The global semiconductor shortage has been playing havoc with supply chains for a good 12-18 months at this stage. And it doesn't look set to be fixed any time soon.
And just recently my colleague Jeff Howell published a fascinating, comprehensive Point of View paper all about the semiconductor shortage. To find out more, I invited him to come on the podcast to talk about it.
We had a fantastic conversation talking about what Jeff uncovered in his months of research into the chip shortage, how it is affecting our customers, what they plan to do about it, and strategies to ensure any future shortages similar to this cause minimal disruption.
You can check out the point of view paper here. And on the podcast, Jeff also mentioned some videos that were developed to support the chip shortage. Here they are for you:
Support the show
Oh, and this is one of the first podcast episodes that I have created chapters for. If you find them useful, do let me know (email@example.com) and I'll make the effort to do it more often!
If you have any comments/suggestions or questions for the podcast - feel free to leave me a voice message over on my SpeakPipe page or just send it to me as a direct message on Twitter/LinkedIn. Audio messages will get played (unless you specifically ask me not to).
If you want to learn more about how to juggle sustainability and efficiency mandates while recovering from pandemic-induced disruptions, meeting growth targets, and preparing for an uncertain future, check out our Oxford Economics research report
I'd like to sincerely thank this podcast's generous supporters:
And remember you too can Support the Podcast - it is really easy and hugely important as it will enable me to continue to create more excellent Digital Supply Chain episodes like this one.
Podcast Sponsorship Opportunities:
If you/your organisation is interested in sponsoring this podcast - I have several options available. Let's talk!
If you have any comments/suggestions or questions for the podcast - feel free to just send me a direct message on Twitter/LinkedIn.
If you liked this show, please don't forget to rate and/or review it. It makes a big difference to help new people discover it.
Thanks for listening.
You know, what do we really see as the as the root cause of this problem? You know, how is it impacting our customers? And, you know, what could we do about it? What suggestions can we offer? What kind of like set a point of view? Can we provide our customers so they have some confidence that yeah, there is a path to solving this.Tom Raftery:
Good morning, good afternoon, or good evening, wherever you are in the world. This is the digital supply chain podcast, the number one podcast focusing on the digitization of supply chain. And I'm your host, global vice president of SAP. Tom Raftery. Hi, everyone. Welcome to the digital supply chain podcast. My name is Tom Raftery, with SAP. And with me on the show today, I have my special guest, Jeff, Jeff, would you like to introduce yourself?Jeff Howell:
Sure. Thank you, Tom. My name is Jeff Howell. And I'm the head of the high tech industry business unit at SAP. And it's basically my job to make sure that SCP is providing solutions to the high tech industry that are relevant for them over the next two to five years.Tom Raftery:
Okay. And I mean, that must be a fun place to be and actually high tech because I I'm very ADD always have been and it is why, I get into science first and then tech afterwards because it's, there's always something new and fun and happening there is are you similar? Or is that what got you into high tech?Jeff Howell:
Yeah, I actually will like, come from a little bit of a lineage of high tech, my dad was an engineer or electrical engineer in Silicon Valley back when Silicon Valley gotten started. And I remember when I was out of college, I had two job offers one to work in a semiconductor company and the other would work in a lock company. And two very different Yeah, opportunities,Tom Raftery:
Choices, choices.Jeff Howell:
And quite frankly, I just wanted a job. And my dad says, you know, this chip stuff is never gonna go away. I don't think lock door either. But it was, I think, a really good piece of advice for my dad. So that's, that's how I got to it.Tom Raftery:
Locks are not going away. But they're becoming smarter. They're requiring chips more, more and more now. They are converging finally. Yeah, that's right. But that that's, that's, that's a nice segue, actually. Because the reason we have you on the podcast is because you wrote a point of view document on the chip shortage. So this is I mean, your father said, chips are never going away. But they're actually becoming harder to get right.Jeff Howell:
That's right. And this chip shortage paper was kind of swimming around in my mind when my wife and I went out earlier in the year to buy some new appliances for the kitchen. And they said, Well, if you want an appliance that doesn't need a chip, like a rangetop, I can get get that one right away. But the ovens Nope, because there's a lot of chips in there. And then we live out in the country also wanted a pickup truck, couldn't get a pickup truck. Same reason. So that you know, this is a real big problem. And you don't have to look very far to find just news of the week where some manufacturer or there was some disruption in supply chain because of the chip. So that was sort of swimming around in my mind said, Well, it's time to write a paper about this, because I think, you know, we can have a good point of view with our background here in the industry.Tom Raftery:
Okay, so what's involved in in writing a paper like that? And what kind of outcomes did you get from it? What what did you find out when you when you went to write this, because we've all read the headlines, and we're all aware of some of the stories that are going on. But that's not from having done much research, just just from reading a couple of articles. What, what did you call, like I said, What, what's involved? And what did you find?Jeff Howell:
Well, what's involved is just, you know, starting with some kind of passion, because it's not like, you know, we all wake up in the morning say, hey, let's, let's just write a paper. I mean, that's something that doesn't come very natural to us. And, and so I would say it starts with some level of passion or some interest in it. And we knew that a lot of our customers at SAP were experiencing a lot of challenges around this, this problem. And then what's involved is just getting the right people together to collaborate on, you know, what do we really see as the as the root cause of this problem? You know, how is it impacting our customers? And, you know, what could we do about it? What suggestions can we offer? What kind of like set a point of view? Could we provide our customers so they have some confidence that yeah, there is a path to solving this. And so I would say it's just pulling the right people together, because it's a very complex problem.Tom Raftery:
Okay. And what were the findings? What did you find? Talk me through the the document itself?Jeff Howell:
Yeah, so it's, it's first of all, we had to say, Okay, what what are we trying to contain, like, how are we going to scope this thing? We said, okay, what are the big, you know, outcomes that our customers could or should expect from us? From a document like this and how it impacts their business, who said, Well, there's really three things? How can they improve the relationship with their customers? Maintaining customer expectations is one, maintaining revenue continuity is another. And then finally, how do we prevent, or at least establish better processes. So the next disruption that happens isn't quite so violent or disruptive. So that's, those are the three things we're hoping that the readers of this document would see as a, as a path to addressing those those issues. And, and what we found when we dug into it was, this is not the first chip shortage the world is seeing. Right? It was one back in the late 80s. We also have, you know, that was kind of geopolitical between the US and Japan. And then we also had one, I think, in the early 2000s. And then, of course, natural disasters will have an impact on these as well. And so but we did find that this was probably the most amplified, of the previous chip shortages, and the longest lasting I mean, they're saying that depend on the report, you read, this is going to go well into this year, and likely in the next year. Yeah. And that's one of the one of the things that we found out of this, but the reason this is so, so much more amplified than the previous chip shortages we've had is that semiconductor chips are so prolific. I mean, they're, they're in everywhere. In fact, we found a source where just in the US alone, semiconductors make up point 3% of the US GDP, but are required for 12% of our national output. So a lever that these chips have in the overall economy just in the US is very significant. Of course, it goes beyond the US. But that's that's one of the big findings through this process.Tom Raftery:
Yeah, I mean, wherever you're, you talked about there being a requirement for chips in your oven. I mean, if I look at my desk, here, pretty much everything on my desk has chips in it, in some form, even a two lights pointed at me, they're, you know, just for shooting video, for example. And they've got chips built into them, because they're, I can turn them on and off and control the color and the brightness from my phone, you know, so they're, they're on Wi Fi. So even the lights in the room have chips in them, everything these days has chips in them. So I gotta think that's part of why we've hit the shortage, the demand has increased. I don't want to say exponentially. But the demand has increased enormously. And there was a wobble in the, in the, in the outputs when we hit COVID.Jeff Howell:
Well, yeah, and COVID is, is the kind of the poster child that a lot of people would like to point to for a number of things. But you know, when we looked at it, you know, sure, it had an effect in so much as a lot of the plants in the manufacturing facilities are used to make chips, or even the equipment they use are used to make the chips, you know, they randomly do shifts and early and COVID. And so we lost a little bit of capacity there. But it was really the confluence of multiple things kind of hitting all at once. And I should say, over a very short amount of time. You know, one you mentioned, you know, in your home office? Well, when we, when COVID hit, one of the things that happened was many employees not just in the high tech industry, but across multiple industries established, you know, a work from home policy. Yeah. So and what that did is you saw in your office, everything, you know, so people are building out their home offices that require chips, and the bandwidth requirements for video calls. Just catapulted right. So that was one of the big ones. And then we also had some, you know, trade activities between, say the US and China was one. And as China was trying to build out their capacity to produce more chips, because they saw the demand increasing. Well, they also put the sanctions were driving some, some nervous behavior with the OEMs. And China, who said, Wow, this is this could be a problem, we may not get the chips that we need from the US because of the sanctions, therefore, we're going to start building up inventory this to get ahead of it. They didn't want corruption. And then we also had some other things around COVID where the automotive industry actually, you know, reduced their forecast temporarily and tried to get back in line again, and that was a big one. But that's you know, that's on the demand side of the equation. But then on the supply side of the equation around the same time, there was a fire in one of the facilities in Japan that produce a lot of automotive chips. And then the US put sanctions on the equipment that's used to make chips on China so China couldn't get the the equipment they needed to make chips that happened in December of 2020. So it's really, you know, I would say that if any one or two of these hit in that same period of time, I suspect that, that there was probably enough slack in the system to absorb most of that. But it was all of that hitting at once. It was the perfect storm. Yeah, that's right.Tom Raftery:
Okay. And you mentioned in the the point of view document, you look at how it has affected our customers. And so yeah, how has it affected our customers?Jeff Howell:
Well, it's, like I said, personal example, you know, all the way at the end of the value chain, that consumer like you or I get a pickup truck was a problem getting appliances a problem, but then they also had, you know, farm equipment was sitting, idle waiting for, for chips to complete it. And it just, it just got manifested throughout the entire economy, a lot of the economy anyway. So there's a lot of frustration, you know, I used to find frustration is when your expectations exceed reality. So people expected to get things when they wanted it, but that just didn't happen. So there's a lot of frustration.Tom Raftery:
And my, my poor 15 year old son was hoping to get an Xbox series X and of course, zero chances of that happening, right? Being a bit glib there, but how so? How are how are our customers responding? I mean, what are they doing to? I mean, they gotta be some way dynamic in reacting to this.Jeff Howell:
Yes. And that's something we point out in the paper, you know, what our customers doing today, and at SAP and our partners system, we're not so bold and arrogant to ever believe that our customers are just standing watching it happen. And that's not at all what's going on. I mean, our customers are, you know, they have a lot of very smart people working incredibly long hours to try and, you know, get through this. And, and there's a number of things that they're doing. And I would categorize it in probably three time horizons, things are working on today, obviously, there's the immediate, that's a lot of firefighting is going on, you know, just looking for daily updates, and even some cases, Hourly updates, you know, when are we going to get these chips so we can finish our production? That's one and there's a near term that's sort of like, well, today, they got to, they got to put something in place, whether it's processes or tools that will help them, you know, through maybe months six through month 12. So it's just kind of that near term. So maybe there's a stop gap, we can we can plan for today that helps us in the near term. And then finally, there's there's some things that they're doing today, for the long term. You know, that's, that's really the other word we put on that as how are they building resilience. So the next, and it will be inevitable, there will be another disruption, it could be an excess of supply disruption, it could be another shortage, but you know, how will they create the systems and the processes? So the next thing that happens, they're gonna be a lot more resilient. So that's how I put it putting those three categories of kind of the immediate near term and long term. Okay.Tom Raftery:
So, one story I've heard, and I'm not sure how true this is, I suspect it is. But it's, it's about Tesla, for example, and what they've done to make sure that they weren't as heavily impacted as other auto manufacturers. And it was because they fully control their software stack, they were able to rewrite it so that it would match whatever semiconductor chips they could get their hands on. And you know, that that's the beauty of being very vertically integrated, I guess. Are there any other interesting examples like that, that you've come across?Jeff Howell:
Yeah, there's there's a couple interesting examples that, you know, we've seen one is what we call match set optimization. And so the there's an old mathematical principle called the bakers problem, where the baker makes multiple things, cakes, pies, cookies. And in order to complete that, they have to decide given this supply. What do I make? Yeah, and there's also a belief that, well, I don't want to allocate all of my chocolate to cookies, because maybe I won't make as much money overall. And if I'm limited on chocolate, that's how I'm going to do it. But some of those ingredients or products actually share eggs and milk and flour. So the question is, given my supply, what do I make, and that's what we call a match set optimization and it also you want to reduce the amount of stranded inventory. So in that example, You know, you want to make sure you're exhausting as much of the inventory as possible, you know that in the bakers problem. You can't put the cake in the oven, unless you have all the ingredients. So it doesn't do you any good to have everything except the eggs, for example. So, when you think about that, it's like when you're making cars, yeah, well, you know, do you want to buy all the inventory? And, and not have the chips? So the question is, you know, given the chips that are available, which cars should I make? And that's something that I think companies are now starting to look at, in addition to their traditional planning methods, where they say, oh, let's let's go and try and capture what customers want to buy. Well, that's, that used to be the primary planning, approach, demand driven. Now we're going to couple that with another voice of the table says, aha, but this is all that I have available. So how do we marry that together to come up with a plan?Tom Raftery:
Okay. And, you know, looking looking forward, as you pointed out, this is unlikely to be the last time that there's a chip shortage or some other kind of big shortage that impacts manufacturers. What can they do to minimize the impacts of something like this? And in the future? Yeah,Jeff Howell:
this, this is a big, challenging question. And there's a couple of our customers. Zed F is a company that makes a lot of the platforms for the automotive industry. And there's an initiative called CatenaX, which is Open Source Initiative, it's basically an automotive network that's starting in Europe. And what Sedef is doing is now talking to their suppliers, like wolfspeed is a company that produces the chips are used to regulate power in those chip and in the vehicle. And they're now saying, well, we'll speed we'd like you to be on this network. So it's the big automotive OEMs, inviting multiple customers on their or excuse me, their suppliers on their network, and their suppliers, suppliers will be on and so on. So that entire value chain is going to be on a single network. And the vision is that when an OEM says, we're going to reduce the forecasts of our vehicles, but we don't want to just necessarily reduce the forecasts of the chips, they're gonna be using the vehicles because maybe I still want some safety stock of the chips. So if demand goes back up, I still have it. But my chip suppliers will know about that immediately. It's kind of like the pebble in the pond, that ripple will will be visible to everybody. And the other thing that is going to happen out of this network, this automotive network is they won't be simply exchanging information about forecasts and commits. This is how many I want, when When will I get them, but we'll also have sustainability information that's going to be provided. So now the the automotive OEMs, in this example, will not only get their logistics information, but they'll also get the sustainability information from from their entire network.Tom Raftery:
Nice. And is that just European OEMs?Jeff Howell:
As of now, that's it. But we're starting to see that there's other initiatives and other networks that are that are starting to build and other countries and so on, but it's it seems to be starting with the Europeans.Tom Raftery:
Okay, so it's, and that's just the one industry that's just automotive. But what, what it says is that one of the big answers to this is access to data, its visibility of data and information. So I mean, bottom line, that's what we need to do. It doesn't have to be CatenaX but it has to be in all industries, that we need to have better visibility of what's going on.Jeff Howell:
That's exactly right. And in fact, that's one of the three tenants that we talked about in this paper. It's, as you said, it's visibility. But it's also insight, and collaboration. And it's all three of those working together. So visibility is like the example I gave, you know, providing a forecast. And then they say, Okay, here's the commit, and it's kind of like Where's my stuff? And when are we getting it? So like what's going on in that network? But insights a little different now. It's, it's subtle, but where insight comes into is when they're producing chips, for the five months that it takes to get your cell phone, for example, it spent in a front end fab, just making the chip. Well, in that four months, a lot can happen even in the best of times, you know, we could have big yield losses, there could be, you know, overrun of capacity. I mean, there's all kinds of things That can happen in these over 1000 steps it takes to make a chip. And the question is in that four month period, you're getting those chips, whether you want them or not, but are you going to get 100 die? What they called good die per wafer? Or is it 120? Same equipment, same period of time. And so there's a lot of data that's used to create those chips. And we found a source that 80% of that data is never looked at, because there's just too much out of it. And, and so what if there were hidden truths in that data, while they're making those chips? That could help the semiconductor company work with the OEMs to say, hey, you know, what, we seem to be falling off the yield Cliff here, you know, or maybe, maybe we have satellite grade level reliability for a cellphone that's only going to last three years in the field, maybe, maybe we can at least use that data to have a different kind of conversation with the OEM to see if we can improve the yields. Okay.Tom Raftery:
And is that is the that awareness of the need for that insight? Is that common, you know, or are people starting to become aware of it? Or where are we in that spectrum?Jeff Howell:
Yeah, well, the way it typically works is, you know, there's process engineers, and they, they will vector in on a problem when the problem comes up. So that you say, no, no, I need to see that data. Because there seems to be some failures out here. But exception management. Yeah, that's right. And it's no fault of the company. These are very, very smart people. And it's just a capacity thing, right? We're Yeah, creating, I think it's 15 gigabytes per wafer times 25 wafers times 100,000 wafers a quarter, we're talking about a lot of data, and no human can be ever expected to, to analyze that, by they have these systems that can, you know, pull it together. And so you would augment that current process with this other approach.Tom Raftery:
Right? Okay. Maybe run it through an AI or something. Right? Look, look for some kind of weirdness and go, maybe you want to take a look at this, because it's a bit out of the ordinary.Jeff Howell:
That's right, a more proactive approach. So right now we've got a very reactive approach. But this is proactive, we're gonna go and like he said, there's there's AI and machine learning algorithms that can be applied to that big volume of data to say, keep doing that. But we also want to bring another voice at the table. And this is what the AI ml Big Data voice is telling us. It's telling us that we're about to fall off the yield cliff. Or we might be able to relax things without compromising quality. So there's a lot of what we think is hidden, hidden truths in that data, but they're just not looked at yet.Tom Raftery:
And what about third party data, things like I don't know, looking at social media, looking at geopolitics, looking at all kinds of external factors, whether it's stuff I mean, which was a big drought in Taiwan, which is affecting the fabs for a while, these kinds of things. Is that part of the mix as well? Yeah, absolutely. AndJeff Howell:
that that kind of gets more into the front end forecast, you know, when we're thinking about, you know, like the home office example? Yeah, I mean, no, I don't know that anybody really predicted that we were going to need to have multiple monitors and mice and keyboards, I mean, that happened very, very quickly. But you also brought up a very good point about the kind of the droughts in Taiwan, Taiwan produces quite a bit of our semiconductor needs on the planet. And, you know, there was a drought. And just by way of contrast, you know, fab probably uses anywhere from two to 4 million gallons of ultra pure water a day. So there, you know, he kind of go back to that network and the insight a lot of semiconductor companies are now working to see okay, can we reduce the amount of consumption of not just power, but that ultra pure water? Can we recycle more of that water? And so that's, that's becoming more and more evident, because if we don't have enough water, we can't make enough chips. So that's,Tom Raftery:
yeah. I mentioned geopolitics. And having a lot of fabs in Taiwan is maybe not such a good idea.Jeff Howell:
Well, yeah, and that's, that's a very good one. There's a number of initiatives to restore these fabs back to Europe. European Commission is now in recommending an initiative that's getting significant funding now in Europe to rebuild Europe's capacity. So they're not as dependent on external countries for their for the chip requirements and what's happened. What's interesting in Europe is this. This chip act as they're calling it, is intended to restore Europe's presence in the in the chip markets. So for example, Europe used to have I think, in early 2000s 20% Of all the chips manufacturer in the world came from Europe, it's now down at 10%, though. So they're trying to rebuild this not just in volume, but also in technology. Well, there's going to be some challenges around, you know, do we have enough of the right people? Who can? Who can staff these fabs? You know, not just the engineers, but also the operators? This is going to be a challenge. And not just for Europe, but also in the US as this reshoring takes place.Tom Raftery:
Interesting. Yeah. We are coming towards the end of the podcast. Now, Jeff, is there any question that I haven't asked that you wish I had, or any aspect of this that we've not touched on that you think it's important for people to be aware of?Jeff Howell:
Yeah, I think the other one is just the emphasis of, of the sustainability requirements on this, you know, when we really look at yields, for example, you know, use that example, where you get 100, good die per wafer, we're going to get 120, or maybe even 150, good die per wafer as an example, you got to remember, if we can figure out how to improve the yields in that very long process for metal and process. If we get more dye per wafer, and we can be smarter about the decision making on on these good dye per wafer, we're going to start fewer wafers, which is really good for the environment, it's ours, I should say it's better for the environment. So and I think, in that case, companies will achieve their goal, maintain more revenue, continuity, greater resilience, and of course, meeting their customer expectations.Tom Raftery:
Okay. So, Jeff, if people want to know more about yourself, about yourself, Jeff, how are about the point of view paper we talked about? or about any of the other topics we discussed in the podcast today? Where would you have me direct them?Jeff Howell:
I would say to on my LinkedIn profile is a good place. Oh, and that's the best way I'm on LinkedIn quite a bit. And so happy to respond to any enquiries or, or even questions or comments. And I'd love to hear from other folks too. This, this paper does have a little bit of controversy in it. So I expect Okay, well, the to react and love to hear from them. And so yeah, I'd say my LinkedIn profiles the best.Tom Raftery:
And where can people find the paper?Jeff Howell:
sap.com. And we'll provide the link for that.Tom Raftery:
Okay, I'll stick that in the show notes. So people have access to it. You mentioned in the prep as well, that there's a couple of interesting videos around this.Jeff Howell:
Yeah, we, we created some videos, because some of these concepts are very complicated things like yield management, and some of the plant maintenance, which we didn't really get into today. But what we did, we took some of these concepts from the paper and created some videos. They're also on sa p.com. They're four or five minutes in length, but they're entertaining. And it's just an easy way to kind of understand some of these concepts. And so that's another resource I'd recommend to you. I didn't read a long white paper. SoTom Raftery:
great, correct. I put links to the videos in the show notes for this podcast as well. So people have access to those two. Great, Jeff, that's been really, really interesting. And I look forward to reading that paper myself and maybe watching the videos too first, so that I have a bit of a handle on it before I before I head into the paper. And thanks a million for coming on the podcast.Jeff Howell:
Well, thank you very much, Tom. It's a pleasure.Tom Raftery:
Okay, we've come to the end of the show. Thanks, everyone for listening. If you'd like to know more about digital supply chains, head on over to sa p.com/digital supply chain or simply drop me an email to Tom Raftery at sa p.com. If you'd like to show, please don't forget to subscribe to it and your podcast application of choice to get new episodes as soon as they're published. Also, please don't forget to rate and review the podcast. It really does help new people to find the show. Thanks. catch you all next time.