From Hybrid Bonding To AI Power: Live At SEMICON West

The floor in Phoenix was packed, and so were the ideas. We sat down with innovators across the stack—equipment makers, metrology experts, logistics strategists, and software leaders—to map the real state of advanced packaging and what it takes to build, measure, move, and power tomorrow’s chips.

EV Group kicked things off with a candid look at die-to-wafer realities: activation on film frame, then 100% overlay metrology that measures tens of thousands of points per hour so every die and corner is verified. They also unveiled LithoScale XT, a fully digital, maskless lithography system printing 300 mm at 60 wph—perfect for massive AI dies and fast design turns. 

Lab14 widened the frame with a portfolio approach: direct-write lithography, single-wafer processing, data prep, and analysis tools working as a coordinated line, with data sharing and AI feedback baked in.

Resilience and regionalization came to life through Kuehne+Nagel’s on-the-ground view: supplier clustering near fabs, cross-border trucking, time-critical services, and 4PL integration that gives real-time visibility and smarter capacity planning. 

ERS showed where throughput meets cost: photothermal debonding with lower stress and reusable glass carriers, demo centers in Taiwan (and planned in North America), plus surge demand for warpage repair as volumes rise.

Process control is moving into packaging with front-end rigor. Nova detailed metrology for hybrid bonding, chemistry monitoring of plating baths, X-ray and XPS/SIMS material insights, and the handling know-how to measure framed wafers and panels reliably. 

Nordson Test & Inspection highlighted AI-driven inspection, ultra-fast acoustic scanning, automated X-ray metrology, and sensor wafers that cut tool downtime and sharpen process windows. Comet showcased its CT and CA20 upgrades for 3D IC and TSV analysis.

Power dominated the later conversations. Siemens argued we need to design for energy from the chip through the blade, rack, and data center, simulating real workloads and cooling to slash gigawatts—then extend that thinking into the fab, where optimizing chillers and facilities already saves serious money. 

Onto Innovation brought it home with execution: the PACE Center now hosts partners’ tools, accelerating experiments for glass, TGV, and panel processes without waiting on public funds.

If you care about hybrid bonding, maskless lithography, CT for 3D ICs, panel-scale packaging, or cutting AI’s energy bill, this one is dense with takeaways and hard truths. Subscribe, share with a colleague who lives in the fab or data center, and leave a review telling us which insight you’ll act on first.

Transcript

Opening And Semicon West In Phoenix

SPEAKER_02 0:00

This episode of the 3D Insights Podcast is brought to you by SEMI, the leading microelectronics industry association with programs that help its members grow their business and address top challenges worldwide. With a global focus on advocacy, the microelectronics supply chain, sustainability, and workforce development, SEMI works with industry leaders to align goals, share best practices, and accelerate progress. Learn more at semi.org. Hi there. I'm Francoise von Traff, and this is the 3D Insights Podcast. Hi everyone. This week we are recording live at Semicon West 2025, and there are so many changes this year for this event. It's in October instead of July, and that's because it's being held in Phoenix instead of San Francisco. And we all know how hot Phoenix is in July, so I think that was a good decision. So in this episode, I'm speaking with a number of our member companies who are here participating in the event. And they have some of them have things to celebrate and news to share. So speaking of Arizona, my first guest is Thomas Urman of Eevee Group, who are celebrating 30 years of their US headquarters in Arizona. So yeah, welcome to the podcast.

SPEAKER_07 1:30

So 30 years EVG in Arizona, right? So it's a little bit younger than actually uh we are.

SPEAKER_02 1:36

So well, but still, you know, the thing is, you've been part of the Arizona semiconductor ecosystem uh for a long time.

SPEAKER_07 1:44

Yeah, absolutely.

SPEAKER_02 1:45

So how do you feel about all these other companies coming here coming here?

SPEAKER_07 1:48

I could actually say that we had a great vision 30 years ago, probably that it it was part of it, right? So um it all started out basically um when it was still Motorola at the time. And now we are since uh several years sit uh we are sitting in the a on the ASU campus. Great place to be actually in the US. Um great collaboration. We have uh access to clean room in the in in ASU, so uh we can do customer demonstration, and uh well, it's it's just very nice to be here in Arizona. Yeah.

SPEAKER_02 2:19

So what are you doing to celebrate?

SPEAKER_07 2:21

Well, we have a couple of party activities uh actually for today and the next days, uh so with the team and also of course with our customers. For example, also gathering out there in the so-called Bourbon Range, which is between the the exhibition centers uh organized by SEMI. And well, we'll be will be just a very good you you will be there, of course. Of course you will be there. Of course.

SPEAKER_02 2:43

Okay, we didn't miss it. So you have some other news to share. You've got a few product launches?

SPEAKER_07 2:49

Yes, actually, we released a couple of products in the last half a year. It was quite uh quite busy for our product marketing team. Um so essentially what we started with, we had further activities on the on the die-to-wafer side, right? Um EVG is market leader for wafer-to-wafer bonding, a lot of our processes also fit to the die-to-wafer world. In this specific case, and if you remember back, and probably some of our your listeners also remember back, so we have launched uh die-to-wafer activation systems a couple of years back, which is called EVG320 die-to-wafer, which are quite nice uh pickup in the market, that does actually activation and cleaning of dice for hybrid bonding on film frame. But now, since uh when the dice are bonded, one of the central questions is also well, how well are the dice actually placed? And that's a little bit different story than wafer to wafer bonding, because in wafer to wafer bonding, we can control very nicely a very big part of the wafer. The challenging part is actually edge of the wafer. Now in die-to-wafer bonding, you could also say, well, actually it's all edge because the die is small, so you're talking about a centimeter by centimeter, nine by thirteen or a little bit bigger. And actually that's act that's quite troublesome. Uh so it definitely needs metrology of overlay, and that's not of one die and not one or two corners, but actually every die and a lot of different uh as as many points as possible.

SPEAKER_02 4:17

So you're doing you're doing the die to wafer um on a film frame?

SPEAKER_07 4:22

Right. So the activation, yes.

SPEAKER_02 4:24

So you have one wafer on the frame and you're picking placing known to die.

SPEAKER_07 4:31

Exactly. That's then done with a partner. So we are strongly populated or strongly promoting right now the a separated approach due to the fact that our queue times uh after activations, they are very long. So we talk here weeks and not uh a couple of hours. So essentially a process separation is uh it works for us. That means you can activate the whole lot and then take it over to a dice wafer bonder. That does actually the bonding part, but then still the meteorology part is uh is is open.

SPEAKER_02 5:02

Okay, so is the bottom wafer already diced on the film frame?

SPEAKER_07 5:06

So actually we have dice on the film frame for the activation. Uh the receiving substrate, however, is a monolithic wafer, right? So where you pick and place the dice. That's uh for us not needed. We stay on wafers still, also for the metrology. So it's you can you can think about it like well, there are different processes, like for example, uh SOIC is called by uh by TSMC. Then for Intel, basically you have uh for Verous direct placement, and so it's multiple different process flows by the big customers. And well, actually, then still the task remains how to do a very quick metrology. So we we upscaled one of our metrology systems that we use for wafer-to-wafer bonding. And in this case, so the the system is called EVG40 in uh die-to-wafer. The original system for wafer to wafer is called EVG40 NT2. And what we are doing is we are measuring several tens of uh tens of thousands of points per hour. So we are for a normal HPM wafer, we are in a range of like 15-16 wafer per hour that we can measure on the tool, and not just individual points or basically just look at uh look at a couple of subsets of dyes, but we are measuring every dye, every corner. So we get a full picture of uh dye, die distortion, run out capability. So it's basically from that perspective, a hundred percent inspection or metrology that we do.

SPEAKER_02 6:29

Okay. Tell me about the lithography.

SPEAKER_07 6:31

So we also launched uh a brand new lithography system called Lithoscale XT. So it's an extension. With uh a couple of years back, we launched the first system that was actually well more for advanced RD, small production lines. Um, and now we basically followed up with the HVM system. At lithoscale XT, we are doing 60 wafers per hour on 300 millimeter, fully digital lithography. That means any wafer, every wafer by itself is processed. So we just get the vector data in, there is no mask, and we can just write any wafer individually, 60 wafers per hour. Maskless lithography, exactly. Yes. And that's a big deal, for example, also if you look for AI, for example, the dice get bigger and bigger, they don't fit the mask, yeah. So with massless lithography, here you can consider each wafer is actually one die. Okay, we don't care, so any part of the wafer can be different. We just print any structure, can also comply with uh certain things, and uh well, we can we can basically make that fully digital now with 60 wafer per hour. Right. That's uh absolute game changer.

SPEAKER_02 7:42

Well, that sounds awesome. Good luck at this show today and the party tomorrow night. See you later, and we'll see you in Europe in a few weeks.

SPEAKER_07 7:50

Yes, all right, looking forward to that. Thank you very much, Francois.

SPEAKER_02 7:58

So, my next guest here at Semicon West 2025 is Niels Weinance from Lab 14, one of our newest premium members. Welcome to the podcast, Niels.

SPEAKER_08 8:09

Thank you.

SPEAKER_02 8:09

Because Lab14 is such a new member, we've never had this conversation before. So I'd love to give you the opportunity to maybe give a bit of the origin story and your role at the company.

Lab14’s Origin And Portfolio Strategy

SPEAKER_08 8:20

I'm the business development uh manager for lab 14. And uh basically lab 14 is uh is a group of companies that are in high-tech uh making uh equipment for um semiconductor manufacturing. So basically, it's micro and nanofabrication and surface analysis tools. And uh the origin of Lab 14 uh is um coming from uh the RAG Foundation, so it's a very large foundation in Germany that uh comes from the coal mining uh actually from the exit is a result of the exit of Germany uh of uh doing coal mining. And uh they're buying small to medium-sized businesses in different areas, and we cover the technology high-tech uh side for doing micro and nanofabrication uh and service analysis.

SPEAKER_02 9:08

So uh what is the relationship between the coal mining industry then? I'm not quite understanding that.

SPEAKER_08 9:14

So um I think 2007 um so the um the German government decided to exit uh coal mining as a whole and uh basically start buying coal on the international market because the cost of mining the coal was uh about twice as expensive as buying it on the on the world market. And so um the idea was then to actually set up a foundation to take care of the uh the eternal obligations out of the coal mining system. So there was a lot of uh social uh and uh economic uh changes uh happening that um come from exiting the coal mining industry, so there was a large large workforce that has to be retrained, things like that. So to finance that um and actually taking care of the remaining mines that still exist, um, and the cost of that is being covered by uh basically investing in companies and then um the dividend pays for these these ongoing costs forever.

