Latest / Oxide and Friends / Framework Computer with Nirav Patel

Transcript

Framework Computer with Nirav Patel · Oxide and Friends

• 0:00
  Speaker 1: Alright. Adam, how is my millennial podcast quality?
• 0:03
  Speaker 2: You got any millennial would be proud of you.
• 0:06
  Speaker 1: Any millennial on that. Actually, that's actually, no millennial would be proud of me. My kids would just roast me. Only only my fellow xers. That's the only place I can find out.
• 0:15
  I was so I I and this is just a bridge too far, and I apologize. And I I know I can get this working if I have a little more time. And I was really just trying to get my audio to work through my terrific new framework laptop.
• 0:27
  Speaker 3: Oh, no. No. But it's
• 0:29
  Speaker 1: not it's not framework's file. Is this is so far ahead of where we are with the Dell. I cannot cannot even begin to tell you. This is basically gonna be a therapy session for me where I I unload on this goddamn Dell that has more or less ruined my life. So, no, this framework's already way, way ahead, and I know I can get this working.
• 0:47
  I just I just just just ran out of time. So I'm sorry about that.
• 0:52
  Speaker 3: Glad to hear that. We can make this a live debug session too if you wanna go.
• 0:55
  Speaker 1: That that's right. I went to the first. Head. No. You I mean, I literally support.
• 1:00
  I I literally just got this. So I actually Adam, did did you describe our story with respect to to framework?
• 1:08
  Speaker 2: No. Let's, let's, like, you know, start at the top or whatever.
• 1:11
  Speaker 1: Yeah. Start at the top.
• 1:12
  Speaker 2: Whatever that whatever that story is sounds good.
• 1:15
  Speaker 1: Yeah. Well, I feel like I I mean, what what what's, do you wanna intro, Nirav? Have you already done that? Or are we No.
• 1:22
  Speaker 2: Well, no. I mean, I was waiting I mean, not to stand on ceremony, but I was kinda waiting on you.
• 1:26
  Speaker 1: That makes sense. Waiting on me to debug my laptop issues. Yeah. Well, Nirav, welcome. It is great to to have you here.
• 1:33
  Speaker 3: Nice to
• 1:34
  Speaker 1: be here. We are, really big fans of of framework. So Adam has been a Adam, you've been a you've been, like, an original Stan, I would say, of framework. You I think you were I
• 1:45
  Speaker 2: think that's fair to say. So our colleague, Arjen, I guess you guys work together in Oculus.
• 1:49
  Speaker 1: Is that
• 1:50
  Speaker 2: right, Nirav? Yeah. Yeah. So right when I I think right when you were announcing a product, Arion sent it over to me, and I'm a little embarrassed that I impulse bought a laptop, but that's sort of what happened. I was so excited about what you guys were doing that I put my name on the list.
• 2:09
  And, you know, it took a little while to get here, but it was awesome. And I put it together with my, with my teenage son. We had a great time. And then, I pretty much never saw it again. Like, he absconded with it almost immediately.
• 2:21
  And he has reinstalled Linux on it, like, 15,000 times, and loves the thing. And and I get to use it every once in a while. But, yeah, I'm a huge fan of what you guys are building. Like a a, you know, from scratch, I think, implementation, consumer electronics, but all serviceable and, you know, you can poke around inside, build it yourself. It was awesome.
• 2:45
  An awesome experience building.
• 2:47
  Speaker 3: I love to hear that. I love the building it with your kids story as well.
• 2:52
  Speaker 1: Oh, that's nice.
• 2:53
  Speaker 3: Yeah. Just the depressing world we would all live in if every child was using a MacBook and none of you know, no one was growing up with that experience of taking apart their parents' computers and hoping this will work afterwards.
• 3:07
  Speaker 1: Yeah. I mean and I wish I could I I I could give you another heartwarming story of assembling a framework laptop with my own children, but sadly, that that's, well, Adam knows my kids. So that's my my my kids are much more into the destroying.
• 3:20
  Speaker 3: They Well,
• 3:20
  Speaker 2: as you when you when you unwrapped it, Brian, you were excited to do it with with your kids, but I guess that might have been a one-sided
• 3:28
  Speaker 1: excitement. I I could not. You know, my daughter was out. I would have been able to I I think, she would have been into I I've got a an 11 year old daughter. She would have been into it.
• 3:38
  The 19 year old and the 16 year old. If I could have somehow told them that, like, if we I have changed the Wi Fi password and the only way to get the new Wi Fi password is to assemble this laptop with me, I think they still would have been like, no. No. No. We're just gonna torture you.
• 3:50
  You're gonna give us the Wi Fi password. I just I don't know. I think they still wouldn't have gone for it. But, no. It was, so and and, Adam, you had ordered a a framework and, actually, perhaps unbeknownst to you, I had actually also ordered a framework, at the, so, Nirav, you, one of your engineers gave a presentation at the open source firmware conference.
• 4:11
  Yes.
• 4:11
  Speaker 3: Daniel.
• 4:12
  Speaker 1: And, which is awesome, and on the some of the Rust based firmware that I think that you have in development, and I I swear that whole conference, like, went to the website and started ordering 1, started configuring 1. And I was trying to figure out if I wanted the the the 13 or the 16, and I'm in a I'm in a very bad relationship. Wasn't a bad relationship, I'd like to say, with a, I had a so, Dron, if you might may I give you my own personal laptop journey for a moment? Yeah. There yeah.
