 Good afternoon data science community and a welcome back to Stanford University. We're here with the cubes live coverage all day long at women data science Worldwide annual summit. My name is Savannah Peterson. Delighted to be bringing you a power pack to line up all day. My next two guests are from ASML Bing and Chen. Thank you so much for being on the show. Welcome. Thank you. Thank you. It's very nice to have us here. Yes. Are you enjoying yourself so far at the show? Of course. Yeah, it's it's you mentioned you've been here before the pandemic. Now things are scaling. Does it feel like there's more and more women in data science? Yeah, I'm so glad to see the community is growing so fast. Back to a few years ago, I'm one of the maybe 200 ish ambassadors. And now we know that there are over 1000 ambassadors. Oh my goodness, 1000 ambassadors, 150,000 people in the community. Super impressive. Really glad to have you here. You both have very interesting jobs. And and ASML is more than just a semiconductor company, more than just a silicon company. So I am going to have you tell us a little bit about those. Bing, let's start with you. Okay. So in ASML that we make the most complex machines to print chips. So for most people, we think that, you know, most like just hardware. But actually, we also have a lot of software engineers to work behind this. And so for me, specifically, I'm working on the image processing part of of the production line. So we do a detect. We do sorry inspection and metrology. So for inspection, basically means that after we get the chips that we want to. Find the anomalies on the chips so that we can quantify this can quantify this chip. Sorry. No, you're doing great. It's essentially quality assurance. It's QA for photolithography and the actual making sure that everything that goes into these machines is going to do what we expect it to do. Exactly. And for metrology also, so they are basically taking measurements of the printed patterns. So if the measurements are within the given range of the specifications, we know that these patterns are printed well. They will do the chip actually will work as we designed. But if without if it is out of the range, obviously we need to rework this chip. So for both inspection and metrology, actually we are doing the quality assurance. Yes. It's going to be a tough job. I mean, we're talking about things on the micron layer too. We're talking about very, very tiny nuance and detail here. Exactly. So we are at the resolution of nanometers. Amazing. Yes. Wow. That's super cool. And Chen, you also have a very exciting job. Yeah. So I work with being in the office. So my job is slightly different. I work on the modeling side inside computational lithography. So you know that our machine keeps pushing the physics boundary and engineering boundary in every generation of our new machines. So the computational simulation part is really important. So what I'm working on is that I'm trying to build highly accurate model to a sub-nanometer level to predict how the optical, physical chemical and how the physical chemical process is happening on the wafer at this highly accuracy. How do we do that? It's actually re-leveraging a large amount of, I should say, exponentially growing amount of data actually through metrology and inspection to build highly accurate models. And this large amount of data actually enables us to unleash the power of machine learning or even deep learning and our production lines. You would have to. Otherwise, you couldn't process all that data to be able to... Yeah. I mean, we can't see what you're doing. You know, it's such a detailed level. It's so exciting. So I'm curious. This whole show is a real intersection of academia, of industry, of curiosity even. There's folks here just kind of coming to check things out. What are some of the career paths in academia for data science? I'll start with you, Chen, and then I'll bring it to you. Yeah. Interestingly, I have both a career in academia and I shifted to industry. Perfect person to ask. So I'm a physics major in undergrad and I took my physics PhD in atomic molecular and optical physics. And then I did one year postdoc in the Lawrence Berkeley lab just up north, 60 miles and so on. So after that, I realized that I really want my research, my work, and my skill sets to be applicable to contribute to real world impacts. So that a few friends referred me to SML. After I entered SML my job in a research division in SML and now in production engineering division in SML, I really find that SML is the best place because it combines all of my academia background postdoc experiences as well as my programming skills, physics skills and so on. Combines everything into contributing to the most challenging machine in the world. At the meantime, I met many brilliant colleagues from multidisciplinary, for example, being an electrical engineer background and many of my colleagues are physics, chemistry, mathematics and mechanical engineer and so on. So it's really, I should say that people who are having academia, doing studies, researches in academia, they will have a very huge amount of opportunity in industry. Nowadays, pretty much every industry requires multidisciplinary contributions from all the backgrounds. I would say that if you're willing to enter a new interest in industry, willing to learn, there will be tons of opportunities from my experiences. It's definitely been a theme of the show, is this intersectionality, multiple disciplinary, cross-team collaboration. It sounds like you were a dream fit for ASML and like you two are dream teammates in general. What sort of paths have you noticed, Bing? My background is a little bit different from Chen's. Even at the beginning of my undergraduate study, I started majoring in image processing. So even starting from then, I have started dealing with massive data. Later on, I also got my PhD in image processing and statistical data analysis. Just a little bit of education sitting at this table with me right now. I'm feeling uneducated by a mile and I couldn't be more proud. It's great. I love it. It's an honor. And also after I finished my degrees, I worked in the university for quite some years. At that time, I actually was doing clinical imaging studies. So I was