SPEAKER_02 10:10

That's a wow. So that's a really interesting business model because it's almost like a non-profit that you exist to pay reparations.

SPEAKER_08 10:19

Well, you don't really maintain the mines, but you have do you have to return them back to the original basically have to maintain the mines, so um they basically have to pump the water out of the mines because once that that water that that uh is in the mines goes up to uh groundwater level, it can poison the whole area. So it's actually very important that forever uh these pumps are running and pumping out away the water. And the cost, just uh cost for electricity is somewhere in the range of uh 350 million euros per year. Yeah, so that that money has to be raised somehow. Okay, and that's how they decided to actually um invest in these small to medium-sized businesses and grow them and then live from the from the dividends.

SPEAKER_02 11:01

That is so interesting. So, why the semiconductor industry as uh a target for those companies? Is it because they're high yielding or no?

SPEAKER_08 11:09

They invest actually in all kinds of different other businesses, but we as Lab 14 are uh uh basically a group that's so that started uh when they purchased Heidelberg Instruments is a lithography company specializing in direct ride lithography. And um basically they have asked Heidelberg to look for um you know partners that that could also be joining this, and we formed this group. And the idea behind the group is really that you can um imagine that um all the equipment in a in a process line, let's say advanced packaging, could be coming out of one hand, basically.

SPEAKER_02 11:46

Okay, so you're building basically uh tools to support the semiconductor supply chain. Yes. You're acquiring the companies.

SPEAKER_08 11:55

So these companies are still individual companies, but it's like uh lab 14 is the um it's sort of the umbrella, the owner of these companies, and we actually look to find also synergies in where we can work together. This could be so simplest ways to if you if you look at a semiconductor process, is you need we have wafers, you need to uh clean them and then code them. And so we have a company called Notion Systems that actually um uh is uh making single wafer processing equipment. Uh-huh. It can do cleaning, uh developing, bonding, and and things like that. And then the next step would be doing some sort of patterning, and that could be done by um Heidelberg instruments tools with doing direct write lithography. Um you need to prepare the data. So we have a company called Genesis that does uh data preparation for lithography, mostly for e-beam lithography. Maybe later along the chain you would need some inspection. So uh we have a company called uh Specs that does uh XPS, so you it can actually look into the layers and find the composition um of the layers. We have NanoSurf that uh is an AFM company that does um uh meteorology for uh measuring surface roughness, for example, for very small feature sizes. Um, and then we have some companies uh uh in the service group that actually um are in the business of uh maintaining uh uh clean rooms and also supplying uh parts for equipment that is uh needed in the clean rooms.

SPEAKER_02 13:28

That's such an interesting concept to me. Um, so where are you mostly finding your customers at this point?

Advanced Packaging Use Cases And Data Synergy

SPEAKER_08 13:36

Um I think uh the nice thing about it is that I I like to speak about its micro and nanofabrication and service analysis because it's not just semiconductor. Um I think the most interesting uh market right now, or a growing market, is uh definitely it's uh semiconductor packaging. Uh we all know that.

SPEAKER_02 13:55

And uh yes, finally.

SPEAKER_08 13:57

Yes, finally, we all know that.

SPEAKER_02 13:59

I've been telling this story for 15 years.

SPEAKER_08 14:01

And so I'm also uh looking into this for uh a couple years now, and I'm uh really excited about um the opportunities that are coming in there, but you could also look at uh micro nanofabrication in many different areas. So we're looking at AR VR glasses where uh you need uh grayscale patterning for making micro lenses, um arrays of uh gratings and uh things like that. And uh so there's other exciting things that you can think of. Uh it and I always say that in technology everything starts with making a pattern. And so I think definitely that is what's all what we're always doing as humans, is we're generating a pattern and we're kind of replicating it, and that is all what the semiconductor industry is about. But that's also in all the other applications. Um uh, you know, if you look at uh how AR glasses are being made, it's the same process, really. Right.

SPEAKER_02 14:56

So you're the group pattern experts?

SPEAKER_08 14:59

I come from Heidelberg Instruments, so my background is really uh also I come uh originally from Heidelberg Instruments, and I'm a lithography fan. Yeah, I would say.

SPEAKER_02 15:10

So in the advanced packaging space, which of these companies is really targeted toward that? Would that be Heidelberg Instruments?

SPEAKER_08 15:16

I would say so. I also like to say that that's a combination of these companies, and I would say in that uh sense, the the say um uh the companies that would target it would mostly be Heidelberg Instruments, Notion Systems, and Genesis at this point. Uh but then uh also because I come from lithography and I don't know enough about the uh analytical side. Uh I have a partner that uh um that uh is looking more into the business development on these analysis tools. Um I think I don't just don't know enough about the opportunities that we would have there. But I see that everywhere where you do uh fabricate something, you also need to inspect it.

SPEAKER_02 15:57

So uh yes, that's very important, especially as the devices become more and more complex and more and more expensive as you move down the line, right? So they need to catch any defects early and you know before the devices are complete so they can pull them out of the yeah.

SPEAKER_08 16:15

And you you you hear also about uh you know now AI in these uh advanced manufacturing processes, and uh we're excited to also think uh about how we can uh chain these tools together and uh share the data amongst the tools and actually come up with uh sort of analytical ways of uh you know correcting a process or maintaining a process, or let the tools decide uh when they need help, right? Um, these are all things that you can even see whole companies built just on that principle. And uh I think that is something that we are also working on. So to actually not only sell individual tools, but actually the whole integration of all the parts of these processes.

Kuehne+Nagel On Regionalization And 4PL

SPEAKER_02 16:59

So this is the first time Semicon West has been here in Phoenix, and it's the first day. Um you've been to you said Semicon West the last three or four years. Right. So what are your thoughts about the move?

SPEAKER_08 17:13

Um I'm still a little bit skeptical. Uh I I think, but that's normal for something to change. Right, yeah. I think the idea to uh you know have it maybe every other year in uh in Phoenix and in uh because it's undeniable that uh, you know, with the TSMC factory coming here, that uh there's gonna be a supply chain here uh in in Arizona, and there even already is. And uh so I think that uh important people in this industry are going to be here and in the in the Bay Area of Sands.

SPEAKER_02 17:45

Yeah, I think you're right, and I'm noticing the crowds seem to be a lot more on the floor this first day. So hopefully the intention of bringing in all of these people from the local surrounds and working, it's easier to have to pay the travel expenses and be able to have them just go from home. I came from home this morning, it was fantastic.

SPEAKER_08 18:04

Well, a change is refreshing too, right? So it is like okay, I've been to I go to San Francisco maybe three times a year because there's like different conferences uh uh around uh micro androfabrication. So uh I think it's nice to also go to uh with such a large conference to go to a different place. If it's if it stays this way, uh two-year cycle is actually also a good idea.

SPEAKER_02 18:26

Well, thank you so much for joining me today, Niels. Where can people go to learn more about lab 14?

SPEAKER_08 18:31

So they can go to the uh website, so that is uh uh lab14.group, and uh also of course uh contact me directly. So I hope uh on 3D Insights there is also contact information. And uh should be uh yeah, uh I'm happy to answer any questions anybody might have.

SPEAKER_02 18:48

Great, thank you so much.

SPEAKER_08 18:49

Thank you, Franz.

SPEAKER_02 18:55

So my next guest is from our elite member, CUNA Nogle, and I'm speaking with Farley Penna Florida. Welcome to the podcast, Farley. Thank you. So before we dive into our conversation, can you just um tell me a little bit about yourself and your role at CUNANAGLE?

SPEAKER_01 19:13

Sure, no problem. Again, my name is Farley Penna Florida. I've been in the industry for more than 30 years. 15 years in logistics in the manufacturing environment. Well, about 20 more than 20 years in transportation, global transportation. And I'm currently the global key account manager, okay, uh managing uh semiconductor customers.

SPEAKER_02 19:35

So I've been at some of the market symposium and um into this morning's talk, and and obviously, you know, the the supply chain is a big topic right now. So can you talk a little bit about some of the current trends from your perspective that you are seeing that's shaping the global semiconductor supply chain?

SPEAKER_01 19:54

The most important trend right now that really impacting the supply chain is regionalization or reshoring or near shoring. Okay. And because of that, but that is being driven by geopolitical situation in Asia Pacific and also the government incentives being given to particularly here by the US government to some of the companies. And of course, resiliency when it comes to uh security as far as supply chains concerned.

SPEAKER_02 20:27

So when you talk about regionalization, it's not just the U.S. that's trying to onshore. We've got it happening in Europe, we've got it happening in Asia Pacific. How much of this is has to do with some of the tariffs that are being set in the United States?

SPEAKER_01 20:45

I cannot give you a specific one, but tariff is also an impact one of the factors that made this company uh rethink about or redesign the supply chain.

SPEAKER_02 20:56

Okay, so how does regionalization improve supply chain resilience for semiconductor companies?

SPEAKER_01 21:02

A lot of customers right now, as far as semiconductor IDMs is concerned, for example, or even the contract manufacturers of semiconductors, they're trying to do the clustering wherein suppliers are being being asked to expand within proximity of the production uh site so that they would be able to support whatever being demanded by those customers from upstream.

SPEAKER_11 21:30

Right.

SPEAKER_01 21:30

And also they're trying to support uh the time to market needs of their customers. So shortening the lead time is one advantage for that.

SPEAKER_02 21:42

One of the speakers this morning, one of one of the keynote speakers this morning from CUNY was talking about the importance of localization, bringing your supply closer, manufacturing closer to your customers. Correct. So how does that impact a logistics company like CUNENAGO?

U.S. Packaging Onshoring And Automation Outlook

SPEAKER_01 21:59

Well, on our side, that is positive. Why? Because particularly Conanago, because we are ready to support these customers, right? Number one, we have plenty of solutions in our service for pollution that would be able to support these customers. Number one, cross-border tracking. Right. In Southeast Asia, Mexico to US, we have that capability. And in fact, we're serving a lot of customers right now in that. And also the local deliveries, right? Domestic delivery. Right. So for example, Malaysia, within Malaysia, because the some of the suppliers are within that region, we have the capability to deliver that services. Same here in the US. Uh time critical uh service, uh like the NFO, the hand carry type of stuff, the white gloves. We have also capability to do that. And the good thing about that is that we are on we are not only addressing the upstream, which is more on the capital equipment side, we are also addressing the finished goods sides, which is downstream.

