#541 – Chip Shortage Denier

Chris Gammell: This is The Amp Hour Podcast. Released May 10th, 2021. Episode 541. Chip shortage denier.

Dave Jones: Welcome to the Amp Hour. I'm Dave Jones from the EEV blog.

Chris Gammell: What? What? What? You know, I'm Chris Gammell. Text electronics. I'm sleep deprived. I'm back.

Dave Jones: Sleep deprived? It's been four weeks. I'm back.

Chris Gammell: Tell us. Baby crying in the background.

Dave Jones: Whoa, hang on. Was that an announcement?

Chris Gammell: That was an announcement, yes. Yes, I am a father. And yeah, I've been on paternity leave for the past month. Woo-hoo! Yeah.

Dave Jones: Paid for, but from your own company. That's right.

Chris Gammell: Analog Life LLC. My boss gave me a solid four weeks off. So don't let them tell you, US corporations aren't generous.

Dave Jones: Congratulations. Thank you very much. There's a baby Gamble.

Chris Gammell: That's right. That's right. Baby girl. Very healthy. Healthy mom.

Dave Jones: Yep.

Chris Gammell: It's a whole new life experience and it's been fun. I like it.

Dave Jones: That's because this is the easy part. You don't have to do much. It's like, come on. Yeah.

Chris Gammell: Yep. One of my friends gave me advice of just like, everything's a phase. And I'm like, yep, that's good. That's what I, that's my mantra over and over again. So.

Dave Jones: Brilliant. All right. So what happens over there with paternity leave? As in, as opposed to maternity leave? Well, let's speak about maternity leave as well. You know, cause this is important. This, you know, people want to. Sure. Yeah.

Chris Gammell: I think it impacts things. Yeah.

Dave Jones: Yep. Worthy topic. What's it like in the U S because here a maternity leave, if you're a full-time employee, don't quote me, but the company has to give you six or nine months or no up to 12 months, I think off with no pay or negotiated half pay or something like that.

Chris Gammell: That's like the federal mandate thing.

Dave Jones: Yeah. They have to be flexible. Like, you know, if you're a full-time employee, that's, that's one of the, I guess, downsides of from like a financial downside of hiring a full-time employee is that you are on the hook for these sorts of benefits. Sure. You know? So, yeah.

Chris Gammell: Yeah. I mean, that's part of a functional society, I suppose, uh, around here, a little less functional, but the minimum is basically the only thing that the, the, the federal leave policy in the U S is it's FMLA family medical leave act. The only thing it guarantees is that you get 12 weeks and if you come back, you will have a job, nothing else. So, right.

Dave Jones: So that's all it is. So is that paid leave?

Chris Gammell: That is not paid. There's no, there's no federal.

Dave Jones: Okay. That's right. Right. Okay. Here, here it's, yep. I'm right. It's 12 months. You, they have, you have to guarantee their job for, and it's yeah. Unpaid, but, or there is, I thought there was a paid or they're allowed to take all the holiday pay or something. I don't know. Yeah.

Chris Gammell: Some, some countries have that. And yes, in the States, there are many companies that do that. Some of my wife's friends are like at like law firms and they have like six months paid kind of thing. Like just like, it's like an employee perk kind of thing. And that's great, but it's just not every company, every company or every job.

Dave Jones: So is it, although I don't believe you're protected if the company makes bona fide redundancies. So if the, if during that 12 months, the company makes redundancies and you're like, that doesn't protect you from being one of the redundancies. So unfortunately, but it has to be a bona fide redundancy and there's got to be, you know, that's has a legal definition and a legal, you know, stance and all that sort of jazz. So. Yep. Yep. Yeah. Cool. And so does that, is that the same for paternity as well as maternity?

Chris Gammell: There is no guarantee. As far as I know, there's no guaranteed paternity leave.

Dave Jones: Right.

Chris Gammell: But again, a lot of companies offer it as like a perk. Yep. Paid or unpaid. And then, you know, in my case, it's all unpaid. Well, it's paid, but it's not, uh, you know, it's coming out of the kitty, you know?

Dave Jones: So did you just stop doing consulting work or what? Did you just.

Chris Gammell: Well, no, I warned my clients about six months ago. So they were some of the first people to know that we were having a baby. And, uh, yeah, so all my projects, I kind of lined everything up as best I could. I have some great clients and they were all very accommodating and, uh, yeah, it worked out. It worked out well.

Dave Jones: And of course you have to tell them, you can't just say, oh, I'm taking a mysterious month off, you know, like, yeah, I'm, I'm just a slacker. I'm going to be on a beach in, you know, like.

Chris Gammell: Yeah. Right. I'll be out. Leave of absence.

Dave Jones: Yeah. No. So you had to tell them. Yeah.

Chris Gammell: I think at that point I'd be like, I'm, I'm no longer a consultant for a month. You know, I'll, you know, I'll be back at some point. If you want to hire me back, if you're desperate for someone in the future, you could hire me back then. But yeah, it is. That is the ups and downs of, uh, of consulting. Uh, and actually, so I should speak to a consulting as well. So I've mentioned on here, I started doing consulting course about three months or well, back in like January, February.

Dave Jones: As in teaching it or get or taking one.

Chris Gammell: Yes. I've, I've been recording it. So I've made a video about it, but basically I've been putting together, I've been recording. I've probably recorded about 50 videos just about how to be an electronics design consultant.

Dave Jones: Wow. Yeah.

Chris Gammell: Because like all this, all the other stuff that kind of, I always struggled with when I was getting started. So, and started making all these videos, had people sign up for it. And I've unfortunately just sat on that mailing list. I haven't even reached out to them. I really need to. And I feel bad about that, but basically people that were interested signed up as like potential beta members.

Dave Jones: Even though you shot the content, is that because you haven't edited it properly or what?

Chris Gammell: No, no. It's all sitting there ready to go.

Dave Jones: It's all canned. Yep. In the can. That's right.

Chris Gammell: It's just, there is a tutoring element as well. I wanted to make sure that it was like.

Dave Jones: Oh, okay. So you're available to, to, right. Okay. Well, you could just release it. You don't have to, you could just release it as a paid thing. Here it is. Take it or leave it kind of, you know? Yeah. There's more I want to make that too. Or do you think there's value add to.

Chris Gammell: Well, this is also the beta. So it was like kind of like early, like I did with the contextual, the first time. Yeah. Right. Contextual electronics. I also had done, you know, basically I wanted people to go through it, ask all the, you know, point out the stupid things I did, ask all the questions that I hadn't answered, that sort of thing. And that's really what, you know, that's pretty high touch. And so, yeah, I didn't have the mental bandwidth for that. Right.

Dave Jones: Yeah. Fair enough.

Chris Gammell: That's unfortunate. But, you know, in the future, it's, it's interesting. You know, I, I've been trying to reach out to people that I know that are consultants as well, just to get, you know, understand what's tough, what's not, you know, I think there's a lot of interest in it, especially, you know, like in the era of COVID too, I think a lot of jobs have gone away as well. Yeah. So people are interested in kind of.

Dave Jones: Or changed.

Chris Gammell: Or changed. Right. Exactly. Or work remote, that sort of thing. And it's just like the nature of work has and is changing. And so just thinking about that from an electronics design perspective is, you know, like, what does that look like? What do you need to know otherwise? That sort of thing. And it's, it's, it's tough. You may, if you, if you amplify the sound right now, you actually may hear my daughter wailing in the background. So. Okay. Yeah. Cool. Yeah. Yeah. Yeah. Yeah. We're doing it live. That's right.

Dave Jones: So we're going from dogs barking to. Oh, the dogs are too. So yeah.

Chris Gammell: He'll kind of at some point, it's a true cacophony of sound. And my wife's an angel. Nice.

Dave Jones: So do you find that there's a significant, how much overlap is there for our industry, i.e. the electronics design consulting industry and other types of consulting industries? Have you been talking to other ones that?

Chris Gammell: Yeah. So this is, I market it as a technical consulting course too, because it is. So basically I'd say like mechanical, anything with like physical goods, I would kind of put in the same.

Dave Jones: Okay. So, so you're actually designing something, be it electrical, mechanical, maybe even, you know, architectural, you know, if you're an architect or something like that, maybe or some other.

Chris Gammell: I mean, I'd say like software, I'd put software in there too. But I think there's just, there's kind of.

Dave Jones: Oh, there's a significant. Okay. Yeah. Right.

Chris Gammell: I'd say there, there are a lot of similarities. I just, I don't know as many software like freelancers or, you know, consultants and things like that, because it feels like the benefits of joining a company full-time are so good. I just, I don't hear about it as much or they're not in my circles or something like that. So I don't, I don't quite get as much exposure to that, but I'd say mechanical, you know, industrial design, electrical design, stuff like that. I, those are kind of the people that I talk to on a regular basis. That's, that's who I have as a kind of a, a background with.

Dave Jones: Right.

Chris Gammell: You know, actually there was, so I've been, I still listen to podcasts and thank you for, uh, I appreciate you having all those guests on while I was away. There was some good shows there. Well, I did skip a week, but you know. That's okay. I mean, yeah. But you know, we've had, Big Clive's been on our request list for, I think. Oh yeah, forever. I think since like six months in, people are like, you start to be Clive. It's like, okay, well, uh, so that was great. And then Fran. And then who was the third one? The last one was, uh, last one was Fran. No, I know who was before Fran. Clive. Who was the, oh no, no, no, no. Sorry. It was the missile man, Bruce Simpson.