doing cancer treatment or tracking for the medical school in University of Michigan. Through all these years in the academic, I'm vigorously trained with all the theoretical background. Also, in academia, most of the time, we are trying to develop a novel algorithm and models to get a specific metric optimized. That's the goal, to get novel technologies. So that needs a lot of theoretical background. And that's very useful when I switched my career from academia to industry. Because I say in a company, most of the time, we are in contact with real world problems. So the data are messier, let's say messier. And the problems are usually at the beginning, less well defined as in academic background. That's a really good point you just made. You're saying that when you're in academia, there's a lot more research that's going to be led up to that moment or things coming together to intersect. So it's almost more scripted versus when you're in an industry business solution. You're kind of just pulling together parts off the shelf and trying to make it up as you go and build what you need to build. So with the data that you have to make the decisions how to preprocess this data, how to clean data and how to formulate your problem to a model or to a real well defined problem. So with the training and all the experiences you get in academia, that will have you great to start and to get the formulation of the problem in the most efficient time. So that's the advantage that we had transferring from academia to industry. Well, I think that's awesome. There's clearly an overlap between academia and industry. I think also we're at a really interesting time just in the state of AI, ML, a lot of what's happening right now. We're going to have new folks entering the space. What you both just said is your background and whatever your background is, paired with some education in the space means anything's possible. You could be working on really interesting projects and that background, that interdisciplinary background could be really compelling. Are there any very specific, and you've both kind of touched on this a little bit but I'm curious if there's some that come to mind. Are there any particular challenges in the semiconductor industry that are uniquely solved by data science or technology today that maybe would have been significantly harder to do in the past before we were at where we're at? Ah, so what you mean is that with data science technologies from data science are inspired by data science methods, we can solve some problem better. Yes, and specifically in semiconductors. Yeah, definitely. So nowadays, many of our data are coming from as metrology inspection and many of our data are coming from the machine-locked data. So those data are massive. So before maybe even five years ago, we can only do down-sampling of those data to extract very limited amount of information to build positive empirical models with a few parameters. So that we are, I should say, back to that time, we wasted a lot of data simply due to the computational power not sufficient or how the modeling model is too complicated to finish the simulation we see in a production needed time-free. So that means we are limited back to that time. Nowadays, we can leverage distributed computing. We can leverage GPU technologies and so on, these hardware computing powers as well as many data science methods which were not well-known before. However, now it's getting much more popular. For example, cluster technologies, image processing technologies, video processing, audio processing, and so on. So all these, I should say, data science make all these technology much easier, accessible, even to industrial applications and for both ourselves and our customers to be more open to try these new technologies and generate more values for them and for ourselves, yeah. So pretty much every aspect of it to send a degree. Yeah, for us, let's say, before we do metrology and inspection that the WIFO will be scanned using an electron beam. Yeah. And okay, I talked about that these days all the resolution is the nanometer scale, so the images are very, very targeted and imagine with the electron beams that the noise of the images, they are much higher than what we deserve to be. Yeah. So the first step is that we need to actually pre-process these images to have better quality and now with current data science technologies we can use all the data that we have like Chen mentioned, we don't have to reduce a lot of the data or done sampling. We can use all of them so we can get the best benefit of it. Yeah. Yeah. It's all about making the world better and making our jobs easier. Yeah, let me add one more thing. Please. By all means. I'm here to learn. So I have a very recent example in my work although I couldn't disclose the details but I use one example. So maybe five years ago our computational capability can only afford a few hundred training examples and now we can easily afford up to million training examples. That means our model's competency back that time should be limited, right? Yeah. But now we can leverage deep learning models and our productions because the growing computational power and growing data amount. I'm really glad you brought that up because what we're able to do now before even four years ago, ten years ago is orders and orders of magnitude more in terms of volume and speed. So I think that's a really good point. You obviously know that sitting at the epicenter in a semiconductor company processing power is kind of that focal point there but I'm really glad you brought that up. We're here celebrating International Women's Day. We are all passionate women in our space. You both do a little bit of work there, don't you, for the working group? So tell me a little bit about how you're empowering women in our field. Chen, I'll start with you. Yeah. So both of us are inside a community called SML Women in Silicon Valley and it's part of the SML Women Bigger Community. So we established this community back to the year 2015 and I'm the third chairman and being worked with very closely contributing to the community. So the community's mission is to build a supportive environment for people to learn, share and lead. And these WIS, the WIS organization is definitely one of