SPEAKER_02 23:02

Okay. So does it reduce the need for a lot of the intercontinental transport that you're doing?

SPEAKER_01 23:12

No, there's still there's still a demand for that. And that's the reason why one of the solutions that we can propose to some of these customers, as far as concerned, is our ability to do the strategic planning as far as capacity is concerned.

SPEAKER_02 23:30

Okay. So they basically make one phone call to you, you determine what their needs, and you set up a plan for them.

SPEAKER_01 23:35

Correct.

SPEAKER_02 23:35

And when you're dealing with like the last mile, are you qualifying those companies to make sure that they have all of the because you're you're using third party at that point. So are you qualifying those companies to make sure that they handle all of those, like you mentioned before, white glove situations?

SPEAKER_01 23:53

We have a very stringent uh procurement uh policy when it comes to selecting vendors, and we are applying that when we try to determine the list of vendors that we will be putting in our pool when it comes to supporting us on those uh last mile delivery.

SPEAKER_02 24:11

Um, how can CUNANAGL evolve from being a service provider to a strategic partner in semiconductor supply chain transformation?

SPEAKER_01 24:18

That's a very good question. That's the one of the big advantages of Kunanagle. Kunanagle is not only focusing on 3PL, but we are also offering 4PL service.

SPEAKER_02 24:27

What is 3PL and 4PL service?

SPEAKER_01 24:30

The 3PL is more on the transportation and warehousing.

SPEAKER_02 24:33

Okay.

SPEAKER_01 24:33

Whereas the 4PL, the integrated logistics, is more on the systems. Meaning to say from point A to point B, we, for example, the service provider providing all the transportation, the warehousing needs of the customers, whereas the 4PL services will allow us to integrate with their systems so that they will have the customer will have uh real-time disability as far as uh the status of their shipments so that they can make more effective and efficient uh decision making in the supply chain.

SPEAKER_02 25:07

I see. Okay. Now, specifically, a lot of our listeners are in the advanced packaging space.

SPEAKER_03 25:13

Right.

SPEAKER_02 25:13

So in the US, specifically, advanced packaging is an issue because we don't have any high-volume OSATs here. Amcore is building one here in Arizona, TSMC is supposedly building packaging facility, and um Intel does some packaging. But most of the packaging that happens, I think the number was 80%, is sent back offshore to the Asia Pacific region for the packaging, mostly due to cost and the lower cost of that. So um what are you seeing as far as the onshoring is concerned from the packaging and maybe the substrate materials perspective?

SPEAKER_01 25:52

Now, before I answer the questions, I want to give you a proper context on this one as far as market share is concerned.

SPEAKER_11 25:58

Right, okay.

SPEAKER_01 25:58

In semiconductor industry, 50% or more than 50% of the market share as far as global transaction is concerned is happening in Asia Pacific. Right now. The races between US and Europe. But because of the push coming from the US government, for example, here in the US, and the incentives given to these customers to bring Fab and also the back end here in the US. I my projection in this one is that backend operations will have will be happening here in the US with the reinforcement or support of robotics inside the manufacturing staff.

SPEAKER_02 26:38

So it'll be more automated.

SPEAKER_01 26:39

More automated.

SPEAKER_02 26:42

You're ready to go.

SPEAKER_01 26:43

Ready to go.

SPEAKER_02 26:44

Okay. Well, thank you so much for joining me today, Farley. It was a pleasure.

SPEAKER_01 26:47

Thank you.

SPEAKER_02 26:52

So my next guest is Laurent Giaminier from ERS, one of our premium members. And you've been members of the 3D Insights community for quite a few years, but we haven't had you on the podcast.

SPEAKER_00 27:04

That's true, Farnus.

SPEAKER_02 27:06

This is our first time, right?

SPEAKER_00 27:07

Exactly. Happy for happy for our first time. Thank you for having us.

SPEAKER_02 27:10

It's so nice to finally have you here. And here we are at Semicon West in Phoenix, which is a change from San Francisco.

SPEAKER_00 27:18

Um, you've been to San Francisco for I have been to Semicon in San Francisco, and I see it as a very positive change from this morning queuing, a long time to come in and seeing so many people on the booth. This is a really good change. Happy to be in Phoenix, Arizona today.

SPEAKER_02 27:35

Yes, and it's gonna go back and forth. Um but I but I love that the goal that I think Joe Stakun has set out to have so many of the local people because we have such a big semiconductor ecosystem here in Arizona and it's growing and getting bigger. So um your US headquarters is in Austin, Texas. No, Dallas, Texas. In Dallas, Texas. Okay. Any thoughts about opening space here in Arizona?

SPEAKER_00 28:01

Not yet. I mean, we have a discussion with uh with people in Arizona. I mean, Arizona is very attractive for a lot of reasons in the SUMI, and uh we have a lot of good contact in the local, in the educational, with ASU. So why not? But not immediately. You know, we have our people traveling to Phoenix. Um but it's a good place to be in the future. And I'm very glad that uh Semicron established an alternative to San Francisco, really.

SPEAKER_02 28:27

Yeah, okay. So news from ERS last February your first opened the Regensburg facility. So how's it going?

SPEAKER_00 28:39

It's going very well. I mean, to the point that we we duplicate this uh this customer demo center uh after the success we had in Regensburg. Uh we it was still too complicated to ask our customers in Taiwan to send the panel and uh and wafers to uh Regensburg. So we duplicate uh the customer demo center, and we now have a customer demo center in Zumbei, Taiwan. Uh and uh I mean we for the time being we only established a photothermal debonding capability there. That was machine with debonding with flash lamp and uh stress stress-free for the wafer of the panel, uh, with a lot of success. I would say a lot of customers' evaluation that I wouldn't have thought we would have uh came to uh to Zube. I don't think they would have traveled to Germany, I don't think they would have sent us uh the wavefer panel to Germany. So confirm my feeling that being anchored in the geography, having our team with a machine locally, so we will continue to deploy machines. America is on the card. Today we have one machine with Pulseforge in Austin.

SPEAKER_02 29:42

Okay. But we do have a demo center in I wouldn't call it yet a demo center.

SPEAKER_00 29:47

I mean it's we have a machine in in the US at Pulse Forge factory. Uh but we we will consider putting one. Uh we are looking at the possibility to do that eventually with uh uh in the with a university. Or in the education sector, getting the benefit of a university interested in advanced packaging, that could be an option. Or on Customer Demo Center in America, it's on the company.

SPEAKER_02 30:11

So it seems, and you know, this is one of the things that they've been talking about in the um in the keynotes, and um, I think it was the key one keynote today from um DuPont, um, which is being rebranded to CUNY, I think is the name of the company. And he was talking about the trend of local um localization, of having a presence where the manufacturing is, helps the customer. It also helps the company. I would imagine it also helps in the sense of sustainability that you know you're the closer you can be to service your tools, you're not doing that carbon footprint of backing, you know, flying back and forth or being able to have your customers come to some of a uh demonstration center that's local to them versus having to fly, like you said, all the way to Regensburg.

SPEAKER_00 31:03

I would lie if I said that was our objective to have a lot of people. Well, I know, but it's like an absolute happy coincidence. No, no, no, but it's a nice point. It's a valid point. But the the the truth is customers like to get easy access to the technology. Okay, flying an engineer from Taiwan to Regensburg, I mean to Munich and then driving to Regensburg, uh, flying a customer from North America to Central Europe, not easy. Having a demo center in the location helps a lot. We were astonished by the power of uh Regensburg. That's why we duplicate in in Taipei or Zhubei, and we are ready to duplicate in North America. We probably underestimate the number of decisions in advanced packaging that are done in North America, even if uh machines are in other place later. Uh so it's, as I said, we're looking at the possibility to also do that in in North America for the North America decision makers, for the education sector, because there is a lot of university in America focusing on advanced packaging, willing to do some experimentation. Photothermal debonding is is very different. That's uh kind of it's changing uh the landscape from laser debonding. I mean, providing a solution that is much less stress on panel and wafer, and that's proven, having a higher throughput, much higher throughput than laser debonding, which is very long as a process, and finally having the reusability of the of the glass carrier. The consumable in the advanced packaging is a big, big equation. Uh, being able to reuse the glass carrier a lot is is very attractive to customers. So, with all of the success we've seen in the recent month in photothermal debonding, we are ready to deploy this technology in other place.

Warpage Repair And Panel/Wafer Flexibility

SPEAKER_02 32:51

You know, the temporary bond-de-bond situation has been a long-standing challenge. And I feel like we're finally gotten somewhere where there's a really viable solution.

SPEAKER_00 33:01

Yeah, there is attraction. The solution was there. The viable, it was viable. It's a time, it's a proper time. Same for panel, for example. I mean, panel.

SPEAKER_02 33:09

Can you do this with panel?

SPEAKER_00 33:11

Yeah, we have the same semi-automatic machine that we have, namely the Loom 600S1, which is a semi-automatic machine, can do wafer 300 and 200 mm, whole panel up to 600 millimeters, 600 by 600. So you're also a customer buying one equipment as a semi-automatic in a in a lab will be able to do panel and wafer on the same machine. Uh and that machine is very affordable uh for customer to start doing their photothermal debonding.

SPEAKER_02 33:41

You know, when I last time we spoke was before I was podcasting, I think. And I've probably interviewed you for a blog. And you know, you've come so far from the shock and lock.

SPEAKER_00 33:53

That's true, yeah. Right? That's the shock and lock. Shock and lock is still there. No, it's still there, but you've exceeded.

SPEAKER_02 33:59

It's still, I mean, you were at that point, you were just the um that the chuck, right? And then you've got the tools around that, and then you built the whole thing.

