New Blog
This blog post first appeared first on old medium publication (https://medium.com/startuprad-io), and was moved to this blog with the relaunch of our website in summer 2024.
Our software can assure, that the chip design you approve for production is correct. … we can guarantee that there are no bucks.Tobias Ludwig, CEO and Founder LUBIS EDA
Tune in to our Internet Radio Station here:
I worked in software development since I was 14. I started as a web designerTobias Ludwig, CEO and Founder LUBIS EDA
Subscribe Here
Find all options to subscribe to our newsletter, podcast, YouTube channel or listen to our internet radio station here:
I started with HTML and worked my way through to PHP and everything related to web.Tobias Ludwig, CEO and Founder LUBIS EDA
Our Sponsor Startupraven
The best way to identify investors and cooperation partners for early-stage startups. Sign up for early access here:
The faster you can get the chips in production, the higher are your profit margins. The first company takes almost all of the profits in this industry.Tobias Ludwig, CEO and Founder LUBIS EDA
A Brief Introduction to K-Town
This time we talk for the first time to a startup and a founder, based in Kaiserslautern (to the GI’s stationed close by known as K-Town). The city is locally also known as “Lautern.” This is a city in the German state of Rheinland-Pfalz, approx. 1,5 hours away from Frankfurt. They are known for their rich history, dating back to the 12th century, and their great Technical University. The Technical University of Kaiserslautern is home to two Fraunhofer Institutes and the Max Planck Institute for Software Systems. They are also home to one of the locations of the German Research Center for Artificial Intelligence (DFKI), one of the world’s largest nonprofit contract research institutes for software technology based on artificial intelligence, which shareholders include Google, Microsoft, SAP, and Daimler.
I found a colleague, working on a new method of hardware design.Tobias Ludwig, CEO and Founder LUBIS EDA
The Founder
We interview the founder Tobias Ludwig (https://www.linkedin.com/in/tobias-ludwig-5b1274a0/), CEO and founder of LUBIS EDA (http://lubis-eda.com,https://www.linkedin.com/company/lubis-eda/). Tobias has developed software since he was 14, starting out writing websites. Ever since he was working, developing software, and studying. He also spent one year at Michigan State University on an exchange program. The company is an early-stage startup, providing software for the semiconductor industry, likely a game-changer for the industry.
If you have a problem with the design of your chips is faulty, you may only find out after you started production. Than you have to trash all the newly produced faulty chipsTobias Ludwig, CEO and Founder LUBIS EDA
The Startup
The software of LUBIS EDA helps semiconductor producers to accelerate their production of new (http://lubis-eda.com,https://www.linkedin.com/company/lubis-eda/). In the chip industry, the first company to offer a new chip gets almost all the profits. Therefore, it is very important for semiconductor producers to be fast. The software of LUBIS EDA helps you to save up to 2 months in your development and assures your chip design is fault-free. With LUBIS EDA there is no need for starting over the chip development shortly before the production is about to start, since an error was discovered.
Conservative time savings estimate with our tool is 10%, which can go up to 20%. This may add up to two months in time savings.Tobias Ludwig, CEO and Founder LUBIS EDA
Venture Capital Funding
LUBIS EDA is currently looking for investors for a seed round, to maintain their competitive advantage and scale their product and sales.
Currently, we are ahead of the game, but we want to build on that.Tobias Ludwig, CEO and Founder LUBIS EDA
The Video Interview is set to go live on Thursday, November 4th, at 16.00 CET
The Audio Interview Podcast
You can subscribe to our podcasts here.
Feedback
Reach out to us, here is our audience survey, to give us feedback, suggest topics, interview partners or just to say “Hallo!” https://forms.gle/mLV6mVKwGwKuut8BA
The Interviewer
This interview was conducted by Jörn “Joe” Menninger, startup scout, founder, and host of Startuprad.io. Reach out to him:LinkedInTwitterEmail
Keep Up to Date
Here is our publication calendar: https://calendar.google.com/calendar/u/0?cid=MDEyaTI3YWs1MjVxaTNzbWdqbDh2OXRiaW9AZ3JvdXAuY2FsZW5kYXIuZ29vZ2xlLmNvbQ
Find all links and show notes here: https://www.startuprad.io/blog/lubis-eda-maybe-a-game-changer-for-the-semiconductor-industry/
Automated Transcript
00:00:00.000 → 00:00:07.760Music.