Dave Jones: Oh, yeah. How did I forget Bruce? Yes.

Chris Gammell: Yeah. Of course. No, those were, those were all fun. So people should check them out. They were good.

Dave Jones: Yep.

Chris Gammell: I, my, my one friend's like, is everything okay, Chris? Dave's recording the intros. Yeah. Yeah. Everything's okay. So you can tell something's up, man.

Dave Jones: Something's up when that weird ass voice comes in for the intro.

Chris Gammell: Yeah. Yeah.

Dave Jones: Well, at least I went to the effort to put an intro in this time. Cause when I've done it before, I always went, nah, it's gonna be asked. You know?

Chris Gammell: That's right. But yeah. So to, to, you know, like in my late nights and, you know, dog walks and all the things I've been doing in the meantime, I, you know, listen to other podcasts and I think it was on, yeah. Episode three, no, not three 70, three 69. Yeah. Three 69. Alicia and Chris over on embed.fm were talking about consulting as well. And there were some good questions about that. So yeah, that's the one. So there was, there was a good discussion in there as well. And they were kind of, the person was asking about contracting or consulting after like five years. And I think like to answer the, to answer their, their listeners questions.

Dave Jones: You're either, it's like Schrodinger's engineer at that point. You're either alive or dead or both possibly. You're both alive and you go, Oh geez, I wish I was never a consultant.

Chris Gammell: Oh, you mean at five years as a consultant? No, this is five years as an engineer. And then they're wondering about whether to get into engineering and into consulting. And I think that's interesting because it is, I I've had people like reach out to me. They're like, Oh, like I've been doing this for a year. And the thing that I always talk about is like, and something we've mentioned on the show here in the past too, just if you can get mentors in that scenario, I think you're still in a good place. But I think the benefit and the thing that people should target always is trying to find people they can learn from. So if they're, you know, if, if you think you can go out on your own and still find mentors and find people to interact with and learn from, then that's fine. And obviously find work.

Dave Jones: Jeez. If there was only a forum for that.

Chris Gammell: That's right. Yeah. Yeah. That's true.

Dave Jones: Go on, plug it. Jeez. I gave you an open shot there.

Chris Gammell: There is a consulting forum. Yes. There's an application. I'll put that in again. So if you already are a consultant, you can apply for consulting forum. Ooh, it's exclusive. You've got to apply. That's right. It is a, is a hidden forum because we want to make sure that people feel comfortable talking about things. So, yeah. But I think it's just about like, I don't know, like, what were you, where were you in your career? Like five years in five years.

Dave Jones: Oh, five. No. Oh, technically I was. No, I was a consultant five years in. Damn. Okay. Interesting. I was. So what was that like? Well, sorry. I wasn't a consultant. I was a contractor. So I was working for a recruitment company. And they hired me out as a, as a, as a, uh, like as a consulting engineer. No, consulting is the wrong word. It is a, you're hired like, like on a six month or nine month contract or something like that. So you, so you're embedded in a company. Yeah. Yeah. So I'm a contract. I'm embedded in a company because they've got a project for nine months or 12 months. And I work on that and then bugger off. Right. Yeah. Yeah. And then, yep. Yep.

Chris Gammell: You're like a mercenary.

Dave Jones: Yeah. Yeah. Mercenary for hire. Yeah. Yeah. So I was, yeah, I was, geez. Yeah. I was doing that five years in. Yes. I believe so.

Chris Gammell: And how did it work? I mean, did it go okay?

Dave Jones: Oh yeah. I, I kept on doing that for a decade. So, you know, yeah, yeah, yeah. Yep.

Chris Gammell: Yeah.

Dave Jones: It had a lot of tax implications. I wasn't aware of like my, my, my, my first tax bill. I thought, you know, like they weren't like, there was almost hardly any tax coming out or something. Yeah. It was like minimal tax. It was like 10% or something. I'm making a ton of money. And I went, this, this, uh, contracting gig is great. You know? And then my tax bill come back and I did not have the cash to pay for it. It's like, holy crap. You know? Yeah. It's.

Chris Gammell: I mean, that'll bite you here too. Like 10. So like in the States, it's a 1099 versus W2. 1099 is like the independent contractor. Right. You know, basically it's telling the government company, a paid individual or company B, you know, $5,000 for this work. And so that they can track on both sides of it. And that's great. But then like W2, that means there's no, no tax taken out of it at all, actually. And so.

Dave Jones: You're fully responsible for it.

Chris Gammell: That's right. Yeah. You get hosed.

Dave Jones: I liken it to the difference between high school and uni, right? High school and university. High school, they hold your hand. Like university, it's up to you, dude.

Chris Gammell: Yeah. That's right. Yeah. Like it's even. You don't know how to study. That's on you. Yeah.

Dave Jones: Nobody cares if you turn up or not. Like, you know, that's on you. Right. Yeah. So, yeah. And it's the same thing on tax. Like nobody cares if you don't pay your tax, but at the end of the year, you're going to get stung if you don't account for it and, you know, put it aside or pay it incrementally or something like that. You've got to do that yourself. Otherwise, you're going to get stung at the end.

Chris Gammell: Yeah. I think that's actually, that's a, that's a good point. And I think broadly as well, like, so things like, like articles of incorporation. So, or sorry, not articles of incorporation. I'm saying this wrong.

Dave Jones: See, that's not something I had to do as a contractor. I wasn't my own company or anything like that. I was just, there's a different law for contract or a different, you know, thing. Yeah. I was just in a person. I wasn't a company.

Chris Gammell: Right. And so, yeah, there are things when you are a company, like that you need to have in place. And if you don't know about it, it's like, okay, well, so like one of the things, I think it was like a monthly yearly report or something like that. The dog just kicked in. Yeah. Yeah. I heard that. So yeah. Yeah. Yeah. Okay. So we're, we're just living the circus here, but yeah, it's the same thing where if you don't have, if you haven't filed the necessary reports or things like that, no one's going to knock on your door that day. However. No, that's right.

Dave Jones: But a few years down the track.

Chris Gammell: But then when they do, yeah, you are, you are screwed.

Dave Jones: Yeah, exactly. Yeah. Hey, the IRS put, it's, they put Al Capone in Alcatraz for tax. It wasn't for anything else. Right. Yeah. So if they can get him, they can get you. Right. So it's, yeah. A lot of people don't think. Oh, yeah. It's great. Yeah.

Chris Gammell: So that was unfortunate that I didn't get that stuff out on time, but hopefully I'll do that in the next couple of months and have that as a thing that people are trying to get into.

Dave Jones: That is a lot of content canned up. Jeez. I couldn't, I'd, I'd explode if I was sitting on that much content.

Chris Gammell: Well, you know, yeah. You and I are slightly different in that way, I suppose. Yeah, we're very different. Oh boy.

Dave Jones: Yeah. All right. So what else has been happening?

Chris Gammell: For me, you know, just chip unavailability. I don't know if you've been doing any projects lately.

Dave Jones: Well, I tweeted this out the other day. A lot of people took it the wrong way. Right. They, they, they thought I was a chip shortage denier. Right. They think I was a tinfoil hat chip shortage denier. Cause I tweeted out that I thought, you know, I've had this on my list for like a couple of weeks and I finally got around to it. I thought I'll shoot a video on the chip shortage. Right. And, and I won't just mention it. I will demonstrate it.

Chris Gammell: Yeah. Like go through DigiKey or something.

Dave Jones: I'll go to the Digi, I'll go to the catalog suppliers and I'll surely, cause everyone's been saying, everyone's been mentioning ST micros, for example, and regulators and all sorts of things. And so I thought, oh yeah, it'd be easy to find examples on DigiKey. So I started recording this thing and I went to DigiKey and then like I searched for S I searched all ST micros and then I sorted by stock level and I'm going, well, there's 57,000, there's 40,000, there's 30,000. These are all in stock.

Chris Gammell: No, that's not the way to do it. Right.

Dave Jones: Yeah. And well, that's the thing. Right. I thought like, I can't honestly do an honest video by cherry picking. Right. I've got to go, well, there's a global shortage of ST micros. And I went, no, there's not. I had to get to page 12 before I found one with less than a hundred stock.

Chris Gammell: Here's how I would do it. I would do it. I would do it from the other direction though. I would find a common design, like a, like a blue, or was it a black pill? Is that what it's called? Like there's a, there's some STM 32 based like Arduino board. And so like that becomes the, that is like the seed crystal for a bunch of other designs. And right.

Dave Jones: See what the shortage is for those individual parts. Yeah, exactly.

Chris Gammell: Try and find those parts anywhere in the world, that sort of thing. Or in my case, like NRF 52, I couldn't, you know, when I just went to go buy a bunch, I think this was the last time we recorded. Maybe I was looking for those and I was digging into my own stock just because like I couldn't,

Dave Jones: right.

Chris Gammell: There was nothing out there. And that's the thing. Like once, once it is gone or once it's like your variants. Oh yeah.