SPEAKER_00 34:07

The thermal debonding, yeah, the thermal debonding, the three tamp slides. But surprising enough, thermal debonding, while laser debonding is is there, photothermal debonding is the next up, next valid option and getting traction, as I told you, either for wafer or panel. Uh, but we we still have demand for the traditional EVLB package, traditional fine-out and and traditional thermal debonding. Warpage become even bigger. Okay, advanced packaging deployment is there, it's happening. Challenges remain the same. Work page, uh die shift in residuals, still the same equation. Workpage is a big deal. Having the possibility to repair work page on a panel, on a wafer remains a very attractive point for ERS on the market. We have a lot more demand now that customers are really facing the volume issue because having some work page in small low volume production, who cares? But having high yields, you know, or low yield, sorry, high, high risk, high problem of work page and low yield on a volume manufacturing is a bigger problem. So today we have a lot more customers coming to us with the willingness to repair some work page in panel and in wafers uh at a speed that I would not anticipate to be fine.

SPEAKER_02 35:25

Well, and it's been a pleasure to talk to you today, and it's uh it sounds like they're shutting down around us.

SPEAKER_00 35:31

Yeah, we better. Thank you, Francoise. Thank you for coming. Are you?

SPEAKER_02 35:38

Yes. So this is like a nostalgia trip to the Semicon Europa will be my final show. My last show, my retirement party.

SPEAKER_00 35:47

Your retirement, are you?

SPEAKER_02 35:49

So the Stamptisch, you need to come to the Stamptist.

SPEAKER_00 35:51

Well, we need to go to Salzburg then, then if it's your if it's your retirement trip.

SPEAKER_02 35:55

So it doesn't mean I'm not gonna travel anytime. No, you will, I'm sure you will.

SPEAKER_00 35:59

So Stamtish is on my card. I think it's on Tuesday. Monday, it's a Monday. It's on my agenda.

SPEAKER_02 36:05

You have a table.

SPEAKER_00 36:06

We have a table. Yeah, we have some visitors and we invited them to come with. So we brought, I think we even have a Korean.

SPEAKER_02 36:12

And we need to put a team together for the ice talk sheets. We will. Okay, awesome. Thank you for the time.

SPEAKER_00 36:16

Thank you very much, François. And I mean, we'll meet in November, but in advance, happy retirement.

Nova: Process Control For Advanced Packaging

SPEAKER_02 36:22

Thank you. And looking forward to me. No, we have time. Thank you, François. Thank you very much. Thank you. So I am here at Semicon West on day two, and I am visiting with our premium member Nova and speaking with Tom Larson. Welcome to the podcast, Tom. How are you? Doing good. It's uh fun this week. It's a little bit different. I'm on my I'm as I told you earlier, I've got senioritis because I'm retiring at the end of the year. But thank you. So this is your first interview, Nova's first interview on the 3D Insights podcast.

SPEAKER_04 37:01

Oh, cool. I thought we had done it before, but no.

SPEAKER_02 37:04

No. So can you first just tell me a little bit about you and about Nova and your um role there?

SPEAKER_04 37:09

Sure. I mean, I'm uh uh I came to Nova about 10 years ago when they acquired a company I was a co-founder of. Um it forms one of the main divisions of the the company. We have three main product groups, and we're headquartered in Tel Aviv in Israel. Um Nova's been around since 1994. Yeah, so long time. And uh transformed a couple of times in that history and probably will do so again. For the early part of the transition into Nova, I was running marketing for the division that was my company, and then I moved into uh more corporate development function about five years ago, and now on the executive team running uh mergers and acquisitions and some of the strategic elements of the of the business.

SPEAKER_02 37:55

So Nova's core competency is metrology solutions, and you said you had three divisions. Is semiconductor one of those?

SPEAKER_04 38:03

We primarily focus on semiconductor, and we're uh almost exclusively in semiconductor. We're not doing anything in you know batteries or outside of that space. So our core competence is definitely semiconductor. We're primarily uh process control focused. Okay. So and and generally process control shows up in metrology aspects of the case. Right, yeah, but it can also show up in other areas like one of our divisions is measuring purely uh aspects of chemistry. Okay, and so it's not looking at wafers or doing any kind of dimensional measurements, but is making measurements of uh quality of the chemistry. So the three groups that we have are fundamentally based on technologies in a way. There's an optical group that does dimensional measurements, that's primarily the original Nova core. There's the group here which is doing material measurements and also doing some dimensional measurements, but using uh charged particle optics like XPS and SIMS and using X-rays as a as a utility as opposed to optics. Okay. And then we have in in uh Germany, we have a couple of business units. One is doing the chemical measurements that I mentioned, and the other was an acquisition we completed earlier this year that is primarily focused in packaging on uh metrology of advanced packaging products.

SPEAKER_02 39:17

So one of the things that I always like to point out is the difference between process control and just doing inspection, is that you're actually using that information to go back and adjust the process recipes to improve the yields that way versus just identifying defects and removing them from the line.

SPEAKER_04 39:36

Aaron Powell Yeah, that's generally accurate. I think uh you know, fundamentally the the mission of process control is a few things. One is to to make sure that the process is is in its window. Right. And the only way you know if it's in its window or not is you either measure it at some point right away or you wait until you test the devices and see if they work or not. Right. So it's kind of expensive to wait until then to find the failure. So we try to find strategic places in the process line to look for where the most variability and most uh control is required. The most extreme today is like CMP. CMP, I like to tell people, is more like uh taking sandpaper to a window. You know, just grinding stuff until it's very polished. It's a crazy process to do to nanometer scale devices, but it worked. Right. But in order to make it work, you need to look at almost every wafer before and after. And you and so there's a high intensity of measurement required there, not just from us, but from other aspects of the process. And then there are other processes that are incredibly stable and they don't move. Once you set it, it just doesn't move and it doesn't need as much metrology. So we tend to focus on the most critical areas that our customers are encountering across the device scale.

SPEAKER_02 40:46

Okay. So now one of the conversations I've been having for you know the last few years is the uh the how advanced packaging, as it becomes more complex and more wafer-level, and the purpose of it is more about connectivity and not about protection of the package. We're seeing a lot of front-end processes. And integration, I would say. Right. And you know, we're seeing a lot of the traditionally front-end processes like CMP are now being used, for instance, for hybrid bonding as far part of the surface preparation because you need a completely planar surface to do a hybrid bond, right? So can you talk a little bit about how that transition has evolved and when metrology suddenly became really critical to not just the front end but front for advanced packaging?

Chemistry Monitoring And Handling Complexities

SPEAKER_04 41:32

I mean, it's been kind of a slow creep until it's all of a sudden. Right. You know, and and I think um it's been the past couple of years when we've seen as we were talking offline the the the need to move the density increase from uh kind of a lateral scaling to a vertical scale. And so now I have to add chips that do different functions that normally I would have integrated right into the silica. Right. But I can't do that anymore because I just don't have room. And I don't have the mask, I don't have the yield, I don't have all the things, so it's better to just do this. So once you start making the complexity of anything uh high, the chances for failure go up. Right. Right. And so you need to start making paying attention to things before you didn't have to pay attention to. And in the case of packaging, you know, we always had RDLs and we always had, you know, putting a chip into a onto bumps or C4 process, and we've had that all along. It's just now on steroids, right? That process, and and it's in three dimensions. Right. And so, like I said, the complexity goes up, the need to find and and gauge for failures goes up, and and so the intensity of metrology and inspection and all that is just naturally following the scaling function of the of the complexity. Okay.

SPEAKER_02 42:53

Is there an optimal metrology technology for advanced packaging?

SPEAKER_04 42:56

Uh just like in in the front end, I mean that every problem has kind of its own optimal solution. Uh even in process, you know, there's not an optimal etch for every given integration. There's not an optimal lithography for every layout. There's not so so customers have to kind of pick and choose what's best for them and what actually fits the need. In the front end, there are certainly competitive instances where a customer can select from a number of different vendors based upon things that go beyond just performance and actual uh capability.

SPEAKER_03 43:31

Yeah.

SPEAKER_04 43:31

Packaging has the same sort of situation. You know, you from inspection to to metrology to to all the things that that you need. It's in some cases we'll offer something that nobody else has. You know, in some cases we'll offer something that others have, but we try to do it better. Right. You know, and and I think that's just the evolution of any kind of product uh effort. So I don't think there's any specific one metrology that solves all problems. Right. That doesn't exist, if that was your question.

SPEAKER_02 43:59

Well, what I was thinking is that in the front end, there are standards for the front-end processes. Advanced packaging, somebody asked at the IMAPS executive summit a couple of weeks ago, one of the questions from a company that makes materials originally focused on front end was are, you know, when are we going to see standards in advanced packaging? And all of us who are in advanced packaging looked around and said, Yeah, never. Because most of the time an advanced packages are speci, you know, it's application specific. And anytime a new architecture is qualified in, it never goes away. It's just another tool in the toolbox. So I would imagine that metrology, just by virtue of advanced packaging is being so diverse, is that you have to have multiple tools based on what you're building or you know, I I'm not sure that the reason that we exist in the front end is because of standards. Well, not no not your reason for existing, but that it's easier to have a specific process. You know?

SPEAKER_04 45:00

Yeah, I don't know. I mean the standards really are on how you interface to things and how you how you land the tools in the fab. And that makes everything easier. And in packaging, true, that hasn't existed, right? You haven't had a Jim Sex protocol in the Right. Well, maybe you have, but you haven't had it like really deeply evolved like it is in the front end. Or all those sort of standardization protocols, those are important to take the cost out of what we do so we can provide and focus on the things that are most important, right? So we don't have to think it's like one of the presidents used to say, I only have blue suits because I don't want to have to make a decision about what suit color I'm gonna wear. You know, it's like that. I don't want to have to decide what the layout I need to fit into, right? So things things like that I think are are advancing in packaging. Now and packaging, the handling is the thing that differentiates from the front end. In the front end, I'm I'm moving wafers around that are 300, 200 millimeters. In packaging, what am I doing? Well, I'm moving frames around, I'm moving reconstituted wafers around, I'm moving die strips around, I'm moving, you know, I've got I've got all these different configurations of handling that are, I don't know that there are many standards for that. Right. That's kind of a custom engineering job. Yeah. And we've we've started to see that in our business uh that's focused on packaging.

SPEAKER_02 46:18

Um, so can you tell me some of the unique technologies that Nova has developed to address these challenges and maybe some examples?