00:00:07.609 → 00:00:20.225The Startuprad.io podcast and you to blog covering the German startup scene with News interviews and Live Events,
00:00:20.267 → 00:00:28.705hello and welcome everybody this is Joe from startup radio darou your startup podcast and you to block from Germany bringing you again
00:00:28.702 → 00:00:39.453interview with the startup from Germany this time we’re going really deep deep deep check I do have to bias here,
00:00:39.549 → 00:00:42.694with lubis
00:00:42.565 → 00:00:57.242Eda or lubis Eda from K-Town known to the Americans who have been stationed there for the German known as Kaiserslautern and they are a start-up who helps semiconductor producers
00:00:57.140 → 00:01:04.633to be faster and cheaper on the market but first let me introduce you to Tobias hey how you doing.
00:01:05.080 → 00:01:11.357Hey I’m doing good thank you very much for the invite you are very welcome
00:01:11.219 → 00:01:23.700tell us a little bit about you as always people can reach out directly to you down here in the show notes there’s your LinkedIn profile so what do people read their and what do people don’t read there.
00:01:24.003 → 00:01:30.118Put what somebody does need to know about you besides you’re being a smart guy with a PhD.
00:01:30.772 → 00:01:37.689Yeah my LinkedIn profile is quite clean I guess maybe that’s a German have it to not.
00:01:38.073 → 00:01:44.441To provide not too much information to social media but I guess some things people don’t know about me is that,
00:01:44.546 → 00:01:58.925I’m working in software development since I’m 14 I started actually as a web designer when I was a kid I got like a small book was just HTML and the way through PHP MySQL and all like everything else related to web.
00:01:59.400 → 00:02:09.881During my studies I was working part-time as a software developer kinda like a Consulting based thing where companies hired me to develop software.
00:02:10.193 → 00:02:15.553So what’s basically always programming and working in the domain of software.
00:02:16.018 → 00:02:21.116And yeah I’m a software engineer at heart now comes the twist.
00:02:21.545 → 00:02:34.466I did actually plan after my studies to go to Industry what most Engineers do some project management job I was also working with Professor Van Til Kaiserslautern I also develop software for them.
00:02:35.022 → 00:02:46.898And I went to Michigan State to do my masters and foreign exchange programs or was one year in Michigan it’s very nice to encase you from the life estate it’s a nice one.
00:02:47.193 → 00:02:56.036Um but when I came back one question have you been there in Winter and did you eat a lot of warm socks.
00:02:56.591 → 00:03:10.205No actually I kind of missed the window a little bit so I didn’t state for fully adjust just nine months so I came just when the snow disappeared I could just see some piles of snow and lying around and I left,
00:03:10.328 → 00:03:18.487sometime November so there was some snow but I think I missed the really crazy really cold Michigan winter,
00:03:18.593 → 00:03:26.770but the fall was just nowso no riding of snowmobiles there unfortunately not no.
00:03:27.100 → 00:03:32.729But I had a lot of fun with the mosquitoes in the summer at the Lakes.
00:03:33.230 → 00:03:47.880Awesome big metal high-five to Michigan okay you’ve been there then he came back and what did he do that when I when I came back to professor already knew me for my for my work as a research assistant and he asked me if I would be interesting in pursuing a PhD.
00:03:48.363 → 00:03:55.856So I gave it some thought and I had a really good feeling about a professor I think he’s a great guy and I think he could really support me.
00:03:56.267 → 00:04:07.451And I decided to go for a PhD however the PHD was in the field of Hardware so semiconductor engineering so to say technically I have a PhD in computer engineering.
00:04:07.826 → 00:04:08.747Um
00:04:09.617 → 00:04:21.647And I was looking through what what everyone is doing at a chair because of course you can kind of find your own topic but you also have the connective for other people’s people are doing at the chair cannot just come up with something that’s not related to what chairs doing.
00:04:22.293 → 00:04:30.614And I found one guy he was working on a new methodology for Hardware design and I kind of looked at it I was like hmm,
00:04:30.845 → 00:04:40.219what he’s doing is kind of similar to what people are doing in the software World a lot so when I was programming I used an agile programming methodology called test-driven development.
00:04:40.927 → 00:04:50.446Case you don’t know what the core ideas that before you write any features of your software you write all the tests for the software so once you implemented the feature you know that the feature works correctly.
00:04:50.758 → 00:04:59.430And that methodology basically kicked off early 2000 and now I think most software is developed in that way more more or less.
00:04:59.850 → 00:05:01.662And I thought what he’s doing.
00:05:01.966 → 00:05:14.005That you could apply the same idea to the main of half of the cost Hardware’s mostly developed a software back in the day so you have like this V shape waterfowl models where you plan exactly everything ahead but in the end nothing works,
00:05:14.245 → 00:05:24.096um so I could see that switching to switching to a this kind of methodology could actually have the hardware world.