Dave Jones: No, you screw it. Yeah. It's a Murphy thing, right? You're, you're guaranteed to get the part and, you know, so yeah, I wasn't implying that it's not happening. I've been saying it's, you know, a huge problem. I've been tweeting. It's a huge problem for like six months or something, but it's yeah. No, I just, I was embarrassed by the fact that I couldn't find anything on. Did you key and mouse for all of these jelly bean stuff I went for? Right. And, and it was just, Oh, this is embarrassing. I can't shoot a video with this, you know? Yeah.

Chris Gammell: I mean, but if you're also sorting, but I would, the problem is that you can't sort by like the most popular parts. This is, this is my ongoing gripe. I always, I always want to see the most popular parts, but yeah, those are gone. I'd say instead you should like look for the most angsty comments on Twitter. Yeah. And then just follow that. Oh no.

Dave Jones: Once I posted that, I got a whole tweet of angry responses saying, Oh, I can't buy this. And you're denying that, and that companies are shutting down because of the, you know, geez, we mentioned on the amp hour, just a month or two back that, you know, General Motors are shutting down plants because they can't get parts. I mean, you know.

Chris Gammell: I think they said, I think, I think my one friend told me that like a billion, like two different car companies are going to lose like a billion dollars each. Yeah. Because of the cars they can't make or something like that. Shortage. Yeah.

Dave Jones: Yes. And for those who don't know why it's happening, one of the major reasons why it's happening, it's not that the, you know, the chip suppliers can't open their factories because of COVID or anything like that. It's because the supply chain is a complex. Beast. Beast of, of a FIFO bar. Like it's a complex, like. Yeah. That's a good way to think about it. Right. It's a very, like a six month long FIFO buffer. And it's like, if you're an Apple or you're a General Motors or you're a somebody else, right? You're going to know you're going to need 10 million of that chip six months in advance. So you put in your order six months in advance. Right. And then, but as soon as COVID started happening, all these companies suddenly went, holy crap, nobody's going to be buying a thing. And they shut down and they canceled all their orders.

Chris Gammell: They went, yeah, nah, yeah, nah.

Dave Jones: Yeah. Yeah, nah.

Chris Gammell: Yeah, nah.

Dave Jones: Have you, you've actually learned the yeah, nah, have you?

Chris Gammell: Yeah.

Dave Jones: But you've got to say it more, yeah, nah.

Chris Gammell: Yeah, nah, yeah, nah.

Dave Jones: No, no, that's, that's far too American. No, you've got to.

Chris Gammell: Sorry, Dave. I, you know, there's only so much I could do here.

Dave Jones: Where have you learned the yeah, nah from?

Chris Gammell: I think Aussie man, maybe. Oh, Aussie man. Yeah, yeah, maybe. All right.

Dave Jones: Okay. Yep. Are they the only two channels you listen to? Me and Aussie man.

Chris Gammell: I mean, how much Australian witticisms can you get, really? Right. Okay. Yeah. Before you go bonkers. Yep.

Dave Jones: Anyway, it's yeah, nah. It's, for those who don't know, that's a classic Australian expression. That hasn't been around that long. It's only been around for like 20 years or something. It's like when I was a boy, there was no yeah, nah. Okay.

Chris Gammell: We don't need the, we don't need the etymology. Right. Okay.

Dave Jones: It's just our way of saying we add yeah, nah before, you know, if you see, if you see an Australian interviewed on a TV, especially one person who's not used to being interviewed or talking or something like that on a camera, they'll go, yeah, nah. Yeah. It's, it's, it's, it's our version of arm kind of thing. Yeah. So it's, yeah, it's fascinating. Anyway, where the hell were we? I was mentioning. Oh, you were talking about FIFO buffer. Yes, FIFO pipeline. So all of these companies, once COVID happened, like March, April, you know, March, April last year, they, they all went, holy crap, we better stop buying parts because this is going to get bad. Nobody's going to buy anything. So they stopped or canceled all their orders. And so all the semiconductor companies went, oh, what do we do? Okay. We'll stop manufacturing. And then, you know, four or five months ago when, uh, no, even.

Chris Gammell: No, it wasn't.

Dave Jones: Actually, it wasn't, it wasn't long after that.

Chris Gammell: It was like, it was like three months after. Yeah. Three months after.

Dave Jones: It was about three months after that. People just all of a sudden started, they started to work from home. So they had to buy all these monitors and computers and phones and new everything. Right. And people were just buying stuff like mad.

Chris Gammell: I think the other thing too, is people had disposable income, like crazy, because you're not going out. Exactly. You're not going out for drinks. You had extra disposal. After work. Exactly. You're not going to lunch every day, you know, and it's like people are just saving money and they're like, well, I want to buy stuff because I'm American. Yeah. Although not buy American. Exactly.

Dave Jones: So bottom line is all of these, all the big companies bet that retail sales would tank, but the opposite happened. There was a boom in retail sales. So once they realized that they bet wrong. Almost every, I don't know if you know of a single company that bet right on this and went, Oh, I'm going to buy up all these chips and I'm going to buy it cheap. That's right. You know, let us know in the comments, but all of a sudden, yeah, there was this massive consumer demand and there was stock in the buffer so they could fulfill it. But once all that stock started to go and people were buying new cars and all sorts of things crazy. And once all that stock was gone, all the companies, they couldn't ramp up again because it takes so long to ramp up production of all this stuff. And then all the companies wanted to order at once. Right. So there was no shuffled pipeline anymore because a huge part of a semiconductor manufacturer's job is trying to prioritize and sort out is the logistics of having 20,000 different major customers, right? All ordering all these parts. And, you know, so they're very expert at staggering all the orders and, you know, pipeline optimizing their production flow and everything else. But when all of a sudden you get zero and then you go to a hundred, every single one of your major clients, and I'm talking the biggies, right? Your apples and your GMs, they all want their chips now because all of a sudden this consumer demand and they've bet wrong. Whoa. It's like they can't just suddenly make that many chips for that many customers. Yep. So, you know, anyway, yeah, all of these companies, they bet wrong. And then the manufacturers couldn't supply them all at once, once their demands kicked in. And it's going to be nine months to 12 months to sort it back out again. Like, it's just.

Chris Gammell: Yeah. I mean, the frustrating thing. So, like, you know, they're trying to like, you know, so TSMC announced a hundred billion that they're investing. The US is talking about a 50 billion investing. Intel is talking about, oh, we're in the foundry business now. We talked about that last time and all these things. And, you know, there's, they're talking about taking like major action at the federal level and all these things. And I'm just like, it's not going to frigging matter. You know, like it is going to, it's going to matter in a year, right? Or a matter in two years or five years.

Dave Jones: No, it's not going to matter. I was going to say.

Chris Gammell: Sure.

Dave Jones: Yeah. Well, it's nice having that extra capacity, but everyone knows what the problem is. Like, no one's going to make that mistake again by canceling all their six month backlog orders.

Chris Gammell: Yeah, no, but they're also talking about, they're talking about a couple of things, right? They're one is they're talking about just the straight ahead sourcing issues. Yeah. There's not enough, not enough silicon in the pipeline. And just, I think the real thing that we need is to un-mothball a bunch of fabs. And I don't even know if that's possible at this point, but basically, you know, there's been fabs that have been slowly shutting down in the States over the years. That's probably the best case scenario, I would think, for getting any kind of capacity online. But I have to imagine that they're just buildings at this point that they sold out all the equipment to places wherever, right? Because there's capital equipment in there. Why wouldn't you sell it?

Dave Jones: You wouldn't leave it there. No. You wouldn't just let it sit there and get cobwebs. You're almost starting from scratch at that point, surely. I mean, it's...

Chris Gammell: Right. Yeah. Well, and so then, so someone was asking me about this too. They're like, well, okay, so there's 50 billion proposed to the States. That's great. And I'm like, yeah, that's great. They're like, well, what would it take to make that into a fab? I'm like, well, if you think it's hard to build Chris's simple circuit boards, let me tell you, ordering a bunch of dry etch machines or a bunch of EUV lithography machines, how long does it take to get those? Those aren't just sitting around anywhere either. Those are all in order.

Dave Jones: I was going to say, yeah, where do the chips come from to power those, right?

Chris Gammell: Yeah, exactly. So it's not like Tokyo Electron is just sitting around with a bunch of stock. They're building to order as well. And so they call this stuff.

Dave Jones: No, it's the same problem.

Chris Gammell: It's a mess. They might have pieces that they could... They have modular etch machines. I was just looking at their site the other day. Like, yeah, so they have modular etch machines. They might be able to piece them together, but those are etch machines not going to someone else. So basically, yeah.

Dave Jones: And if they've already got the orders in and it's in the pipeline, you can't just bump orders out. Although that brings me to one of the reasons why I wanted to do the video on the component shortage is that force component supply shortage. I got an email from Bryman, from the CEO of Bryman, because I'm one of their major distributors. Anyway, to whom it may concern, shall I read it out? Component supply shortage notice. No. Why not? Come on. And dear valued customers, Bryman has been receiving delay shortage notices from worldwide suppliers on integrated circuits, passive components, as well as plastic materials. It's not just components. Plastic materials as well. Lately, the delivery situation is getting worse. It's not getting better. And some of our vendors have issued force majeure notices concerning deliveries. Force majeure. Force majeure. For those who don't know what that means, that's French for something. I think. Don't quote me on that.

