms of being able to enter that market. So first, can you just maybe confirm that? And second of all, I mean, you've clearly made some additional accuracy improvements there, what's the likely go-to-market strategy? Have you thought about that any more in terms of whether you go with an exclusive provider or whether you open it up? Just curious to get your thoughts there.

[Subodh Kulkarni:]

All good questions, Greg. So first on the technology side. So as we have said before, we continue to make rapid advancements in our ability to measure smaller and smaller features. So those stuff, that is all commercial right now, whether it's SQ or what we sell to KLA or Nordson, is all over 100 micron features, just to set the baseline. So that's what is commercial today. In our lab, about three months ago, we said we were able to do more than 50 microns. We have made advancements, and now we are able to do 30 microns and above. And that certainly has peaked, and we are showing those sensors and our capabilities to select customers right now. Particularly, the semiconductor industry because that's where they evaluate the most and mid and the front-end is obviously available to the semiconductor industry. And they certainly see a number of applications once you are in the 30-plus-micron range, things like cracks that happen and they take a silicon wafer and dies it. There's a lot of micro cracks that happen at the edges. And those micro cracks, it seems the transistors are coming so close to the edge. Those micro cracks are better when it comes to the chip yield. And right now, the only option they have are -- go with some kind of a microscopic technique, which is extremely slow or not do any inspection, in which case, they take it and yield it. So when they see a technology like MRS becoming evident in the next year or so, that can help them do things like crack inspection that gets them pretty excited because they are significantly faster than any microscopic technique. So we feel pretty good based on the customer feedback we have received, that this 30-plus-micron sensor, which is what we are referring to as mid end, if you will, we should be able to generate significant commercial dollars in 2018. And the other -- crack inspection is just one that we cited, but there are many other inspection areas that need those kind of dimensions. So we feel overall very good about hitting those kinds of numbers and time line that we have disclosed in our investors presentation for mid end and front end.

[Gregory Palm:]

That's great color. Can you remind us in terms of quantification, what you see is sort of the mid-end and front-end opportunities again? And how that compares to the opportunity in the back end?

[Subodh Kulkarni:]

Sure. So what we have said in our investor deck is that once -- for the back end inspection, which is more than 100 micron area, we have said that once we are -- and our only go-to-market route for the back end inspection is through KLA-Tencor since we have a mutually exclusive long-term agreement with KLA-Tencor. We have said that in -- next year, once we are scaled up completely with KLA-Tencor, that's a roughly $10 million per year opportunity for us, in the back-end area. Mid-end area, we have said that is 2 times to 3 times size of the back end area but will lag by a couple of years. So by 2020, we are saying the mid-end opportunity for us is a $20 million to $30 million opportunity. And what we are saying for front end, which is a soft end micron area, we believe it's as much as -- the next opportunity. So we are saying roughly it's a $100 million per year opportunity in 2022 time period. So when you say it's a semi area for us with MRS this year would mean high single digits, next year will be in $10 million neighborhood. The year after that, the $10 million will continue but mid end will start contributing to some level in 2018 and significant level in 2019. So by 2020, that you'll have a big number from mid end and the back end will continue. So overall, the semi numbers are going to look like an exclusive growth category for us with MRS sensors.

[Operator:]

And we'll now take our next question from [Miles Jennings].

[Unidentified Analyst:]

I'm sort of interested in this WaferSense for fabs, and I know that in the last year or so, this is really the time period when you have first tried to address that market. And perhaps, you could give the percentage of revenue in the second quarter that came from fab. And also you mentioned that there is sort of a largest order of $500,000 or so from a fab in the WaferSense area, was that from 1 fab? And perhaps, you could give us a sense of the market for the fabs in WaferSense.