00:05:24.399 → 00:05:31.208So my research was mostly focused on the one hand side writing the software that generates something and the other hand side.
00:05:31.691 → 00:05:35.142Translate the idea of tests to the domain of hardware.
00:05:35.913 → 00:05:49.627And there’s something very similar that transports that everywhere which is called formal property checking I don’t want to go into any details but in the end it gives you a clear yes or no answer similar to test and software whether something works correctly or not.
00:05:50.047 → 00:05:54.947And I wrote a tool that generates you all the tests you need for your Hardware development,
00:05:55.187 → 00:06:05.912that’s also the basis of our smart IP generation today that my company is selling its so basically you PhD thesis gave you this idea and.
00:06:06.242 → 00:06:16.129What do you guys are actually doing can you really dumb this down for us Bonnie style and explain it like you would explain it to little child or business student.
00:06:16.702 → 00:06:25.726Yeah to business doing okay that’s a real challenge right.
00:06:26.497 → 00:06:35.637So to dump it down you can split the development of a semiconductor or microchip roughly in three phases there’s a specification phase.
00:06:36.363 → 00:06:42.983Where you basically just right information it’s like it’s just writing down the requirements and a really specifying what they should do,
00:06:43.205 → 00:06:56.153and then based on the specification you create something that’s called a design design is basically source code that it’s a digital model that specifies how the transistors are connected just on a digital model.
00:06:56.672 → 00:07:02.977And then the third phase you produce based on that design description the actual semiconductor.
00:07:03.496 → 00:07:12.546Now the problem and this is interesting for the for the business people the business students is if you make a mistake in your design and you have a problem in there.
00:07:12.904 → 00:07:20.919And introduced the chips and realize that the problem is they’re after you produce the chips all the produce chips can go straight to waste.
00:07:21.240 → 00:07:29.894And we have to be aware that producing a ship’s is crazy expensive because getting from a step from the Digital model to basically have a.
00:07:30.557 → 00:07:40.471Production of the chip that’s that cost a lot of money because you need to create something that’s called a mask and creating this mask is really really expensive and.
00:07:41.144 → 00:07:46.368What we can do is we can ensure that whatever you chip to the production,
00:07:46.392 → 00:07:56.486this correct so you don’t have the risk of having to throw away all your produce chips because we can guarantee that there’s no no bucks and no problems then you produce.
00:07:57.015 → 00:08:05.678So first of all you safe of course all the labor that goes into creating a design but also that’s even more important for the entire industry is that.
00:08:06.179 → 00:08:21.063Earlier get to market the higher profit margin is because usually companies are waiting for you to get a chips ready and whoever company gets the chips done first gets almost all the prophets so if you come late to Market your profit really drinks to bare minimum.
00:08:21.510 → 00:08:28.760So yes you’re safe to cost of labor but you also gain the time to Market benefit over your competitors were using our solution.
00:08:29.054 → 00:08:41.633In your personal experience what will be the time Advantage say cutting down from three months to three days or how could could a semiconductor producers for here
00:08:41.585 → 00:08:44.919conservative estimation is 10%.
00:08:45.222 → 00:08:54.020Savings of time but if you really fast and you really have experience with our Tools in our generation I think you can save up to 20%.
00:08:54.755 → 00:09:04.102In time this is considering that the creation of the design takes about 30 weeks so this is between a month and two months of Time Savings.
00:09:05.144 → 00:09:14.797And that may not sound that much but sometimes it happens something that’s worst-case for every Semiconductor Company is something called The Reason.
00:09:15.118 → 00:09:18.875So before you hand over your design to the production.
00:09:19.665 → 00:09:31.190You would still run a last pick test on your design and if you find a bug then you go back all the way and have to go through all the 30 weeks again and that may happen several times so.
00:09:31.962 → 00:09:44.469Tip on time can actually accumulate if you really doing a poor job in your design some of us you only go through it once you even faster and there’s no need for reason because we can guarantee that there will be no bugs in the design.
00:09:45.105 → 00:09:57.603Who that is pretty cool and how do you guys actually do this here you have some magic dust that goes over the hardware and then it all works out fine so we do two different things.
00:09:58.338 → 00:10:04.571So first of all we can generate the design so you can start with a graphical user interface,
00:10:04.829 → 00:10:12.862and you can kind of design your your your what you want to do it graphically means and then you can push button and generate a design.
00:10:14.155 → 00:10:18.227Bit works for simple and easy designs but,
00:10:18.395 → 00:10:30.416usually every or most semiconductor companies have like some Niche they’re really good at for example some companies are really really good at many writing really power efficient design or other companies are really good at,
00:10:30.440 → 00:10:37.933writing really performant designs that just can crunch a lot of data so for them like a general MIP generation.