Chris Gammell: Major forces, basically. Yeah.

Dave Jones: It's basically, yeah. It's a legal definition, which is either written into contracts or it's embedded in case law or some other law in your particular country, which means that you can get out of a contract if there's circumstances beyond your control. I.e. You can't get like you promised. Yeah. We're going to ship, you know, 10,000 multimeters on this date, but sorry, we can't get the part. So you can't sue us. It basically means you can't sue someone for breach of contract.

Chris Gammell: Well, you can. You just. Well, you're not going to win.

Dave Jones: You can sue anyone for anything at any time, but you're not going to win because force majeure is a legal thing that allows people to get out of basically fulfilling their contracts. Yeah.

Chris Gammell: No, it's the fulfillment part. Yeah. So it's like, so my buddy, I wish my buddy was able to talk on the show about it because he's like in this world, but he's like, he was talking about someone who pulled the same thing and he's like, it's not like, it's not like Steve Carell in the office declaring bankruptcy. You can't, you can't just, I declare force majeure. Yeah. No, I mean, you still have to like, it's basically like mitigating circumstances type of thing. You still have to deliver. It's just, you don't have to deliver on the same timeline. Oh yes.

Dave Jones: No, no, no, of course you can't necessarily get out of the contract, but it means that you can't be held to that date that you promised. You know, you can't pay penalties because you didn't meet. Like there's a lot of contracts. People don't know this built into contracts that you have to pay a penalty, a financial penalty if you can't deliver on time.

Chris Gammell: That's right.

Dave Jones: So, you know, that's yeah.

Chris Gammell: And it's like, oh, what if you don't pay that penalty? It's like, okay, then you go to court and you pay it through your lawyers basically.

Dave Jones: But yeah, force majeure is one of the ways. Anyway, basically the CEO of Bryman is coming out and saying, Hey, all of our suppliers are saying force majeure. Sorry. We, we, we can't deliver those chips as promised, you know, whether or not that's the situation that like a more important customer came along than Bryman, whether or not, you know, like, cause you can't really prove that behind the scenes, but you know, it wouldn't surprise me if that was, you know, that was actually happening. But yeah. So it's not just electronic components is as the important one there was the plastic parts. They're having problem getting plastic materials as well. So it's, it's not just us electronics.

Chris Gammell: Right. But that's, that's probably more of a capacity thing than it is like a, it's not like.

Dave Jones: Yeah. But it's the same thing. It's specialized pellets. No, it's exactly the same issue. It's not like the fabs have shut down. Right. It's not like the fabs can't make these chips. They can. There's just so many people demanding them all of a sudden. That's the problem.

Chris Gammell: I would disagree with that because there's interchangeable parts and there's non interchangeable parts. Right. I, I, I posit that many silicon parts are not interchangeable. Right. There are some, right. Of course there's non-silicon parts, right. You can go and get different resistors and you know, there might be 150 versions of that 4.99 K resistor that you need to get. Fine. Great. That's interchangeable part. But that TI switcher that you've been trying to get that just literally doesn't exist anywhere in the same footprint. That is a non interchangeable part. Sure. Plastics, unless they're using some like plastic mix that is unpurchasable. Right.

Dave Jones: Yeah. But ramping up and going to another supplier, usually the manufacturer holds the molds and things like that. Right. They're really expensive stuff. You can't just bugger off to another plastic supplier and say, give me 10,000 multimeter cases. Right. You've got to tool up again.

Chris Gammell: Yeah. I guess, I guess, I guess where that, where that difference comes in is the, you know, how you define the, like the, the true cause of interchangeable, not like not being able to buy an interchangeable part. Right. So in your case, you're saying that like, yeah, the plastics you can't get, it's a non interchangeable part because you're buying a finished good from, which is the plastic, the injected plastic from the supplier. Yeah. I guess from that perspective, it is.

Dave Jones: And unless you already had that manufacturer qualified, right. And, and they have the tool in and they're qualified and they're ready to go, or you're already ordering from multiple, the same part from multiple, multiple suppliers, just for a backup. I've worked for companies that have done this, right. We would not single source our plastics, right. We would actually have multiple suppliers ready at any one time, just in case something happened to that one company. We didn't want to be called short. Right. But most companies won't go to that sort of effort. They go, you know, the one hung low factory is just giving us 10,000, you know, things a month and then they don't give a thought to them actually going under or be, or some other crisis happening.

Chris Gammell: Right. Or we'll deal with it when we need to deal with it when we need to. Yeah, exactly.

Dave Jones: It's not something, you know. Yeah.

Chris Gammell: Yeah. That's true. I guess I, I was thinking about it in terms of like the pellets that are the raw material, but yeah, you're right. If it's a, if it's purchasing a finished good kind of thing, then it is similar.

Dave Jones: And, and then it's qualified. It can change the whole look and feel of your product, right. The customer might notice, Ooh, this plastic looks a bit different. Ooh, it's a bit more brittle. Ooh, it did like, does it pass all the qualification, you know, drop tests and, and ruggedness and vibration tests and everything else. Right.

Chris Gammell: Yeah.

Dave Jones: Who knows?

Chris Gammell: Yeah.

Dave Jones: Just because it's plastic, plastic ain't plastic.

Chris Gammell: Right. I think that we're entering an age of people taking that step, probably not out of like prudence, but in, you know, obviously someone's going to be reactionary or it's mostly reactionary, but it's like the cost of waiting is much, much higher than the cost of revising. Right. So like to use my own example, that switcher chip that I mentioned I couldn't buy, it's gone. It's designed out. So TI, you've lost a, you've lost a socket. Yep. And, uh, I designed in an, an LT part or analog devices. Ugh. Uh, uh, and because, yeah, because I could buy it.

Dave Jones: What's wrong with analog devices?

Chris Gammell: No, it's because they're got rid of the name, Dave.

Dave Jones: Oh, well, okay. Right. Got it. Yeah. Yeah. Okay.

Chris Gammell: Cause they bought LT, you know? I know. And yeah, I mean, and, and so in this case, the, the thing that is interchangeable is my ability to lay out a new circuit board and requalify it based on that. Right. That's that is, that is the lever that I'm able to turn, uh, based on that. And the risk is, you know, that the next part goes out of stock or whatever, but basically, you know, having multiple revs of a board. Now, this is where you start getting into like letter revs and number revs on top of it. Yeah. Like, so I have a rev rev C1, C2, C3, C4, and all the numbers are because I could only get certain parts with certain builds and it's just, Oh, what a mess. Yep.

Dave Jones: So, so are you putting down multiple footprints and multiple, have you specified multiple bomb items now? Have you learned your lesson or, or you, or you've just switched to one that happens to be available because your volume is so low. You don't, you just need to do a one-off thing.

Chris Gammell: That's right. Yeah. I'm still in, this is even just for, you know, low run prototype type stuff. But at this point, what I do is I don't buy the board before ordering all the parts, making sure I can get them. Yep. And even today, actually Jay Carlson, former guest Jay Carlson, he actually tweeted about the same thing and said, basically he will only send out for a board once he has the parts in hand. And I think I'm going to do that as well. Yep. Yep. Yeah. I just, I can't risk it anymore.

Dave Jones: I can't, I've lost count of the number of times where I've looked at what literally one day I've gone, Oh, there's the, oh, did your key have 20,000 of these parts? No worries. I'll order them tomorrow when I'm ready to order the rest of the parts from tomorrow comes around and they're gone. They've got zero stock and I'm screwed.

Chris Gammell: Oh yeah. See that's, so that's, that's what I'm doing right now. That's I'm taking a similar tactic. I'm saying I only, when I've gotten, get the PR or whatever, will I, or get the invoice, will I send out the board? Yes. This is actually saying once I actually physically have the parts, because even having the, you know, until they're actually shipped to me, you know, they don't, it's a Schrodinger part type of thing, you know, it doesn't exist until it's in your hand.

Dave Jones: So you were, but, but you will trust the shipping notice from DigiKey and Mouser. So then you don't.

Chris Gammell: No, that's what I'm saying. Not anymore.

Dave Jones: Oh, you, you won't even trust the shipping notice.

Chris Gammell: Nope. Nope. Nope. I will, I will have parts in my lab validated.

Dave Jones: What do you think they, they, they're going to ship you the wrong parts?

Chris Gammell: Sure. What if I, or what if I ordered the wrong parts? You know what I mean? Like it's just, at this point, yeah.

Dave Jones: Well, if you don't want to trust yourself at that point. Yeah. Okay.

Chris Gammell: I mean, no, this is just, this is just being burned a bunch of times, you know, and it's just like, there's, it's just the next step in the progression of like, of being paranoid. It's just because it's not like the, it's not like the, the boards are expensive, but the time's expensive. So it's better to basically burn two days in order to get the parts in hand and validated and then send out the board versus get everything to the assembly house. And, uh, and it turned out actually, no, that one part wasn't included in your order. And now you have to go spin your board again and then do everything again, over again. And it's just, I can't, I can't burn that, you know, two weeks of the cycle time, because if, even if I'm using a, you know, a board house that has a four day turn with shipping from China, it's like, okay, that's a solid two weeks. And by the time, you know, the assembly house looks at it and says, Oh, nope, sorry. You can't get this part. And I have to go re-spin the board. It's another two weeks. And so it's just like, it's better to burn two days up front than lose two weeks on the back end. Yep. And it sucks. I mean, it sucks because.