[Subodh Kulkarni:]

Okay. So first, WaferSense in general, right? I mean, we have said in the past, Miles, that there's about 170 active fabs worldwide right now and new ones are being built. Right now, semiconductor industry is in a very strong investment cycle as many of you know. And so the number continues to increase rapidly. Most of the new fabs are being built in Asia as we speak. So that's the active number of fabs. In each fab, there's a roughly 500 pieces of, what they call, tools or stations, and we serve six applications right now. So if you just do a multiplication, that will give you about $0.5 million potential places where we can sell WaferSense. And so far, we have sold of significantly lower number than that. Actually, are in the 2,000 to 3,000 WaferSense installed so far. So we have plenty of upside room to grow with WaferSense going forward. So that's at the global market total time, if you will, level. Regarding your specific question, yes, in Q2, one of the larger semiconductor fab manufactures basically commissioned a new fab in Asia. And then they usually commission a new fab that's when the final equipment is being purchased and accepted and installed. We get a natural benefit of a new fab coming online and there was a large order. They'll obviously, come back and order more, and they are one of our larger customers right now. They are also one of the largest semiconductor companies in the world. So they will continue to buy smaller number but that fab does not need another $0.5 million this current quarter. I think just what Jeff was referring to when he said, we had a large order that won't repeat WaferSense in Q3.

[Jeff Bertelsen:]

Yes.

[Unidentified Analyst:]

One more question. And I realize this isn't fair, but I'd like to ask it anyway. In the last conference call, regarding this large OEM, you said that they are continuing to evaluate the product and that your sense was it could be two weeks or two months. And since it's now three months from when you said that, are you still further thinking in terms of two weeks to two months?

[Subodh Kulkarni:]

What we can just say, and I think that's what we said is that they have our sensors, they are in their -- they're subcontracting locations. They don't actually own the factories. And so they are in their sub-con locations. They are collecting data as we speak. And the data is being compared primarily with manual inspections. So right now they're operation users for a lot of stations where they do manual inspection. And they are testing the long-term viability of whether MRS can be the answer to replace manual inspection stations. And that's what they are using the sensors for in collecting data. Everything we see, we feel very good long term that MRS is indeed the answer to replace manual inspection but regarding then the hand that pulled the trigger, if you will, to get this manual inspection with MRS that's totally up to them obviously. So -- and because they are such a large company and have the potential, they don't think -- we don't use any of their potential future numbers in our guidance just because it will skew the numbers quite a bit.

[Unidentified Analyst:]

Yes. I do realize that. So as far as you're concerned, it's still a very active potential customer over the long term?

[Subodh Kulkarni:]

Exactly. So I view them as a very strategic customer because they notably can either direct customer but they have huge influence on the ecosystem or the products spending period, and that does clear all in how many SKUs and CyberGage as we sell. So they have indirect influence on their ecosystem quite a bit. So they are an important strategic long-term partner and clearly, we continue to work with them to see if MRS will be a viable solution to when they go full automation and want to get rid of manual inspection, which is what they are really looking for in the next five to 10 years. But I do view them as a long-term customer, strategic customer.

[Operator:]

And we'll now take our next question from Ash Birla with Dougherty & Company.

[Ash Birla:]

Subodh, I wanted to ask you guys about DRAM and WaferSense. So there's a lot of DRAM capacity expansion and technology expansion that's going to happen in the next 12 to 18 months. Historically, it sounds like WaferSense is more relevant for logic in 3D NAND and those kinds of tools that go into those kind of fabs. Can you help us understand how this new DRAM investment coming. Is that going to drive WaferSense investment at all?

[Subodh Kulkarni:]

The answer is yes. I mean, any investment whether it's DRAM, NAND or logic helps WaferSense without a doubt. Regarding WaferSense, and conditions for more with logic and NAND, I don't know if -- if I said that in the past, I was not right, okay? I mean, WaferSense is really -- is needed when you are using more advanced equipment. So 300 millimeters more than 200 and 150 and their use matter. So if you take an old fab that is using 150 millimeter and making low end parts made for them, WaferSense value is not that high, but if you take an advanced logic or NAND or DRAM, fab, it's 300-millimeter state-of-the-art equipment, that's really where WaferSense value shows up just because the equipment is expensive, time is of the essence to them. And that's really how WaferSense becomes valuable to them. So again, putting back -- all the investments that are going to happen in the fab environment in the next few years, certainly will help WaferSense growth.