00:10:38.416 → 00:10:49.159It doesn’t really help them so we can do something else that relates to what I talked about earlier the test-driven development approach we can generate them something that we call a verification IP.
00:10:49.705 → 00:10:58.476And if I talk about IP it’s not a pattern or something bigger it just means it’s a digital digital component you can use a source code basically.
00:10:58.852 → 00:11:10.971So we can generate this verification IP and they can use that to verify whatever you implement in your design and its really quick because it uses different technique than it is used today in the industry.
00:11:11.590 → 00:11:19.272And nobody is using this technique until now because it requires some expert knowledge to create this verification IP.
00:11:19.548 → 00:11:27.095But we can adjust automatically generated so everybody can use this improve technique and get that done faster with the process.
00:11:28.406 → 00:11:32.280That already sounds a pretty Magic
00:11:32.133 → 00:11:42.092can you give us like some idea he said a certain knowledge so if now an executive of a Semiconductor Company
00:11:41.945 → 00:11:48.439listens to this audio podcast or watches the YouTube video the question is right now.
00:11:48.868 → 00:11:57.000What would they need to change in their processes in order to be faster in almost error-free yeah.
00:11:57.843 → 00:12:04.274If you listen to me and you are you are CEO of a big company you don’t need to change that much I’m all you need to do is,
00:12:04.496 → 00:12:09.441instead of getting your verification team on board when the design is partly done,
00:12:09.510 → 00:12:23.647you will get it on board right away when the design process starts so as soon as the first line of RTL code is done you will have your verification IP in place so the design verification team actually would start ahead of the designers and specifies what the designer should do,
00:12:23.824 → 00:12:29.921and the design I can focus on is like main task of writing really efficient and really good designs,
00:12:30.153 → 00:12:36.970that’s the only thing that would change and of course you will need our tool to generate the verification IP that’s that’s of course.
00:12:37.795 → 00:12:52.157Yeah mandatory disguise that that sounds pretty magic basically you don’t need to change a lot of your process your faster and you also minimize the potential errors.
00:12:52.577 → 00:13:01.528That’s already customer work with you product or are you in a bitter face I think we are good you could call us we under seat says.
00:13:01.822 → 00:13:08.127Right now we have two core Novation projects running with customers they’re really interested in our tool but of course,
00:13:08.376 → 00:13:13.744the first couple of tries of customers we cannot really say that our tool is perfect yet,
00:13:13.786 → 00:13:28.616it’s a good better face so what we will do most in our use it in house but also the customer uses it and we do some projects together we kinda like have like the role of a little bit of a Consulting and the software company and we run in parallel with the customers through a project,
00:13:28.685 → 00:13:34.036and show them how it works and if they have questions or if there’s a bug in it will we fix it right away.
00:13:34.384 → 00:13:41.156Um this is our current state I think coming next year early next year we should have a,
00:13:41.387 → 00:13:47.116like a ship of awards and ready that the teacher can just use out of the box with like a day of Consulting.
00:13:49.426 → 00:13:57.189We are there talking right now we have to tell our audience that we were calling this towards the end of
00:13:57.150 → 00:14:07.299August but it will only be published in fall and when you talked about that we talked about like first quarter 2022 yes exactly.
00:14:07.773 → 00:14:17.138And basically then people can get in go to your website and reach out to you one question when you’ve been,
00:14:17.379 → 00:14:23.611telling us that you’re something between consultant and a.
00:14:24.373 → 00:14:34.954Enter software startup what is your aim at the end you want to be software you want to be Consulting so we see ourselves as a software software company,
00:14:35.203 → 00:14:38.825maybe I can talk a little bit about to team so we have four people.
00:14:39.344 → 00:14:50.564Three software developers and then one guy Max is mostly working on the administrative side of of things he could be considered a he’s considered our CFO I’m the CEO.
00:14:50.885 → 00:14:51.743And.
00:14:52.649 → 00:15:02.798So we are mostly working on improving the software but I’m not sure if this is special to the semiconductor industry or this is the case for many other Industries.
00:15:03.479 → 00:15:05.760For the customers they see the benefit,
00:15:05.991 → 00:15:17.571and they would like to use our tool but for them it only makes sense to use to put tool in production right away but as explained earlier of course the time to Market pressure is always there.
00:15:17.883 → 00:15:24.466So our aim is to start with projects now show the two works and.
00:15:24.760 → 00:15:35.152That’s basically how every Eda startup is working they do they collect their success stories and once they have a bag full of success stories customers just buy because they’re not work.