Dave Jones: Make sure you have those spares as well, because when they manufacture, there's going to be X amount of loss. And well, at this point, they may be scraping them off the pick and place assembly floor, you know, and they might be like, and then hand assembly.

Chris Gammell: Oh, I've, I've been sending in parts. I think I mentioned this last time, but I've been sending in parts from my personal stock to just, yeah, just to fill out the order because I couldn't get parts anywhere else. So I just had to send in, I had to send in the parts I had and it's fine. You know, slight variance on microcontrollers, things like that, you know, just got to get something in hand. Cause like, what else are we going to do? This is what we do for a living. Yeah, exactly. Yeah.

Dave Jones: Oh boy. Anyway. So yeah. And it's not going to get better anytime soon as the, once again, the president, sorry, not CEO of a Bryman said the chaotic set will not, he can't see it improving, even improving in the next few months at least. Yeah. So he's, he's basically saying, sorry, but we have no timeline when this is going to get better. So that, that that's the CEO of a major corporation, you know, telling us directly.

Chris Gammell: Yeah. The other, the other way I've explained it to people, I've explained this to family members. Like, they're like, why are you stressed? I'm like, this is also, it's like, you know, like you mentioned a FIFO buffer. That's good for people on this show.

Dave Jones: Yeah.

Chris Gammell: But I think explaining it to other people is like, it's like a car crash, right? So you're sitting in traffic now and all of the, you know, the car crashes happened. They've cleaned up all the, the bits of the car crash and you're still sitting there for the four hours.

Dave Jones: And you're stuck in the traffic because everyone's going slow. You're stuck in traffic. Then you have to wait for the person in front of you to move on. And then that's right.

Chris Gammell: Right. Right. So the, the car crash in this case is getting, getting all the silicon necessary in to the wafer starts. And then, you know, my car moving is, but by the time the chips have been all, all the way propagated through the supply chain and sitting on a distributor shelf. And I happen to catch it in time because it's very likely that even if the, you know, the order date is, you know, if, if, if DigiKey or Mauser says, oh, there's 20,000 parts on order, that's great. If I get, if I get scooters, I swooped, I called it swooped on Twitter the other day. If someone swoops in and buys all those parts underneath me, like I'm screwed. Right. You know, like, so I could, I could outlay the cash now, but it's just like, it's a cashflow nightmare. And also someone had, uh, someone had proposed that it'd be called reeling. I'm reeling from being swooped, you know, like reeling. Really? Like a reel of parts.

Dave Jones: Ah, we're here all week. Yeah.

Chris Gammell: Yeah. I like it.

Dave Jones: It's all right. Once again, nobody outside the industry will get that. No one will get it. It's like reeling. Okay. You're like, as in reeling a, you know, a fish in line or whatever.

Chris Gammell: Like Steely Dan, like Steely Dan reeling in the years, you know?

Dave Jones: You're right. Yeah. Yeah. Yeah. Yeah. Yeah. Oh boy. Yep. Yeah. Anyway, big problem. Can we label this one? What was it that I labeled myself at the start of people were labeling? Well, you know.

Chris Gammell: Oh, chip, chip shortage denier.

Dave Jones: Chip, chip shortage denier. Yeah. Yeah.

Chris Gammell: I like that. Yeah. I think that's good. I'm sure there's a photo of you somewhere with a, you know, tinfoil hat.

Dave Jones: There is a photo of me with a tinfoil hat. Yeah. I can, I can send you that. There is.

Chris Gammell: Great. I will take that. All right.

Dave Jones: All right. Remind me at the end of the show. Good.

Chris Gammell: Yeah. Yeah. Yeah. Really, really important stuff that we're doing here on the Amplified. Oh yeah, totally.

Dave Jones: Yeah. Because I'm an Aussie and it's like, I don't give a shit about taking the piss out of myself. You know? Oh, I'll, I'll, I'll post that thumbnail of me wearing a tinfoil hat. You know, it's just for shits and giggles, right? Yeah. Anyway.

Chris Gammell: It is, it's going to be bad. I mean, so, and like I said, if you're not, if you're not being careful now, there's someone else too on Twitter was like, you know, when you saw these shortages happening three months ago, why didn't you buy six months of stock? Which is part of the problem as well, by the way. And it's like, man, I do, I do two designs a month. Like, like I would love to have bought all the parts that I need for the next six months, but I just, that's not, that's not how this works.

Dave Jones: Well, you've got to have a cash buffer in your business to take care of that. That's part.

Chris Gammell: Oh, cash is a different problem. That's not, that's not a problem for me. Right. Luckily. Then what was the problem? The parts that I, so if, if they, they said by six months with the components that I'll need, you know, the components that I'll need in six months from now, like, I don't know what sensors I'm going to be designing for some clamp. Oh, okay.

Dave Jones: Oh, you're talking about new design. Okay. Right. Okay. Oh yeah. Right. Right. Not, not, not, if I'm building the same thing. Yeah.

Chris Gammell: Oh, sure. Yeah. If it's a manufacturing thing, then yeah, of course. But that, that's a, that is a cash problem.

Dave Jones: Yep.

Chris Gammell: Yeah. No, this is, this is a, this is a design problem. I'm, I play a little, little towards the front of the design cycle. Manufacturing has got a whole bunch of other issues. Mm-hmm.

Dave Jones: And this leads us into the, uh, often, this was like a whole Altium's vision and everyone else's is the modular electronics thing. Oh, if you've just got this little switching controller here, you know, if you've got, if you just standardize on this switching controller and you do your layout and that's it, it's done. You can use that in every design forever. It's like, you know, eh, no, not quite.

Chris Gammell: I mean, that's, that's fine. But then you also have to be sure on like a, to the, to the microsecond, you know, accuracy type of thing that you, that your switching controller has the exact same behavior. You know, if it has any different behavior that could just propagate all the way through

Dave Jones: your system, or if you need a different efficiency, you need a different price point. You need a different, whatever insert difference in design here. There's a dozen of them.

Chris Gammell: Yeah.

Dave Jones: Right. Then, then that entire concept falls apart.

Chris Gammell: Right. I like, I like the idea, like modularity in general, like I like the idea. There's always higher costs to it. That's fine. You know, like some, some industries can handle that. That's great. Like I think about it, what it really is, is basically you're making modules much like, you know, standardized chip pinouts are basically, you're making little modules that now have its own standard pinouts. And, and that could work. It's just that there's a whole lot of investment to do that sort of thing. If you, if you did invest in that a year ago and you made your whole design modular, then you're probably doing pretty great right now. Right. You, you're, you know, you can't get switcher ABC one, two, three. Well, the, you know, you've got it.

Dave Jones: It depends on the business you're in. You know, if, if you're a consultant, you know, if you're a consultant, as we're talking about here, you're working on vastly different designs from one, you know, every couple of months. Right. Maybe. Sure. Right. Right. And you can't do that. You can't just, you know.

Chris Gammell: Well, and that would work in a prototyping contact context for sure. I mean, if you were, yeah. So like you just plop down your switching converter, you know, five volt to 3.3 volt footprint. That's basically just a tiny circuit board. It's going to drop on there with like castellated edges or something that works great. And, you know, then it's super switchable and you can, you know, you can swap in things as needed, but are you willing to, you know, use that space and the cost and the overall complexity to put that into your final, final goods? You know, it's like, okay, probably not.

Dave Jones: Yeah, exactly. And there's a difference between what type of products you're working on. Are you working on a consumer product? Are you working on like high volume, low volume? Is it a one off? Is it a 10 off? Is it a hundred off? Sure. You know?

Chris Gammell: Yeah. Yeah. Yeah. I think, I think the modularity makes a lot of sense towards the, you know, the lower volume side of things for sure. If you, you know, because you probably have more margin because you probably have weirder requirements, you know, you can, you can really, you could really play with that well. But I think that a lot of people that are doing electronics, you know, they're, they're getting paid because there's, there's higher volume needs and they're, you know, they're getting paid to introduce efficiencies that, that modularity is kind of the opposites of.

Dave Jones: Okay. Okay. Can we get onto, because, well, we've got the whole bloody show about the component crisis and which is all COVID driven. Can we talk about remote working? It, it was on the list.

Chris Gammell: We can.

Dave Jones: America's super cities are coming back.

Chris Gammell: Oh, we've talked about this.

Dave Jones: Remote work is overrated. Have we talked about that?

Chris Gammell: Well, we, we called it before this article was written, I think. I mean, yeah, basically I don't think people are leaving the cities, you know, like young people want to be around young people. We've, I'm pretty sure I've said that exact thing the last time we talked. So there's nothing new there. Did we, I think actually we might've talked about this. Yeah. This is from a month ago. So maybe we talked about this, but yeah, personally, I, you know, I think, you know, so my, my workspace is coming back and I'm vaccinated. I tell you I'm vaccinated. Yes. Yes.

Dave Jones: You're vaccinated. Yeah. Woo. Woo. Get the vaccine. I wouldn't celebrate too much, dude. It doesn't necessarily give you a lot of your freedoms back. So.

Chris Gammell: Oh yeah. I mean, that's yeah. Just be careful.