Nordson T&I: Growth, AI Inspection, And Sensors

SPEAKER_04 46:26

Yeah, from kind of the highest end, we've talked a little bit about moving from the front end into packaging. We're seeing that drive as well. So we've taken our core OCD platform, and it has a couple of applications within packaging now, especially at the kind of the beginning end of the packaging process where it's really critical. Hybrid bonding is one of them, and there's a couple others where the metrology requirement there is actually as front-end as it gets. It's just happens to be at the beginning of packaging. And so the need is is really critical. And we're starting to uh we've sold many products into that category at this point. The the main sort of packaging business that we have, we have two. One is in the chemistry. So anytime you do plating, most of all of the metal lines that are deposited, even in the front end, but in pack the lines that connect stuff, are plated electrochemically. And the chemistry depletes as you plate, and you have to replenish that chemistry. We're involved in helping customers see when the to replenish and how the quality of the chemistry is as that uh sequence kind of goes forward. So it's it's really critical to the packaging process.

SPEAKER_02 47:36

So that's measuring materials. It's measuring like viscosity and things like that.

SPEAKER_04 47:40

It's think about it like composition. We're measuring the how much copper is in the in the bath, how much crap is in the bath. We're looking for all of this. Stuff that you don't want is in the bath. Uh byproducts we call them. So but it's you know, it's it's a very fundamental uh requirement. And then we have the product group that is in Germany that we acquired here this year that is measuring metrology aspects of all of those uh, you know, from silicon thinning to TSVs to even looking at some of the bump heights and things like that. We've got a we've got a handful of really critical measurements that they're making in the proper packaging process. And they and they actually can do handling of panels, they can handle reconstituted wafers, they can handle frame wafers. I had only seen a framed wafer for the first time when we were talking to these guys as we were acquiring them. I've never held one in my hand. It's crazy. It's it's actually a really thin wafer sitting on a membrane that's held in a frame, and it's like a drum head.

SPEAKER_03 48:43

Yeah, yeah.

SPEAKER_04 48:43

Yeah, it's amazing. I wouldn't hit it though. Yeah, you don't want to hit it. But you can imagine the handling of that is completely different than what you do in a photo. Yeah, right. And so, and all of that affects the measurement performance because how stable a substrate is when you're making the measurement is very critical to getting the result you want.

unknown 49:00

Okay.

SPEAKER_02 49:01

Well, I'm afraid we have to wrap up, but hopefully we can continue this conversation in about a month at Semicon Europa. And uh hope to see you there. And thanks for joining me today.

SPEAKER_04 49:11

Thank you.

SPEAKER_02 49:15

So, my next guest today is Carla Ferrana from Nordson Test and Inspection. They are one of our elite members. So welcome to the podcast. Thanks, Franz Fah. Good to see you. So, you know, we've planned a lot of podcasts together behind the scenes, and you always have your subject matter experts, but we've never had you on the podcast. So, can you just tell a little bit about your role there? You've been there a while.

SPEAKER_09 49:37

I've been there for, I guess, collectively around 11 years. Okay. So I worked for Cyber Optics prior to Nordson, and we were acquired back in uh November of 2022. So I lead global marketing for Nordson testin inspection.

SPEAKER_02 49:50

Okay, and here we are at Semicon West 2025, and it's the first time we've had it in Phoenix, but I see you every year. So you are a veteran of San Francisco. So how do the two compare so far?

SPEAKER_09 50:03

I am. I have to say I was absolutely delighted to see the energy or to see all of the people here. It's very busy. Um, we've had excellent customer meetings. I think it's one of the best semi-buss I've been at.

SPEAKER_02 50:15

Yeah, you know, I've been trying to spend more of my time on the floor to hang out with the members and and do the podcasting. And this year I didn't get a table because I've missed being on the floor. It didn't really do much for us to have a booth where we're chained to the booth, you know, for us. Right. I'd rather walk around and see everybody. Yeah, it makes sense. Does mean knocking down and you know, setting up the podcast equipment, but not a problem. So, but I have noticed that the activity, the energy, people are packing into the hall. And I I think it's because Arizona has the ecosystem.

SPEAKER_09 50:50

Yes, I would agree with you.

SPEAKER_02 50:52

And you don't have as many people. I mean, even in San Francisco, people have to come up from San Jose. I think Joe said yesterday on his keynote like Silicon Valley isn't where semiconductors are being manufactured anymore. It's really starting to shift, and Arizona is identifying itself as the semiconductor center of the United States.

SPEAKER_09 51:10

So and it surely makes sense to have semi-west here.

SPEAKER_02 51:13

Of course, you're in Minneapolis.

SPEAKER_09 51:15

Yes.

SPEAKER_02 51:15

So there's other regions where semiconductors, you just had the Heartland event last year, right?

SPEAKER_09 51:19

Right, right. And we have um Vidya Vijay with Nord's Intestine Inspection. She is involved with the semi-Midwest chapter. So she's very active as well.

SPEAKER_02 51:29

Yeah. So how are things at Nord's Intestine Inspection? Let's have an update.

SPEAKER_09 51:33

Yeah, thanks for asking. Um, so business is good, I guess. Stepping back, Nord's Incorporation overall, uh, Q3 versus Q3 of the prior year, we were up 12% in top line. Um, the advanced technology segment, which is um part of the Nordson corporation, was up 12%, and test and inspection is within that. So we're pleased with the results, and we think we're well poised for continued growth. Um we're very excited about the market. It's a great time to be in semiconductor. It really is. As you know very well, um, the the whole march to a trillion dollar market is exciting for everybody. It's going to happen.

SPEAKER_02 52:07

The jury is still out onto where it's going to happen, I think. I mean, in my opinion. But we know the growth is there, the drive is there, um, the need is there, right? But there's still some headwinds that we're facing as an industry, I guess. But sin, let me ask you this. So as a company that is global, yes. How are things like the tariffs and the um H1B visa uh fees that are being suggested? How do you think this is going to impact your business?

Comet: U.S. Footprint, Talent, And CT Launches

SPEAKER_09 52:38

Um, that's a great question. I mean, I guess I would say that as you mentioned, there's headwinds, there's tailwinds. Um, in terms of headwinds, we we do recognize that the tariffs and the visa situation, those create challenges for us. Um, you know, certainly the geopolitical issues and the macroeconomic issues are our headwinds, but we see a lot of tailwinds as well. Um in terms of you know, fab expansion. We're excited, I think 105 new fabs will come online before 2027. The advanced packaging market, as you know very well, um is a high growth market, and that's all driven by um AI. Right, right. Um and so that that whole push to AI is exciting, and we see a lot of growth opportunity based on that. And we develop, you know, within our inspection and metrology solutions, we have advanced our AI capabilities to um address the customer needs. So we're accelerating processes. Um at the end of the day, we're able to improve their yields, improve their productivity, improve their operational efficiencies, and improve tool uptime at the fabs. So we see all the high-level customer benefits are really helping to drive our business. Um, but we are certainly mindful of the challenges as well.

SPEAKER_02 53:49

So since you have customers all over the world and those 105 fabs are all over the world, there's a huge chunk of them here with TSMC, right? And then the Amcor facility, they just broke ground yesterday or two days ago. So there's an opportunity on the packaging side. Where are you seeing the orders coming from?

SPEAKER_09 54:09

Yeah. Good question. We're we are truly global. So we have a large business in Asia, North America, our our overall America's business as well as Europe. So we have a good distribution overall, we have a great global footprint. We um are pushing manufacturing in region for region as well as a key strategy from an operational standpoint. And within North Intestine inspection, I I see we have really three pillars. So operational excellence is a core competency for us, but we also consider ourselves um technology leaders in the marketplace. So we have developed and pioneered several of our technologies within the portfolio. We have um a number of really brilliant PhDs in math that do all of the algorithm work for us, which we call a secret sauce. We have a significant amount of IP that creates competitive barriers for us. Um, we have 2,500 patents within Nordson and 250 within Test and Inspection. So we believe we're really strong on the technology side and we continue to invest in RD to ensure that we're continuing to stay ahead of the customer needs. So technology leadership coupled with that whole operational excellence piece. And then the third is really the customer intimacy or customer centricity. So I believe we have um a leg up in terms of really truly understanding the customer applications and and having this more intimate relationship with them, understanding their their needs, um, and ensuring that we're developing and bringing and deploying inspection metrology systems that address those needs, not just today, but into the into the future, right?

SPEAKER_02 55:45

Right. So the in-region four region, that that means expansion. So you I'm I would imagine have some plans for expansion in different regions. And would that involve putting in like centers of excellence um where people can do tool demonstrations or actually manufacturing in different regions?

SPEAKER_09 56:02

Yeah, actually, um center of excellence. So we're looking at expanding in regions where it makes sense. We haven't made any formal announcements to that degree, but that's on the on the radar.

SPEAKER_02 56:12

With all of this growth, I know that Norton has a really robust like internship program. You know, so how are you making sure that you're gonna fill your talent pipeline?

SPEAKER_09 56:23

Yeah, so you mentioned the CDLP program. It's our commercial leadership development program where we've brought in you know recent graduates. Right. It's a wonderful opportunity for us to have them come on board with fresh eyes and and desire and that the whole um keen factor. So that's great. Um it gives them the opportunity to work in different divisions and get experience within Nordson and then ultimately ideally convert to a full-time employee and and grow their career with Nordson.

SPEAKER_02 56:54

So is this something that all of your employees go through, or is this just one program that people can apply to?

SPEAKER_09 56:59

That's just one program people can apply to.

SPEAKER_02 57:01

So newcomers to the industry or new students right out of school. Right, exactly.

SPEAKER_09 57:07

Yep. Our HR team recently launched a sales leadership development program. So that gives people the opportunity to jump right into sales, familiarize themselves with the products and the technologies, and really understand the customers and the landscape and such. Okay. Tell me what's new at North then. Yeah. So I guess we have solutions that fit, as you know, on the front end, semiconductor, mid-end, and back end. So in our mid-end, we have the spin SAM, and I know you've done a wonderful podcast with Brian about that, but it's a great example of that whole innovation factor where we've gone from this raster scanning to very similar to a record. Okay. Like where it's going in a circular motion, and that leads to this industry-leading throughput. So um it's a game changer technology, it's proprietary. We developed it, but it's really um addressing a need in the industry from an acoustic perspective. So um so we're very excited about that solution. And then also in the mid end, we have the XM8000, which is the automated X-ray metrology system. Okay. And that's for TSVs and uh wafer bumps and such. So that's also another technology where it's all about the image chain inside that's really powering that technology. Um that's uh yet another innovation. Um and then we have a plethora of offerings in the the backend semiconductor. Um as you know, the dynamic planar CT has been a game changer. It's two to three times faster than the the traditional CT for X-ray systems. And then in the front end, um we've just recently launched a centering sensor. So uh we have about seven modalities that we address, or seven applications. So leveling, vibration, relative humidity, gapping, and such. Um so we've extended the portfolio for these front-end sensors to address centering. So it's essentially centering wafers. Okay. Waifer transfer positions. Okay. Um so that's exciting because these devices, they just float through or travel through the semiconductor tools anywhere away for travels, and it's all for tool setup, tool maintenance, and increasing uptime.