our channels that we believe resources from WIS can be accessible to our members for their career growth, for their technology, for the pathways and career policies and so on, improve all of them. That's why we come here and we not only introduce ourselves but meet the community who are either academia or other industries to join the bigger community and use that as our members' growing resources. I love that. And you obviously get excited working with her on this? Exactly. And also like Chen mentioned that we came here and this is my first time at WIS and I'm so excited to okay, this morning listen to all the talks and meet all the other peers, let's say. And this community I think that it will do huge contributions to promote women's in data science and we'll see, yeah, we're looking forward to all the brighter futures. I was just going to say the future is bright. One more thing I want to add is on the inside there are many, many different departments. So mechanical engineering, electrical engineering, software engineering and so on. Many of our colleagues does not even know how their work already involves a lot of data science elements. So that's why we want to bring the knowledge, observations, learning from this conference to our community members, our colleagues to let them know, look, your work is how they're leveraging data science already. How about joining this community to learn more? Yeah. Oh, I love that. Well, it sounds like everyone's invited into your community. On that note, a couple of last questions to close us out. For women who are just about to embark on their journey or looking at you, thinking how cool your job sounds, what's your advice? Bing, I'll start with you. What's your advice to a young woman in data science right now? Yeah. Or are you thinking about it? Yeah, I think that, oh, okay, the first thing is that when a young lady is trying to get into this field of data science, I'm sure that she has the passion about it. So with passion, that's the insight driving for you to get to learn new skills to get more, to broaden your view. And so I'm thinking at the beginning if you haven't been into this field yet, so just to explore your area, you know that you can talk to all the women. I want to specifically say women, all the data science that you, scientists that you know, and talk to know how their path was, how their professional path was. So that will give you some insights that you can reach your goal maybe faster. Beautiful, beautifully said. What's your advice, Chen? Yeah, so one thing I want to emphasize, not specifically for women who want to enter data science career, it's actually for every woman, is that how to never be afraid, be bold and be brief. Think about what we can do if we are not afraid. Once we feel safer, how to leverage our potentials, how to maximize our potentials best. Beautifully said, ladies. Two more questions. What is your advice for someone who is looking to be an ally to women like us? They've got a young woman or any age woman in their life who's trying to fight the good fight and get in a career in STEM or in data science. What do you think, Chen? What would be your advice? The question is about if someone wants to be ally, how they can contribute. Let me use one example I like most. I share with everybody. One of our executive sponsors is actually a male executive in our company. And he is, I should say, his Christmas and he's the best male ally I've ever had in my entire academic or plus industry career. So what I observe is that he is a champion, sponsor and supporter and mentor of all the female colleagues in our company and even globally. So I think to be an ally, you can feel free to take all these roles and what else? How to be brief enough to stand up to help women to advocate. Well said. Any advice you'd like to add to that, Bing? I think that she said too well. That was very well said. So I'll just say that I mean for women that already in this engineering or data science, specific data science world, that we need to be more open to the, let's say, newcomers. Because for them it's a little bit intimate at the beginning, right? Of course it's intimidating. I was intimidated even though they're coming in here today because you're also smart. Yeah, totally. It's a male-dominated field. Totally. We have to donate. But we can also see that there are so many women engineers, women data scientists already there and they have done all the work. They have been through the path. And yeah, I'm thinking just when we are already in this field we can get connected with newcomers and help. Yeah, always be willing to bring that person up and water level rises with those all together. I think that's a really great note. Anyone who's helped you on your journey that you'd like to give a shout out to? Yeah, I would say that when I was working in medical school and my advisor and he's my mentor. He's always very helpful for me that his name is Chuck Meyer. Well, thank you. Thank you, Chuck. And what about you? Anyone who's helped you on your journey? So I had quite a few mentors inside SML. Some of them are female role models. Yes. And some of them are male allies. And some of them are peers. From all these mentors I see has a genuine support attitude and they really want me to success. So they guide me and share me with many things. And later I become a mentor myself and mentoring a young lady in my company. So I really think that this type of... I asked one of my mentors, why do you want to mentor me since you are already very, very high level? He said I want you to mentor somebody else when you grow. So I want to say that to everybody that you are either in a growing journey or you are already well established or you are very green, very young in this field. Keep that mentality in mind. This community will grow. I love that. And your legacy is the impact you have on other people. It's why we invest the time being in China. It was so nice to have you here. Thank you for the work you're doing at SML and within the WIDS community. Just brilliant. I'm inspired and I hope you're all inspired wherever you're tuning in. Live to the Cube's coverage here at Stanford University. My name is Savannah Peterson. You're watching The Cube, the leading network for empowering females in tech.