Chris Gammell: It's a shit sandwich. Don't worry. Don't worry about that. Yeah. But I'm very happy to have it all the same. Right. So, but yeah, I mean like the M hub is coming back and you know, so that's nice. And I just feel like there's people want to be around people. I don't know. Maybe, maybe less so in our listening.

Dave Jones: Well, yeah. I was going to say engineers are like, you know, I don't care if I don't see another human for six months, you know, it's like, yep. Yeah. Yep. It's just engineers. Yeah.

Chris Gammell: Sure. But those engineers need work. And I think a lot of that work comes from other people being around one another and generating economic activity. So.

Dave Jones: Right. Well, there's lots of economic activity in a toaster design.

Chris Gammell: Oh yeah. Oh, what about it? Chip of the week. Is this a chip you can actually get?

Dave Jones: No, zero in stock on DigiKey, unfortunately. Oh, that's 32 week lead time. So there you go. Here's the example, you know? So if you're, if you're in the toaster business, you're screwed because you can't get your PT 8A2514APE chip, which is an eight pin dip. None of that SMD rubbish, eight pin dip. Because, you know, like it's still cheaper to do dip single sided through hole than it is, you know, surface mount in a lot of cases. Depends on where you get it manufactured. Anyway, this is a toaster controller chip from Diodes Incorporated. And it's, it's kind of cool. It's just an eight pin tip chip and it's got a bagel pin. I just like, I just want it because it's got a bagel pin. Just, it's great. I don't know what else to say about it. I just, I just think it's chip of the week.

Chris Gammell: I'm pretty sure we've actually talked about this chip on the show before.

Dave Jones: Oh, really?

Chris Gammell: Because I remember talking about a bagel pin.

Dave Jones: Oh, okay. Oh, good. Is it 25 days ago? Oh, okay. Right. Our list is.

Chris Gammell: No, no, no. I think actually a long time ago.

Dave Jones: Oh, really?

Chris Gammell: Yeah. I don't remember though. I just remember a bagel pin. Okay. Maybe it was just.

Dave Jones: Maybe it's just come around again.

Chris Gammell: On Twitter or something. But yeah, it is interesting. Somebody discovered it again. It's like super, super. Targeted. Niche. Yeah, exactly. Application specific, right? Yeah.

Dave Jones: Oh boy. Yep.

Chris Gammell: Speaking of targeted chips and people making them. So this is actually something I mentioned, meant to mention on the show the last time we recorded, but I completely gapped on it. But Matt Venn, past guest of the show here, he's doing the zero to ASIC course. So he's teaching people how to actually use like all the tools to target, make silicon for the shuttle runs that Tim Ansell and Mohamed Qasem has been on talking about. And so he's been doing this course and he's also doing a podcast where he's talking to all the people, not all the people, but a bunch of people that are actually making the chips that went into that first shuttle run of the chips. This is so if people don't know the Google had sponsored a shuttle run through Skywater, the Skywater PDK we've mentioned on the show before as well. So it's 130 nanometer design node and basically Google sponsored people to go and submit chip designs. And so Matt has been documenting it.

Dave Jones: They had to be open source though, didn't they?

Chris Gammell: That's right. Yep. That's right. Yep. And yeah, it's really interesting kind of hearing why people are doing all these different things. You know, there's some familiar names, some new names. And so there's a zero to ASIC podcast though, and I've been listening to it and really, really enjoying it. It's great. So another, another podcast add to, to your listener, if you haven't already done it. And there's also videos. So cool. Yeah. Cool bananas. Highly recommended. Yeah. Dan Rodriguez. Oh, this is, oh, this is actually, yeah. So Dan Rodriguez is the one this past week. I was watching his thing. And at the end of the show, that episode rather, they mentioned the, I was going to almost put it on our list, but I figured I should just mention it. They mentioned the 2016 talk called the ultimate, the ultimate Game Boy talk or something like that. It's an hour long presentation from 2016 chaos communication Congress. And oh my God, it is amazing. It's like a deep dive into how everything works in a Game Boy, like all the sprite generation, all that stuff. Yeah. Oh man. Highly recommended. I still have to finish it, but like, just like, you know, just eight bit graphics and all that stuff. I just never really understood. And it's just such a great, a great summary of that sort of thing. So we'll have lots of links in the, in the show notes for this stuff. Yep. You can tell I've had a lot of time to watch videos. Can you, can you tell us a lot of, a lot of slow, tired days in my, in my past month.

Dave Jones: Yep. One of the most amazing videos comes from somebody I've never heard of, uh, TK. It's, it's hand drawing a risk five CPU. And sorry, I forget if it's risk five or risk V it's risk five.

Chris Gammell: You started right this time. Yeah. So that's good. Right. Risk five. You always, you always ask.

Dave Jones: Yep. I always think, cause I don't. You're getting there. I just don't care. You're getting there. Hey, you're getting there. It's good. Very good. Risk five CPU and playing. It's somebody hand drawing a risk five schematic, right. An actual risk five processor. And then it magically goes from paper to CAD and, and then it, and then they simulate it and, and then, and then they play a thing on it. They actually run code on it. It really is. Jeez. It's like, yeah. Hats off. It's amazing. Yeah. Absolutely incredible. And it just, yeah. I mean, it simulates processor.

Chris Gammell: And like seeing all the, I mean, the fact that it's actually like logic level, logic gate level, I expected to be more like, when I first saw the title, I was like, oh, it's going to be like high level blocks, but no, it's, it's down. It's down in the gates.

Dave Jones: It's got individual gates. It's got muxes. I don't know how they're actually doing it. Well, how would you feed it? I don't know. You've got to feed in the pixel data. It's, you know, it, yeah. I, yeah, I don't know how they're feeding in the pixel data that allows it to, yeah. Cause it's doing all this animated, it's doing an animated display thing. Um, so the trick is like feeding in the data and stuff like that. So it's not as advanced as you might think it is, I guess, maybe, but it's a bloody impressive. Jeez. I just, I'm blown away. Yeah.

Chris Gammell: Yeah. I mean, I don't even know where to start on this stuff. Like, uh, you know, like, uh, chip architecture and things like that. I just, I'm, no, there's nothing going on up top for me.

Dave Jones: Well, well, this was all like when you, when you learn digital electronics, all these basic building blocks of what you learn, you need, you learn your ALUs and your, you know, your arithmetic logic units. And you learn your, you know, your adders and your subtractors and your, you know, you, you learn all these, uh, building blocks of stuff, you know? And, uh, yeah, I, I don't know if they still teach that these days, do they? I don't know. Any young whippersnapper students out there learning digital electronics 101, do they still teach you, you know, ALUs and whatnot?

Chris Gammell: I think that, I mean, I think they're doing logic level and they're doing Carnot maps and all of that kind of stuff. But I feel like that's a, that's a big leap up to then piecing those together and building, even just building an ALU from a bunch of discrete logic elements is like, uh, is something special. So I think I mentioned on the show before, but there is a course at the university of Michigan that one of my friends took that is basically in this, in a semester, you have to go and design a fully custom, uh, processor all the way from scratch. Right. Okay. Cool. It's like, yeah, yeah. Yeah. I definitely mentioned on the show before, because it's like, that's the course. If you'd pass that course, you're going to get a job somewhere. Oh yeah.

Dave Jones: That's a specialized course. That's not something you do in a general degree.

Chris Gammell: It's a computer engineering. Yeah. It's like, it's like the advanced, the most advanced version. Of course.

Dave Jones: If you're doing computer engineering, that's different to just a Joe blogs, you know, engineering degree.

Chris Gammell: Yeah. Right. Right. You're only going to do that if you're really into it. Yeah. Yeah. Exactly.

Dave Jones: Cool. Uh, that's yeah. That's just, yeah. It's just great. Just, you know, the amount of effort to hand draw that and then, uh, yep. Enter it in and then get it working. I'm sure it's, you know, like the, the five minute video actually belies the number of hours. Um, it's involved in this. It's, uh, yeah, it's, it's huge. Yeah.

Chris Gammell: There's other videos we have on here too. So like in terms of like computer architecture and kind of following all this stuff. So obviously we've had Ben on the show before Ben eaters on the show, you know, talking about this sort of stuff he's building up from 65 or two. This is basically taking down another level.

Dave Jones: He builds the breadboard ones. Yep.

Chris Gammell: Yeah. Right. And so this is taking down another level and like doing it from there. There's another video from January of 2020 of this, this, of last year rather, uh, where there's a relay computer. And that's another good one where it basically, it's like, it's basically like at the Altair 80, 80 level where they're building out things from a, you know, just watching logic operations happening. And it's super possible to do this. Right. So the person who did it on paper, um, that's one way to do it. Obviously you could do it with the relay method and watching how, how, you know, bits propagate through. I there's, it's really interesting to see this stuff. I think then building it all the way back up. So then Matt and the zero to ASIC course, right. That's basically enough saying, let's do it from the, the software level and define it at the software level, or even going above that. When like Michael Gieldo is on the show talking about some of the, uh, the simulation things happening with re node, I believe. Yeah. So there's just like all these different layers and like where people want to specialize and, and learn these things. It's kind of like how much, how much under the hood do you already understand? Or do you want to understand? I feel like if someone wanted to submit something for a future, like Skywater PDK or Skywater shuttle run, you know, you probably want to, you're probably going to be doing that at the level of Verilog. You're going to be using some of the design tools. Of course. If you don't understand the stuff underneath, you might end up overdoing it where, you know, you might have more logic in a certain area. You wouldn't be able to check your work underneath, you know, it's kind of like. Yeah. Yeah. Right. The equivalent being like, like, can you read the disassembled code?