SPEAKER_02 59:16

So you're actually doing inspection on the tool itself. So it's not necessarily is it like for calibration?

SPEAKER_09 59:22

It's not necessarily inspection. It's more um they gather real-time measurements. Okay. So it's a measurement size.

SPEAKER_02 59:29

But it's a measurement of the tool. Correct. Not the wafer, not the device. Yes, correct. We're actually doing metrology on the tools itself to make sure that the processes are running properly.

Siemens: Cutting Compute Power From Chip To Data Center

SPEAKER_09 59:40

Exactly, precisely. The portfolio is very unique, and again, we have a lot of um IP behind it. Essentially, for a process engineer or equipment engineer, it's all about saving a tremendous amount of time. So they could be leveling a tool or addressing vibration, then the tool comes down. That's so expensive, as you know, for hours having a tool down. That's a that's that's not good. So using the wafer sense, it goes right in under vacuum and it takes real-time analytics so that whatever they need to solve the issue can be done really quickly. Okay. So that's a a huge benefit. So those are deployed in all major fabs and um we also sell to equipment OMs, but another innovation, I guess we're proud of. Yeah.

SPEAKER_02 1:00:21

Well, I know that um I'm gonna see you in another month. Yes, in Product and Productronica, and we're gonna have a chat, and you're gonna be joining our Stamtish, which is gonna be my retirement party. So we're gonna have lots of fun. And congrats on your retirement. Thank you. I'm very excited. I mean, as much as I've loved all of this, it's gonna be nice to move on and do something different. You know, a new chapter. A new chapter. No pun intended. Well, you'll write a book.

SPEAKER_09 1:00:47

You'll surely be missed.

SPEAKER_02 1:00:49

Thanks. You never know where I might turn up, though. So I'm never I'm not gonna say goodbye out to Alfie Hussein.

SPEAKER_09 1:00:56

Right, very good.

SPEAKER_02 1:00:57

So we'll see you in a few weeks.

SPEAKER_09 1:00:58

Okay, thanks, Francois. Looking forward to it.

SPEAKER_02 1:01:05

I am speaking with Isabella Drolt of Comet, which is one of our elite members this year. And you might recognize it as Comet XLO, but things have changed. So welcome to the podcast and let's chat. Yeah, thank you very much for having me. So you've been on the podcast before many times. Can you tell me a little bit about your role at Comet and why we are now Comet and not Comet XLON?

SPEAKER_10 1:01:32

Well, there is this thing because uh Comet Yxon is actually our product brand for our X-ray systems. Okay, and uh within Comet we also have Dragonfly as our AI deep learning uh product brand. And together we are all in the division of IXS. And therefore, when you have like so many product brands in one division, you only talk about comet anymore because that's also what we are representing.

SPEAKER_02 1:01:58

Okay, so now we're talking about comet. And we are here at Semicon West in Phoenix in October versus San Francisco in July. So, how have it been for you?

SPEAKER_10 1:02:12

So I have to say I'm very much impressed. It's it's a great show, it's very well organized, a huge crowd. We did not expect so many visitors, and uh yeah, I have to say I'm a fan of Phoenix.

SPEAKER_02 1:02:26

I'm a fan too, because I live here, so I don't have to travel. Oh, that's perfect. Yeah, I get up in the morning, I get an Uber. I've got um Alberto, he's been picking me up every day. We now we just pay him directly and um have a little nice chat on the way in, and it's been lovely, and I actually do find it a nice change. I know we're going back to San Francisco next year, but it's also going to be in October, not July. So that will be interesting. Um but I won't be there because I'm retiring.

SPEAKER_11 1:02:53

Yeah, I heard. How sad? Oh, it's not sad for me. So how do you have any plans in? Yeah, this is your interview though, not my interview.

SPEAKER_02 1:03:03

I'm gonna write a novel. But uh, let's talk more about you. Comet has established a president in the US. Now, is that is not recent, is it?

SPEAKER_10 1:03:12

No, we have been already represented in the US for quite a while, especially with our uh division uh plasma control. Okay. So RF uh technology. And uh we have been there uh quite a fair amount of time.

SPEAKER_02 1:03:26

Because I've started seeing your booth at uh local events, like at the semi-west breakfast in Arizona in May, and then at IMAP Symposium in San Diego. Are you based? Is the office based in San Jose or Arizona?

SPEAKER_10 1:03:41

The office is based in uh in San Jose, but especially when it comes to our division in IXS, we are placed all around the US, and we also have someone here in Arizona who is doing great a great job connecting with people in the semiconductor industry and with semi.

SPEAKER_02 1:03:57

Okay, and your headquartered is in Switzerland. Exactly. Okay, so which leads me my you know to my next leading question, which is how have the tariffs and the visa fees that they're going to implement um impacted Comet when you're trying to conduct business in the United States?

SPEAKER_10 1:04:17

Well, so for us, actually, it has not affected us uh that much. We are producing out of Germany, and uh we are have a good relationship uh with the US, and uh we have uh a lot of US employees, so we have not been that much impacted by uh by that so far. But what we also have to say is the American market is a very important market for us, and we can see also the potential in it. And well, of course, we have to obey all regulations, and that's what we're also gonna do. But we see that the US is really going to pick up in the in the next couple of years uh when it comes to semiconductor and also the potential we see when it when it comes to X reinspection systems.

SPEAKER_02 1:05:00

You're not concerned that all of these things are going to impact the US's actual ability for growth? I mean, for Comet, it's fine, right? You're not experiencing anything. But what about I mean, I'm hearing a lot of other companies struggling with these very things that, in my opinion, I think it's going to drive the growth in other regions of the world versus the US. Even though we've got all of these fabs going in and construction happening, but we still have to have the talent.

SPEAKER_10 1:05:32

So I think we have to see it in two ways. So one thing is, as you mentioned, uh for comment. I think how it's geopolitically really gonna play out. We know when we know in the future. Uh of course, I I'm with you that talent is a very important and critical point. And to get the right talent and the good talent, and that's something where I think we uh especially the US needs to find also a solution, like Europe has also a huge talent gap and getting people attracted to the semiconductor industry. Right. So I think it's like in a two-way street. Yes, there are uh geopolitical tensions, and we see that all around the world, and that's something that needs to be monitored and looked at. But we also have this thing that a lot of people do not even know about the semiconductor industry. We see a lot of investments here in Arizona, but how do you join the industry if you don't know it?

SPEAKER_02 1:06:23

No, right. And you know, one of the things I've been thinking about on the workforce development side and talking to people is that you know, Intel had some layoffs earlier this year. And those people are looking for jobs. So how about recruiting uh people who have been in the industry? There seems to be a lot of effort going into raising the next generation of semiconductor engineers. How about um experienced engineers that are now out looking for work?

Foundry Efficiency And Fab Optimization

SPEAKER_10 1:06:50

So we have uh also uh work for spoof here at semiconvest in Arizona, and we are looking for all kinds of uh engineer types, and yes, we are also actually if you can hire somebody from your former or from your customer, that is like a win-win for us because then you get a better understanding and uh have somebody who has been on the other side and can give you the requirements that they need.

SPEAKER_02 1:07:13

So okay, so if you're out there in Arizona looking for a job, Comet is hiring. Exactly. Okay, well that's a good message. Last year at Semicon Europa, we were at the launch of the CA20. And now I understand that you have had another launch related to that tool in Semicon Taiwan.

SPEAKER_10 1:07:33

Yes, and we also launched uh a CT system, especially for the semiconductor industry. So now we have a portfolio that covers root cause analysis when it comes to 3D IC, chip-on wafer inspection, also TSV inspection. But also uh we improved the CA20 when it comes to frucal time, when it comes to dose uh management, uh to protect the memory, because that's a big concern in the industry. Uh and uh that is what we've launched at Semicon Taiwan. And actually, we had quite some good feedback here in the US.

SPEAKER_02 1:08:06

So, how has it been having this stack up? Did you do Semicon India? No, we did not. Okay, so you had Semicon Taiwan. Yeah. And now here a month later, you have Semicon West. So are you hearing, like getting feedback from Taiwan now when you meet customers so soon after an in-person meeting?

SPEAKER_10 1:08:25

It's very funny because uh we introduced it at Taiwan, and actually, customers from Taiwan were just at our booth today here. There you go. And uh wanted to talk about the C A 20. So it's the semiconductor circus is doing a good job here.

SPEAKER_02 1:08:40

So four weeks from next week, so five weeks away, we'll be at Semicon Europa. Can you give us a preview of what you're gonna be showcasing there, or is there something new coming? Or well, we are fast, but we're not that fast. Well, I know that you paste these things out because you want to have something to introduce. What can we look forward to there? We can talk about it there, but what can we look forward to seeing there?

SPEAKER_10 1:09:02

So semicon Europa, you will see the C20 in person, as you I want to put it. And uh, we also will talk about the FF35, our CT system, especially when it comes to root cause analysis and failure analysis. And uh yeah, we are looking forward to having everyone uh at our booth to join us.

SPEAKER_02 1:09:22

Okay. Well, I will be there on my final stop in the retirement tour. Looking forward to seeing you. So we'll see you then. Thanks. So my next guest today is Michael Muncie from Siemens DISW, which is digital industry software. Um and last week we spoke with Keith Felton of Siemens DISW also, and he recommended that I talk to Michael this week because he gave a talk at Semicon West called Challenging the Status Quo from Chip Innovation to Compute Optimization. Welcome to the podcast, Michael.

SPEAKER_06 1:10:02

Thank you so much.

SPEAKER_02 1:10:03

So before we dive in, can you just tell me a little bit about your role at Siemens?

SPEAKER_06 1:10:07

Sure. So I handle semiconductor strategy in digital industry software, which means not only the software used to design the chips, the EDA part, but also the PLM software, the um 3D CAT software, the multi-physics simulation. So it's looking across our whole portfolio of products and coming up with a software strategy that handles all of our products in the portfolio.