[Ash Birla:]

Okay. The $3 million in orders, you don't actually have those orders, you're expecting those, right?

[Jeff Bertelsen:]

That's correct.

[Subodh Kulkarni:]

Right.

[Ash Birla:]

Okay. Just one final one. When this is like more of like a high level question, 3D MRS, that sensor with AI and everything is driving SQ3000 and KLA is buying that. What is the difference between ASPs of ICOS tools and SQ3000? And what else is KLA bringing to the table for their customers?

[Subodh Kulkarni:]

So we obviously can't disclose the exact number of that. We still (inaudible) KLA for. But suffice it to say, typically in SQ system and this price of $170,000, but after average price that we see in the marketplace is closer to $125,000, $130,000, including some options and stuff like that. That's typically the range for SQ3000 today. Obviously, we sell the sensor to KLA-Tencor at a lot lower price than that, a few tens of thousands of dollars. KLA in turn puts -- it's a critical component of their inspection system, and sometimes they have one sensor, sometimes they have 2 sensors depending on the applications. But they -- if you look at their ICOS line, which is where our sensors go in. They are a fairly complex mechanical part handling system. So they have traced, what they call, dirac trace coming in of its multiple modules and packages. And the robot is bringing the train, picking the packages up, putting the packages in the fill on few of the sensors. One sensor on the top, one sensor on the bottom is inspecting, then the packages go back into the dirac trace in the pass and fill categories, and they come out. So the system inputs dirac trace, typically output dirac trace, with pass and fill components in different areas. There's a lot of mechanical handling for this fine small parts. Typically, in ICOS system, you will have to talk to them, but it typically goes for hundreds of thousands of dollars, but they are complex systems. Our sensor is, obviously, 10% or something like that of the total system.

[Ash Birla:]

Okay. Sorry, just one last one. You were going 50 micron inspection in the lab, and now you're at 30. But you weren't able to or you're not commercialized at 50. What is the kind of time line for 30? And some of -- have your SQ customers or CyberGage customers, have they requested 30 micron inspection?

[Subodh Kulkarni:]

SQs, not really because in the PCB manufacturing, you don't have parts that go down to 30 micron, and when I say 30 micron, the full feature has to be 30 microns. And so the Sigma, if you will, the standard deviation, we typically look at two to three microns, three Sigma of standard deviation. So we've got extremely accurate fine resolution part. So typically PCB does not have anything close to that. So PCB is fine with 60, 70 micron and above. And right now it's typically does 100 micron. The higher resolution SQ that we launched earlier this year, that does like a very accurately up to 60, 70 micron or so, which probably is good enough for the PCB industry for the foreseeable future. We don't think that 30 micron-type stuff is needed for PCB industry, for 3D AOI-type stuff. CyberGage, as I mentioned earlier, again typically, manufacturers are not looking at these kinds of technology for 30 micron-type features semiconductor industry obviously cares a lot. And then maybe some plastic and metal manufacturers are being -- may not have run into it right now, that care about it. But so far, what I see is 30 microns, 50 microns, all these kinds of stuff is important for the semi industry. We haven't quite seen applications outside yet, but there maybe some. It's just I'm not sure.

[Ash Birla:]

So what is the time line for commercializing then because 50 is not commercially yet, right?

[Subodh Kulkarni:]

50 is not commercial but since 30-micron development came pretty fast, I mean, obviously, the semi industry pushes us towards the small a number as we can give them. So we're pretty confident we will commercialize the 30 micron and above so that may includes the 50 micron stuff, in 2018. So we are confident we'll have some significant dollars coming in 2018 from the mid-end area right now.

[Operator:]

And this is the conference operator. It appears there are no further questions in the queue at this time.

[Subodh Kulkarni:]

Well, thank you for your interest and questions. We look forward to updating you at the end of Q3.

[Operator:]

And ladies and gentlemen, that concludes today's conference call. We thank you for your participation.