Dave Jones: Yes, that's right.

Chris Gammell: And, uh, understand like whether it's efficient or not. And that might have big implication at, you know, if it's a, you know, if you're writing C code and it generates a crazy amount of assembly or machine code after that, it might not be a big deal until you're at the end of, you know, like you're, you're, you're filling your, your chips filled up 96%.

Dave Jones: You're trying to optimize it. Yeah. Yeah.

Chris Gammell: Yeah, exactly. And then, so same thing happens with, you know, the, the skywater shuttle runs, right? You're, you're going to have a certain amount of space there. If you get to the edge, you're probably like, oh, I have to make this more efficient. That's when you really need to understand the stuff underneath. And so it's kind of like. Come on, Chris.

Dave Jones: Everyone knows you just buy a bigger memory chip, bigger memory process.

Chris Gammell: Well, when you only have a certain size, I mean, this is, this is the rubber hitting the road on that side of things, I think, you know?

Dave Jones: Oh boy. Yep.

Chris Gammell: Yeah. It's this crazy time we're living in that this is available though. Like the fact that people can just go and build ASICs and that's, that's great.

Dave Jones: You want to know what defines crazy times that we're living in.

Chris Gammell: What is that Dave?

Dave Jones: When your LG dryer, i.e. the dryer that dries your clothes, connect, not only connects to the internet, but uses one gig of data a day.

Chris Gammell: Yeah. Like. That is, that is, that is some software written by a JavaScript engineer.

Dave Jones: I, I, I almost can't.

Chris Gammell: We're logging everything.

Dave Jones: Yeah. I almost can't believe that's real. Like some, like the person who posted this on R slash R smart home, they postulated that have, do they have a compromised device? I.e. is it mining Bitcoin in the background or something, something like that. And I, you know, you would have to suspect that because you, you would, as an engineer, you would think, how could you possibly design a dryer like that, that needed one gig of bandwidth a day? Like, like surely you would know this during the development and you'd go, Hey, this is not, this is not right. You know, come on, this is bullshit. Right. So.

Chris Gammell: Yeah. No. I mean, I think, I mean, if you, but if you had some like crazy logging, like still turned on. Yeah. Right. Say your firmware version, I just had logging every, I don't know, 10 milliseconds, 10 microseconds, whatever. Like it's, I think it's pretty plausible that you could get up there pretty fast.

Dave Jones: Well, yeah, I guess.

Chris Gammell: How much data you're sending back.

Dave Jones: Well, yeah. Like it doesn't. Well, I was sending video, I guess. Back when I had a, you know, a data cap on my internet connection here, I could not leave even a low resolution 320 by 200 webcam running 24 seven here. So that people, you know, I used to, you used to be able to log in and see me working in the lab for whatever reason I couldn't do that because I would use up my entire monthly data capping like a day. If I actually streamed, you know, streamed at 24 seven or two days or something, it was just, and that was only low resolution, let alone high resolution. So just stuff adds up if you're doing it all that bit like your dryer, presumably only logs this data when it's actually operational, which is like what an hour and an hour a day at most. It's just not like it's. Oh, sure.

Chris Gammell: Yeah. Yeah.

Dave Jones: I guess. I mean, it's just, it's insane. Like.

Chris Gammell: Yeah.

Dave Jones: Yeah. I just like.

Chris Gammell: I mean, so, okay. So let's, let's do a little back of the, I mean, I'm going to add a calculator in front of me. It's not back of the napkin, right?

Dave Jones: So you don't have a calculator in front of you. Christopher J. Gamble.

Chris Gammell: I do have a calculator in front of me. I do have a calculator in front of me.

Dave Jones: No, no, no. None of that Windows bullshit. Real physical calculator.

Chris Gammell: No, I, it's, it's a, it's a window. It's a Windows bullshit. Yeah. Nah.

Dave Jones: See? Nah, nah, fail, fail.

Chris Gammell: I don't do calculators. My last, my last calculator was a TI-83.

Dave Jones: Nah, nah, I'm sorry. You're going to have to hand in, hand back your engineer's ring. That's right.

Chris Gammell: Nah. Engineer card. Yeah, that's right. But that would be 10 or 11 megabytes. You know, some of my math's probably off here, but that'd be about 11 megabytes a second. Right? So.

Dave Jones: I, I, I said, yeah, it's, it's giga B, it's capital B. So it's one gigabyte of traffic. Right. So if you've got one gig, one times 10 to the power of nine gigabytes. Right. Right. And divided by 24. Divided by 24 hours. Divided by 60. Divided by 60. Yep. Divided by. That's minutes. 60 again.

Chris Gammell: 60.

Dave Jones: Uh, second. Second. Right.

Chris Gammell: Yeah. That's, yeah. That's, that's 11k. 11 and a half megabytes. Right. No, megabytes. No, I get it. Oh, sorry. Oh, kilobytes. You're right. Yeah, you're right. So 11. Oh, sorry. No.

Dave Jones: 11,000. Yeah. Bytes. 11,000 bytes per second.

Chris Gammell: Yeah. Let's just stay in bytes. That's a good idea. Yeah. It would be kilobytes. You're right.

Dave Jones: It's 11 kilobytes per second. That's actually not a lot. But as I said, this doesn't run 24 seven.

Chris Gammell: Right. That's a good point. But, but if they had logging turned on. Right. So if, so imagine they didn't turn off the debug flag in their compiler. Right. We've got a segue here. Yes.

Dave Jones: I believe.

Chris Gammell: Yeah. Yeah. And, and like then, and so then, okay, 11 K per second, but then say you're doing it every 10 milliseconds. It's very verbose.

Dave Jones: You know, if you've got lots of verbose logs. That's now a thousand bytes. Yeah. Yeah. Yeah.

Chris Gammell: But yeah. But I mean, if you're writing, you know, say you're writing log messages that could very easily get it. And human, human readable log messages. So there's strings. Right. Right. Right. Right. Right. So a hundred, a hundred bytes per 10 milliseconds. Like that's, I've done that, you know, that's, that's easy to do. Yeah. And it adds up though. Right. I mean, if you're logging it back to a server, this is kind of crazy.

Dave Jones: This is not the Voyager 2 space probe where every bit matters, where they pack every bit and you have to, the amount of decoding you have to do on the other end to deconstruct each individual bit is, you know, incredible. Now it's just bandwidth is infinite. Right. So you send out text strings.

Chris Gammell: Yeah, exactly. I ask you. Yeah. Of course you're going to send Esky. Yeah. It's like. Yeah. Yeah. Yeah. Is there, there a segue to a, whether or not you do print F debugging. Is that the. That is the segue.

Dave Jones: I believe I haven't looked at the segue, but I believe there's a segue for that because even, even 20 years ago, I was putting in text, you know, debug log outputs. Cause it's just easy. Just hook up a serial terminal and it's like, there's your text, right? You can read it. You don't have to remember that. Oh, this bit means this. I mean, that's just dumb. Right. Cause then you'd have to get another step to process it. So you just turn on your debugger and your code spits out the message. Oh, it's, you know, sample one taken or something, you know, like.

Chris Gammell: Mm-hmm.

Dave Jones: So it's just easier, but you wouldn't necessarily put that in production code, but I've, I've, I've taken that sort of life. I've put that in during development taken out in production because, well, it's just sacrilege, you know?

Chris Gammell: Right. Yeah. So I, I, this is kind of written from a higher level language kind of thing. I mean, I think, I think that's the person who wrote this, the printf debugging thing. I actually think that it's a big switch to when you do step debugging or when you're doing, you know, you're running a debugger, you're, you got a J link plugged in that sort of thing. I feel like that's a, that's graduated. It's like important for my, from my perspective, but I, it's, it's going to be a mix of all of them. Right. I think that having, having a serial terminal spitting out stuff, as long as it's not like super timing critical, having LEDs available, having a printf out to a, out to a serial terminal, like that's all really important.

Dave Jones: But, but is it worth a couple of K because, but people who don't know, I've probably talked about this before in, yeah, on a, on a microcontroller where kilobyte, you know, you have kilobytes of memory, not megabytes, right? You know, it's measuring killer. If you're really luxurious, you might have 128 kilobytes or 256 or something, but we're still talking kilobytes.

Chris Gammell: I was going to say, right. Dave's, Dave's working with the, with older parts here.

Dave Jones: Right. No.

Chris Gammell: It's cheaper than it used to be. That's for sure. But you're right. No, the 128, 256, you're right. Yeah. I think that's right.

Dave Jones: Right. I'm talking about your eight bit micros, right? I'm talking about your, your AVRs and your picks and your whatnots, right? Your 8051s and whatnots, right? We're talking tens of kilobytes, sometimes under. And just the act of linking in the printf code adds a kilobyte or two to your, you know, I, I, I wouldn't be the first person to write my own printf routine because I didn't have enough space available. Right. I couldn't link in printf because I couldn't afford the one or two kilobytes that link it because when, when you just using your printf once, just once in your code, it means you've got to link in the full printf library. Right. And there are optimized compilers that kind of, you know, I know, I know about all that sort of jazz, but you know, but still, yeah, it, it can take up a couple of kilobytes just because you added one line in your code.