SPEAKER_02 1:10:31

And are we talking not just chip design, but also into the package space?

SPEAKER_06 1:10:34

Yeah, into the package space also and also into the product space as well, because the chips eventually go on a board which go into a product.

SPEAKER_02 1:10:41

Right. Okay, so in your title, From Chip Innovation to Compute Optimization, that's basically the end-to-end of the chip manufacturing cycle.

SPEAKER_06 1:10:51

Exactly, yes.

SPEAKER_02 1:10:52

Awesome. What are some of the challenges this industry is facing right now with regard to that?

SPEAKER_06 1:10:56

Yeah, so you you know, I started off my talk, you know, looking at the amount of power it takes right now to make a chip run. And something happened on Monday that kind of made me think twice. I read the there's a press release about open AI and AMD.

SPEAKER_02 1:11:12

Oh yeah, okay.

SPEAKER_06 1:11:13

And for the first time ever, they didn't talk about MIPS or throughput or flops or number of cores. They said that AMD was gonna supply six gigawatts worth of compute power to open AI. And I'm going, that's the first time I ever heard compute power in terms of the amount of energy it's gonna take to run the chips.

SPEAKER_02 1:11:36

So they're classifying the chips in terms of how much power they need?

SPEAKER_06 1:11:40

Exactly. Right. So they're saying that the total compute power that they're going to offer is equivalent to six gigawatts of electricity demand over the course of the contract.

SPEAKER_02 1:11:52

Why would they term it that way?

Onto Innovation: PACE Center And Packaging Reality

SPEAKER_06 1:11:53

Aaron Powell Well, I think now because you know, we're past the point when we talk about you know traditional ways of measuring, you know, performance, right? Right. That now it's really how much electricity it takes to run a data center. Right, right. So now and so that has a lot of ramifications. You know, six gigawatts is you know enough to run a small town. Right. And for how long? For you know, I think the the rough it's about two gigawatts per, you know, couple thousand people to run their houses and everything.

SPEAKER_02 1:12:27

On a daily basis.

SPEAKER_06 1:12:28

On a daily basis, right? The ramifications are you know, you talk to a colleague of mine that looks after the energy uh industry, and he's talking to customers right now about building micro or mini nuclear reactors that could be strapped on the sides of these buildings just to provide the power. Data centers right now consume about four percent of the entire electricity generated in the in the United States. By 2030, that's gonna double to 8%, maybe go as high as 10%. Now, we've been on a slow ramp, and now that ramp is starting to increase. So if it's gonna double within the next five years and it doubles after that, we're gonna get to a point where it's gonna be very difficult to produce enough electricity to drive these compute farms.

SPEAKER_02 1:13:13

Right. So it's not about finding new ways to power at the compute farms, it's actually finding ways to reduce the power needed in the compute farms itself.

SPEAKER_06 1:13:23

Absolutely, right? And the thing is chips don't you can't look at them just as a chip anymore. You have to look at them within their ecosystem. So there's chips and uh data centers, there's chips that go into automobiles, there's chips that will go into laptops, right? And you so you can't generically think about what am I going to do to reduce the power. You gotta think about what is the application that the chip's being used for, what software is running on it, and then start to think about how you optimize things in terms of the actual use model of that chip and its environment.

SPEAKER_02 1:14:00

So this has been kind of a hot button for me for a while, ever since I found out how much energy it takes to do an AI, generative AI query and concern about frivolous use.

SPEAKER_06 1:14:10

Yes.

SPEAKER_02 1:14:11

And I had this conversation last week with Subu Ayer, and he says it's really not about controlling how people use it, it's about reducing the power requirements of the chip. So the OpenAI press release talked about like the amount that AMD was saying, is that a good amount? Is it show that it needs less power than other like what was it comparing it to?

SPEAKER_06 1:14:31

Well, so that's a very good question because it wasn't comparing it to anything else, and it actually makes it difficult for other people to actually know what is AMD's power consumption compared to NVIDIA's power consumption, right? And and then compared to the chips that Google are producing for themselves or Amazon are producing themselves, you know, it almost obfuscates a comparison now that you start using metrics like that, because six gigawatts of power, but doing what?

SPEAKER_02 1:15:00

Right. So what's the compute power of that? That's the that's the consumption.

SPEAKER_06 1:15:04

And that's the actual consumption. And and and that's tied back to the actual use model, right? So that chip is still using a certain amount of power, right? But um, you know, when when I But what's the performance? But what's the performance of the chip, right? And exactly, and and how much is it actually doing out there? You know, what's the query? My favorite metric was back about 10, 15 years ago, they said a single search in Google uses uh about six kilojoules of energy, which is the amount of energy stored in a double A battery. So every time you ask the Google question, you burn it a double A battery. And and that made it like Right. Yeah.

SPEAKER_02 1:15:37

It's for the lay person can understand this so that we know.

SPEAKER_06 1:15:40

Right. But now, you know, six gigawatts, most people the only thing they remember is not per query. Not per query, but in terms of, you know, you mention gigawatt, and you know, certain people go, oh yeah, back to the future, right?

SPEAKER_02 1:15:52

You know, you know right, exactly. The flux capacitor.

SPEAKER_06 1:15:55

And all we need is Mr. Fusion, and it solves all our problems. So yeah, yeah.

SPEAKER_02 1:16:00

Yeah, I don't even know how many kids these days watch. I mean, I made my kids watch Back to the Future. They loved it. They're like, Oh mom, we love watching old movies with you.

SPEAKER_06 1:16:07

Right. Yeah. But but when I talk about in context, when I talk about data centers, I said, you know, there's a there's an ecosystem, and you go the chip to the package, to the board, to the blade, to the rack, to the data center, right? So you have a chip, right, and and we could quantify how much power that chip uses, and we could do it generically. But ultimately, what's gonna happen is that chip is gonna ultimately be running certain types of software. So we can actually model the software that's gonna be running in the data center to do a much better job of predicting what the power profile of the chip's gonna look like, because it's gonna be very different if that chip was used in some other application. And then by doing that, then we could do a better job of how what it's gonna take to cool cool that chip in the actual blade, and then you stack the blades in the rack, and then you look at the whole data center, right? So now the optimizations start at the chip level, but now go into what are the cooling requirements of the data center, which means what's the requirements of the HVAC systems and can we optimize those? What's the electrical distribution over time of providing power to that data center look like? And can we optimize that? And that's why it's very important to understand if we're gonna start to reduce the entire power consumption, it's really the holistic field we need to look at. Yes, we could do everything possible at the chip level, but that's the starting point. And now we have to go follow it through its ecosystem and see how we can optimize at each step of the process.

SPEAKER_02 1:17:36

So is this what your paper was about or your presentation was about?

SPEAKER_06 1:17:39

This was absolutely what part, you know, is part of my presentation. And then then I took it to the next step, right?

SPEAKER_02 1:17:43

Right, okay.

SPEAKER_06 1:17:44

Because you have to build these chips also, and you have to look at the foundries. And the foundries consume a lot of power and resources, so water and chemicals. And there used to be an adage of, you know, if it worked right the last time, we'll build the next data center the same way. And you have this cookie cutter approach, but that's not the most optimal way of doing it. And if you look at the macro level, for example, one of the most energy consumption parts of the whole process is what's known as the water chillers that provide the chilled water to the fab. And a lot of people think, you know, it's just you know part of the HVAC system, it's just part of the facilities of the building. And they turn them on and they let them run, and like anything that gets older over time, it becomes less inefficient. And they don't spend a lot of time monitoring it. So we've worked with a lot of these companies building fabs, looking at these systems like this. And we had one one company that came to us and we save them roughly$700,000 a year in electricity usage. So by saving them$700,000 a year in electricity usage, we're able to have them now put that money towards other upgrades they could do, and they could start making their older fabs run a lot more efficiently. So now when they build a new fab, we could take a lot of this knowledge and start teaching them how to be much more efficient about the way they build the fabs, have them measure a lot more things. Ultimately, we want to get to the point where if you're a company that's handing off a certain type of chip, right? So you're A D and you're fabbing a one of these processors for AI, we could predict exactly what the power profile of manufacturing those chips are going to look like and help you optimize the fab to also reduce the total amount of electricity.

SPEAKER_02 1:19:34

So this is what the work that Siemens is doing.

Tariffs, Supply Chains, And What’s Next

SPEAKER_06 1:19:36

Exactly, right? So we're taking the holistic view. So that's why it's breaking the existing paradigm. Because like anything, silos have developed over time. You have the design silo and you have the manufacturing silo and you have the packaging silo, right? They all have to come together and they all have to work together. Yes. And that's what we're doing is that we're looking at the entire semiconductor ecosystem from design through manufacturing into packaging and figuring out how we optimize the entire process. So you're producing the most energy efficient chips, you're producing them in the most energy efficient way, and you're packaging in such a way that they could be cooled most effectively, right? And that's I love this.

SPEAKER_02 1:20:16

Yeah, what a great concept.

SPEAKER_06 1:20:18

Yes.

SPEAKER_02 1:20:18

I'm so glad we had this conversation.

SPEAKER_06 1:20:20

Yes, me too.

SPEAKER_02 1:20:21

So it's nice. I like I am really been concerned about this, so I'm happy to see that Siemens is taking such a stand in doing something about it from their perspective or from the software side of things.

SPEAKER_06 1:20:32

Yes, absolutely.

SPEAKER_02 1:20:33

Well, thank you so much for joining me today. I enjoyed the conversation. Where can people go to learn more?

SPEAKER_06 1:20:38

So you could go to the Siemens website and go to the semiconductor section under the industries, and we will have the full range of information, like I said, all the way from design through manufacturing into packaging there.

SPEAKER_02 1:20:50

Awesome. Okay, thank you, Michael.

SPEAKER_06 1:20:52

Great, thank you.

SPEAKER_02 1:20:56

So here we are on the last day of Semicon West here in Phoenix, Arizona. And I've just sat down with Tim Kriman of Onto Innovation. The last time Tim and I talked was during COVID when I was doing video interviews. So um I'm really happy to have you here again this year.

SPEAKER_05 1:21:16

Yeah, it's great to speak with you. A lot going on this year.

SPEAKER_02 1:21:18

You know, okay, so here we are in Phoenix. Now we're through the show, three days. How does it compare to San Francisco for you?