00:15:36.130 → 00:15:47.674Rotary spreader you have your recommendations on your website you can go to conferences and people just know okay this stuff works so the question I’ll do I mess up my production cycle or not doesn’t occur anymore,
00:15:47.851 → 00:15:53.102so our goal is that in two to three years there will be no more Consulting projects,
00:15:53.163 → 00:16:02.078then we then we will only be focusing on sales and try to scale the business app so long-term version Maybe.
00:16:03.173 → 00:16:11.836In 2027 so in about five years from the recording now we we aim to hit 25 million in revenue and 50 employees,
00:16:11.923 → 00:16:21.433so we really going for cross strategy here we try to get now VC funding to grow fast because also you have to know.00:16:23.050 → 00:16:34.621Democritus really competitive right now is some really big players in our Market I can just put up some names in case you want to check the stocks because they’re quite interesting is Cadence eiling synopsis and of course that’s also Zeman’s.
00:16:35.293 → 00:16:42.228And now we have an advantage we are head but actually we want to keep the advantage and build on top of it so when I think,
00:16:42.360 → 00:16:55.344if a classic bootstrapping a strategy that wouldn’t work because at some point the big companies would just throw money at the problem would out beat us we want to crawl fast now and take our or keep our competitive advantage,
00:16:55.467 → 00:17:05.355So currently you’re looking something like a seat series a investment right now I think we would be happy with the seat investment and.
00:17:06.927 → 00:17:11.026I think once we really have our success stories in our pocket
00:17:11.023 → 00:17:22.954and we can get rid of the Consulting projects then it would make sense to us to get a series a funding with a good evaluation have a like really good can really really prove to product-market fit so there’s no questions there,
00:17:23.023 → 00:17:29.598and you’re going to get a seat series a with a like a high evaluation and then scale it from there.
00:17:30.738 → 00:17:39.221Sounds pretty good do you think a Kaiserslautern where you located and we’re you guys are actually the first startup we
00:17:39.065 → 00:17:50.438are interviewing from there so they a little bit under the radar but that’s a very good very big Technical University do you think it’s a good place for start-up to start.
00:17:50.885 → 00:18:01.268Well we had this discussion as well we were also thinking about maybe having location and interests and exact same because there’s a big semiconductor industry of Germany.
00:18:01.859 → 00:18:07.849But on the other hand side since there’s not that many startups in Kaiserslautern but you have a lot of students,
00:18:07.972 → 00:18:19.750many students are from the region or like The Wider region I think we have a good chance of keeping talent that doesn’t want to go to Berlin Munich or treatment wants to stay close,
00:18:19.918 → 00:18:25.646the area maybe Frankfurt Heidelberg Mannheim sub clicking is a bigger bigger city around and,
00:18:25.769 → 00:18:38.969I think we have a nice niche there and also we have a good connection to the university since we two of our team actually did their PHD at Yale Kaiserslautern so we have a good connections to the chairs where first students come in first Ellen’s coming in,
00:18:39.182 → 00:18:41.850so I’m not worried too much about,
00:18:42.072 → 00:18:56.812finding talent and of course if there’s not that many startups there’s not that much competition for resources right everyone is happy if there’s a start-up funding the business in Kaiserslautern so we get really good support from all sides and.
00:18:57.511 → 00:19:03.203For us I think it’s smart to maybe stay in car lot on and and use the benefit of the local.
00:19:05.460 → 00:19:14.204Great awesome closing words only thing left for me to say everybody who would like to learn more you can go down here in the show notes
00:19:14.021 → 00:19:24.764there will be as I said in the beginning a link to your LinkedIn profile link to your company website as well as all the social media channels you provide to us so,
00:19:25.022 → 00:19:33.163thank you very much it was a pleasure having you here as guests Tobias and best of luck for lubis,
00:19:33.196 → 00:19:36.674Eda or lubis Eda what what’s the right pronunciation,
00:19:36.815 → 00:19:45.613Eda come from electronic design automation so we usually go with Eda but I think everybody’s saying it differently so Louisiana is also fine
00:19:45.601 → 00:19:56.479if you type it into Google in some ways I think we are the always ranked 1 so whatever you do you will find us lubis Eda is the name.
00:19:57.763 → 00:20:09.856Great thank you very much best of luck with the startup and I hope to talk to you soon after a successful fundraising yeah thank you very much thanks for the invite was a pleasure thank you bye bye bye.
00:20:10.420 → 00:20:22.198That’s all folks find more news streams events and interviews at.
00:20:22.480 → 00:20:40.234Music.
Comments