Chris Gammell: Yeah. Totally. I actually, I have a call back to earlier with sourcing stuff. So the parts that I had on hand, I had a different code of part, right? So it was actually half the memory, half the RAM, half the flash, half the RAM. That's what I had available to then send, you know, like I said, I didn't, I couldn't get the parts anywhere and I needed to send in a part, send in, there's an NRF 52 that I was using. Right. So now I'm, my code base now has to shrink in half. Right. And, and Nordic does not have a small Bluetooth thingy. Uh, you know, the binary blob is huge. Oh yeah. So basically. Yeah. And so I'm doing that now and I'm going through and I'm just turning off everything just to try and squeeze stuff down. Yeah. Yeah. Yeah. Been there. Yeah. All the print apps, there's multiple ones in there. There's different libraries. There's like all of their libraries, which are super, you know, the. They're bloated.

Dave Jones: Come on.

Chris Gammell: They're bloated. Yeah. I just say nice to written. And I'm like, they're not nicely written. They're, uh, they're thorough. They are all inclusive.

Dave Jones: They have, it's not, I'm, I'm, I'm not bloated. I'm thorough. That's right. Yeah. Right.

Chris Gammell: And yeah. And so they've got tons of options in there. And so it's like, I don't, you know, so it's like either I go write my own, which is, you know, it's got its own pit of vipers, but, uh, uh, it's like, yeah, it's, it's tough to, to do this stuff. So you have to decide like, do I need that printout debugging or not? Or do I, do I just, you know, rip everything out? Maybe I have like a pre-processor thing that I don't include it or I do include it or whatever. So yeah. In that case, I specifically had to then go and rip everything out because I couldn't fit it because I had a smaller part than I thought I would. Yeah. Because of sourcing problems.

Dave Jones: Yep. Hmm. Yeah. Yeah. Those were the days. No one cares anymore. You know, everyone's using their bloody, you know, STR micros with half a megabit of bloody megabyte of memory in them.

Chris Gammell: Well, unless they can only get the 256 K and then.

Dave Jones: Oh, well shock horror because you can, you know, cause your component shortage, gosh, darn it. That's right.

Chris Gammell: I mean, that's just, I mean, that's exactly what just happened. Yeah. That's what I just described. Yeah. I mean, like, yeah, it, it sucked. Yeah. So I don't know.

Dave Jones: That is one of the advantages of choosing a micro that has tons of variants like that, because like a lot of the times the footprints will be completely compatible and you won't get caught short because, okay, look, I've got to spend a bit more money, but Hey, I can buy the 512 meg version in stock where I only need the 64 K, but you know, heck, you know, it's like, it's the same pin out everything, everything else is the same. And I just have to, you know, tweak a line in my code to like, or a compiler switch or whatever. Or resistor strapping. Yeah. Or, you know, yeah, yeah. It's, it's nothing burger. And you can continue to use your part. So yeah, there's advantages to that. Yeah.

Chris Gammell: Yeah. I mean, it depends on different, different parts that have different number of variants. Some of them have, I feel like the ones that have the most variants are the ones where they're just turning things off with like laser.

Dave Jones: Yep.

Chris Gammell: What is it called? Like the blown fuses at the end of the process. Right.

Dave Jones: Yep. Although I've been bitten by this before because I used like a, I think it was like a 16 K variant or something and it worked fine. And I decided, Oh, just, just for production, I'll go to like the 64 K variant to leave me plenty of room or whatever. Right. Oh yeah. So I go to the 64 K variant didn't work because it, cause the 64 K variant had silicon bugs that the 16 K variant didn't, even though it was an identical chip, identical pin out, right. It's just a different slash number at the end or whatever. It was a physically different, you know, silicon in some way. And, and it had silicon bugs that the other one didn't.

Chris Gammell: Oh wow. So somewhere in that 48 K extra memory, the, that's where the problem lurked.

Dave Jones: I don't know. Somebody goofed up something, you know, when they actually compile. Or maybe like an address line, one of the bits, like the address line or something. I don't know, but yeah, that can happen. That's crazy. Like that can ruin your day. Yeah. So.

Chris Gammell: Yeah.

Dave Jones: Yep.

Chris Gammell: Yeah. So when, when Sammy, when Sammy from FNX was on, he was talking about the different FPGAs that they make. It was interesting looking at the masks that they have because it like, I had never really like looked at, they had like a chart that I think I linked, but they had a chart that showed kind of like the different part variants, all that other stuff. Like, you know, the different BGAs you can get and the different sizes, whatever. But it actually, it, when you zoom, when you zoomed out, it actually showed like, oh no, there's actually like three different masks that they bought. So there was like three discrete chips that they were making at the fab. And then you could choose within there. And I never really thought about that before. Like why, why wasn't it, you know, so like if I go to a, an NXP or something, why aren't they all just the same part? And it's because that same thing, there's different actual mess sets that are, maybe it's just, maybe it's not the entire mess that maybe it's just part of, maybe it's the metalization layer is different on some of the parts, but they are different chips. And so the customization or the, the part number variants can only be as much stuff that's actually in the silicon. I'm not sure if anything I'm saying is making sense, but, but the, you need to be able to actually have the silicon under the hood that you then can, you know, blow a fuse and enable or disable some of it. Yep. So yeah, I never really seen it presented like that before though, because I think, I think other chip companies kind of hide that stuff. Got it. So yeah. It's interesting to hear about that.

Dave Jones: There you go. I think we're way over, dude.

Chris Gammell: Yeah. We wouldn't know. Cause we had a, yeah, we had a few goof ups.

Dave Jones: Yep. Zen, Zencaster spat the dummy on us a couple of times.

Chris Gammell: Yeah.

Dave Jones: At your end, not mine. Nothing wrong with Australia. Yeah.

Chris Gammell: Yep. Yeah. Yeah. Australia has perfect internet.

Dave Jones: Yeah. Yeah. Anyway. Yeah. I have to try and figure out how to get internet into my new storage unit, which I got. Oh yeah. So yeah, that's a bit of news. Yeah. Breaking news. Got a, got a storage unit. Did a video on that. Mm-hmm. And yeah, I got to figure out the most, well, cost effective and easiest way to get internet down through five stories of concrete down to my storage unit. So yeah.

Chris Gammell: Sneaker net. Sneaker net is the answer. Sneaker net.

Dave Jones: Anyway, there's a thread on the EV blog forum for that. If you want to, if you're want to throw your suggestions, people said wireless, get really powerful wireless transmitters. I'm going to, five floors of concrete, you know, kind of down that. Stick it on the outside of the building. Not quite.

Chris Gammell: Oh, actually, one thing I did want to mention before the end of the show is I have a robot bassinet. I didn't, I meant to mention this earlier.

Dave Jones: You, yep. You told me, you told me about this and it rocks the baby back and forth and you're about to enter the, yep. You're about to enter the baby market, aren't you? I can, I can see the cogs turning, dude. I can see the cogs turning that you're going to. Oh, you mean as a, as, as, as in a product development.

Chris Gammell: I think I mentioned you when you and I talked about this, it would scare the hell out of me just because of like the product liability. Yeah, liability side of things. Yeah, yeah, yeah. Yeah. Yeah. But it's very interesting. Just like, I think it's one of the first, I mean, it's, it's like a high-end thing. So it's called a snoo and I rented because I can't afford to buy it. Uh, it's cause it's expensive. Cause it's, uh, it's a motor. It's a motor in the bottom of a bassinet. I know, but it's co-designed by Eve Behar, Eve Behar, Eve, one of the Apple designer people. And so it's like super fancy, uh, design stuff.

Dave Jones: It's a motor in the bottom of a bassinet folks.

Chris Gammell: I mean, yeah, yeah, that's right. But also, no, it's also got sound detection and then an app. Of course it's got a fricking app. It's, it's fantastic. Let me tell you, I love the app.

Dave Jones: And, and, and how much data is it uploading per day?

Chris Gammell: That's a good question. I don't know someone should reverse engineering, but it, uh, it is awesome. And it's, it's some electronics that I didn't expect to have in my life. And now I do. So.

Dave Jones: Unbelievable. I don't know. What's the world come and do. Yep.

Chris Gammell: Robots.

Dave Jones: Yep. Oh boy. Yep. We're, we're, we're still nowhere near, you know, robots and AI and all that sort of thing. But while, while, while you're away, Chris, Elon must finally admit it. That, yeah. The self-driving thing. Yeah. We're nowhere near the AI, uh, required to actually do it. You know? Yeah.

Chris Gammell: I think he owes you like 20 bucks or something.

Dave Jones: Yep. Yep. He owes me. Bastard. Yeah. Duh. Anyway, that's it. Okay. Welcome to the fatherhood. Thanks.

Chris Gammell: It's, uh, it's quite a club, quite a club.

Dave Jones: And it just gets easier from here. Trust me. Yeah.

Dave Jones: That's right. Yeah. That's right.

Chris Gammell: Yeah. All right. I'll be back and asking you more, uh, more, more tips next week.

Dave Jones: Catch you next time. Thank you.