SPEAKER_05 1:21:27

I have to say, overall, I mean, I've been coming to Semi-West now for 15 plus years. Um, I I thought they did a fantastic job. The venue is great.

SPEAKER_11 1:21:37

Yeah, I agree.

SPEAKER_05 1:21:38

The local community, restaurants, the ability with large spaces to be able to network with people, different events going on. I thought they'd it was fantastic.

SPEAKER_02 1:21:48

I completely agree. I'm kind of sad that it's going back to San Francisco next year.

SPEAKER_05 1:21:52

But I really think that it's easier here.

SPEAKER_02 1:21:55

Yes.

SPEAKER_05 1:21:56

Everything is very well, you know, positioned in terms of location to be able to get in and out, um, get to the different events. Um the venue is wonderful. Yeah.

SPEAKER_02 1:22:07

Um you know, once I got used to where everything was, because I've been running back and forth to the press room and to the floor. And it's nice that most everything is in one floor, and then there's that 700s. I wouldn't want to be up there. I don't know how much traffic they're getting up there. But anyhow, what is your role at Onto Innovation?

SPEAKER_05 1:22:26

Senior director of product marketing. Okay. So yeah, I'm on the marketing team and and our role on the product marketing team at Ontu is really identifying the landscape. We're directly interacting with with customers, um, but not so much on a on a from a sales perspective, but more so to understand their product roadmaps, sort of the inflections that are going to be occurring looking out. And then we line that up with our strategic marketing and our product managers that are actually driving the products to get real alignment with what the customers are doing, and then also have a vision of where things are going so we can stay ahead of the curve.

SPEAKER_02 1:23:03

Okay. And so your core competencies are in a bunch of different tool sets. You have lithography, you have metrology.

SPEAKER_05 1:23:13

Yeah. So we have um advanced packaging lithography for wafer and panel, inspection, 2D, 3D inspection, um, and also uh 3D bump metrology. Um we have um high-speed IR inspection. Uh we have OCD thin films. We also get into the materials side of the business as well.

SPEAKER_02 1:23:33

Okay. And so the last time I actually spoke with anybody from Ontu, I was at the Entu Innovation Facility in Wilmington Mouse, and you were just launching the Pace Center, the packaging center of X Men. That's correct. So we're one year in. Yes. How has it been going?

SPEAKER_05 1:23:48

It's been going very well. The opportunity to bring in folks within the panel supply chain on the materials side, on the equipment side. We've had uh partners that have placed equipment within the facility. So it's not just an on-to innovation.

SPEAKER_02 1:24:03

Because when I was there last year, it was only on two tools.

Season Preview And Closing

SPEAKER_05 1:24:06

No. So we have brought other tools into the facility at this point in time. So it really gives us a chance to work with equipment makers, material suppliers within the supply chain. They can come in, they can use our equipment, some of the equipment that our partners have brought in to really test some of the new um packaging techniques. Right. Um, we've worked with CCL, obviously, the move to glass, TGV. So it's really been a good opportunity to do that. And then, as well with customers, the lab offers our customers an extension to their RD team. They can send us material, we can run that material with them and then send them the results.

SPEAKER_02 1:24:45

Okay. And how many companies do you have partnering with you at this point?

SPEAKER_05 1:24:49

That's a good question. I think it's north of six or seven at this point.

SPEAKER_02 1:24:53

All right. And now all of this did not tap into any CHIPS Act funding.

SPEAKER_05 1:24:58

No, it did not.

SPEAKER_02 1:24:59

You know, what was the decision around that?

SPEAKER_05 1:25:01

I think that, you know, at the time, the concept first pace was prior to the CHIPS Act. Okay. And our CEO, Mike Blazinski, really felt as though it was an investment that we needed to make to show uh our customers we're invested in their success and give them an opportunity to work with us directly. And as I said, to be an extension of their of uh of their RD team.

SPEAKER_02 1:25:26

Okay. You're probably able to move a lot quicker than you would have if you were waiting for chips funding.

SPEAKER_05 1:25:33

Definitely. There's no question. There's no question. That likely played into it as well because we obviously had, as all did, had the opportunity to try and make the case why an investment would be worthwhile. Um, but it was unnecessary.

SPEAKER_02 1:25:47

So I feel like the companies who can go it alone and not have to rely on the CHIPS Act funding or government funding, uh in the end of the day, are probably gonna be better off.

SPEAKER_05 1:25:59

No strings attached. Exactly. No strings attached. And I think right now, if you look at the climate, there was money that was assigned, money that was given, money that was not given out, right? And then there was clawback.

SPEAKER_11 1:26:10

Right.

SPEAKER_05 1:26:10

So at this point in time, I think it's a wait and see to how this is going to play out. There are definitely investments that are being made into US semiconductor manufacturing. The question is, how does that play out long term on the customer side of it and from a political perspective? How are those things all going to play out? Ideally, I think my perspective is a lot of the companies are taking the wait and see attitude.

SPEAKER_02 1:26:37

I went to the market symposium on Monday and it was an interesting collection of presentations. It seemed like they were telling the audience how they were going to benefit from the tariffs. And then there was another presentation that was telling them how they were gonna benefit from the big beautiful bill. And then there was one that was telling on how to navigate for foreign companies to do business in the United States. And but I did see a lot of references to investments gonna happen in 2028. We're gonna see 40 billion in 2028. And I put 2028 together with Wade and C, and I think we're gonna be in a different administration in 2028. My take is that the industry itself is on a growth trajectory.

SPEAKER_05 1:27:26

No question.

SPEAKER_02 1:27:27

I'm not convinced, even a little bit, that the UN that United States is going to be a leader if we keep going the way we're going in that industry.

SPEAKER_05 1:27:38

I think that's an accurate statement. And I also believe it that's part of why you're in some of the sessions here this week, those are some of the messages that you're hearing. You know, there's a big push for onshoring at this point in time. And clearly, if you look at the global situation, there's a huge need, and I believe there would be a huge advantage if we could successfully do that. Right. I don't know that that's something that the US government can effectively facilitate by creating an environment where the investment comes. Right. Right. And the companies are allowed. But ultimately, the companies that we have in the United States that are manufacturing, they're gonna have to stand on their feet and get to a point where we can compete at the bleeding edge again.

SPEAKER_02 1:28:24

You know, one of the things Jan said, she was very straightforward, Jan Vardiman. Yes, which is always wonderful. And she doesn't pull any punches or mince any words, and she talked about how the advanced packaging sector is still heavily in Asia. The ad the they've invested in the infrastructure in the ecosystem, they invested in the technology, all of the most advanced stuff happens in Asia. And here we don't really have the substrates, we don't have the um the advanced packaging at the large scale, high volume, even though MCOR just broke ground, it's gonna be a few years before that facility is up and running. And one of the points Jan made that I thought was genius is that it's always going to be less expensive to send these offshore to be done. And it's more expensive to do it here. And unless companies like NVIDIA and AMD are willing to spend more for their packaging, we can build as much as we want. It's not gonna happen here.

SPEAKER_05 1:29:23

Right. Yeah, I I don't quite see how we win the cost structure side of packaging.

SPEAKER_02 1:29:30

I think we have to accept the fact that if we want to do it here on shore and not be making chips here and then sending them offshore to be packaged, the customers have to be willing to pay more for their packaging. Yes. Yes. Otherwise they're just gonna keep sending it offshore, right?

SPEAKER_05 1:29:45

Sure, sure. And completely logical. I think what it may come down to is offshore will be leveraged to the extent that it is not impeded by global issues. Right. But I think the concern right now is not having any capacity whatsoever to be able to support our own national security or have a path. So between the US and Europe and some of our other partners, I think that's where there needs to be a collective plan. How do we manage a potential negative situation in Taiwan? Right. Right. Right. Um, so but it's very hard to see how packaging takes off in the United States from a cost standpoint. Right.

SPEAKER_02 1:30:33

Now, Onto is primarily or essentially a tool manufacturer.

SPEAKER_05 1:30:37

We are tools manufacturing.

SPEAKER_02 1:30:38

And you work on, you know, you develop processes as part of that. But your products that you sell are your tools. Yes. Right. So you can sell them in the United States, you can sell them in Asia. You can are you doing the you know manufacturing locally so that so currently our primary manufacturing is in the US.

SPEAKER_05 1:30:58

Um, but there is an initiative in multiple locations um across um Asia to have manufacturing capability to be able to balance this out. Um obviously with the tariffs and how that will work out, the supply chains impacted. So it's we're looking to be able to um hedge our bets, so to say.

SPEAKER_02 1:31:18

Yeah, yeah. So how would the tariffs impact on your exports, exporting tools?

SPEAKER_05 1:31:27

Yes, it could it could be on our exports, it could obviously in the supply chain, depending on where components are coming for. Um we're fortunate a large portion does come from the US. Yeah. But again, you have to look across how this plays out. And, you know, we've seen the um tariff landscape change multiple times over the last six, seven, eight months. So what is it finally gonna look like? What do we need to do? How do we prepare ourselves? So I think our company is taking the position, listen, we've identified these potential outcomes, and we're gonna put mitigation in place to be able to make sure that onto can support our customers and be successful for investors at the same time. Okay.

SPEAKER_02 1:32:07

Well, I want to thank you for your time today and your candor.

unknown 1:32:10

Yes.

SPEAKER_02 1:32:11

And I wish Ontu continued success.

SPEAKER_05 1:32:13

Thank you so much. Same to you.

SPEAKER_02 1:32:15

Even though I'm retiring, you know, I'll probably be around because this is family.

SPEAKER_05 1:32:19

Well, we hope so.

SPEAKER_02 1:32:19

So we'll see you.

SPEAKER_05 1:32:20

Okay, thanks so much.

SPEAKER_02 1:32:25

It's hard to believe that we're in the home stretch for this season of the 3D Insights podcast. Coming up, we'll be talking with some contributors to a new book, Empowering Women in STEM, Pioneering Pasts to Shape the Future, and we'll be talking about their STEM journeys. And then we head to Munich for Semicon Europa, which is celebrating its 50th anniversary. We'll be reminiscing with Semi-Europe's Leith Altamime, Luc Vandenhoe of IMEC, and Christian Koitsch of ESMC about Semicon Europa past, present, and future. There's lots more to come, so tune in next time to the 3D Insights podcast. The 3D Insights Podcast is a production of 3D Insights LLC.