• 4:40
  Sorry. I I it it I it ends with the, the good news of, my framework my new framework laptop. But the I had a so I was a little embarrassed actually that when we started the company but when did you all start at framework? It must have been 20 20 maybe?
• 4:55
  Speaker 3: January 2020. That's right.
• 4:57
  Speaker 1: That makes me feel so much better because we started the company in 2019, and I'm like, I need to buy a laptop. I'm gonna buy a Dell. I'm, like, I'm sitting here roasting Dell in this pitch deck, and I'm basically gonna sit here and buy a Dell, which I felt bad about. So the and and because you did I, I think I also and I looked system 76 also, I kind of looked at I I they were also at the Oak Shore Conference, but, I would have loved to have not gotten a Dell. But I got a Dell and then kind of even worse, like, it was actually the next PS 13 and it was actually okay.
• 5:32
  I I kinda liked it and felt dirty about it, and then the wheels started kinda falling off the thing. And, Adam, did you see when the the batteries got real bloated on that thing? Did you
• 5:41
  Speaker 2: Yeah. Yeah. That that was very exciting.
• 5:44
  Speaker 1: Yeah. I didn't quite realize how exciting that was. So I you know, one of the double e's is like, god, what's going on? Like, why is your laptop kinda, like, rocking on the desk? I'm like, I I don't know.
• 5:53
  It's a I, you know, I probably haven't treated it very well. And then someone else, like, is it looks like your keyboard is, like, distended a little bit. And finally, someone was like, I think you're, you need to look at your batteries. I think it was Nathaniel. Oh, no.
• 6:08
  It was like an intervention of, like, this thing is unsafe and, I think or maybe it was Matt Keter who was like, you need to go to the spicy pillows Reddit. Have you ever been to the spicy pillows Reddit, Adam?
• 6:21
  Speaker 2: Never.
• 6:22
  Speaker 1: Okay. So a spicy pillow is one of these batteries that has become distended. And it is a, and Rob, I'm sure you know, all the physics involved, but it is, it's a fire risk. And so we I went to the spicy pillows Reddit, which has all of these kind of bloated batteries, and I realized that the battery that I had was not just considered to be like a spicy pillow or an extra spicy pillow, but it was, like, ghost pepper spicy pillow. I literally, like, I did not the thing was outside on a cookie sheet that night because I was so scared it was gonna burn.
• 6:53
  I mean, I've been sleeping with this laptop more or less. Anyway
• 6:56
  Speaker 3: I'm your fellow.
• 6:58
  Speaker 1: The the literally, it's my bill. So the the and but, you know, the the but, you know, to kinda like the the kind of the the part of the core of framework, it's like, this laptop was basically functional, and little things started breaking that I wanted to replace. And in particular, one of the keys broke, and I'm like, you know, I'm gonna do, like, I'm gonna, like, do the right thing and actually go buy a replacement key. And I bought 6 different replacement keys for that thing that were supposed to be the exact model number, and they didn't. They were, like, slightly wrong.
• 7:29
  And, you know, you're not like, Dell is not supporting you trying to go do that. I mean, I replaced the battery. And so I'm like, I'm kinda, like, proud of myself for barely keeping this thing together, but I can't replace this key. My laptop was
• 7:41
  Speaker 2: I have to interject just because it was so delightful, the months when you couldn't type an e. I just really enjoyed those months.
• 7:47
  Speaker 1: Let's see. Okay. Look. It was never quite
• 7:49
  Speaker 3: that bad. I mean yeah. Like, you know, keyboards, what, like, 2 or 3% of the bomb value of that laptop. The battery is, you know, maybe 5% of the bomb value, but any one of those things breaking forces you to have to go to extreme measures if you can't repair it.
• 8:04
  Speaker 1: Go to extreme measures if you can't repair it. Yeah. So I and I was really, again, frustrated because I was trying to repair this so many different times. And, like, I'm I'm buying what should be the exact key, and it's just, like, subtly wrong. And, of course, like, Dell is doing the thing that always happens where you're getting, like, they will subtly change the bomb without changing any and we've obviously seen this on the server side where this stuff changes without changing the actual SKU or model number.
• 8:32
  So, actually, there are, like, 18 of these that were made and that have the same model number, and you're not gonna you're not finding the right part. So it's like, frustrating.
• 8:40
  Speaker 3: That's right. Yeah. And there's no incentive for cross compatibility because, really, it's only for their own internal service purposes.
• 8:46
  Speaker 1: Totally. And their view, of course, is like, well, if if, you know, if the t key breaks, you should just oh, also, Adam, it was the t key, not Oh, it
• 8:53
  Speaker 2: was the t key. It was the tiki because of all I don't remember. Right. Right. Because you would explain about the tiki and, like, tiki torch, tiki bar.
• 9:02
  What are we talking about?
• 9:04
  Speaker 1: Right. There's a lot of, like, threes company antics around here. I know that's referenced. I that I'm I can no longer make talk about, like Oh, David. Millennials spray blanket.
• 9:12
  But, like, people are just like the t like, a tiki lounge? No. Like, a tiki. Like, a tiki. I mean, no.
• 9:18
  At pause, pause, pause pause key, k e y. Like, oh, and you're t e. I'm like, yes. The yeah. It was it it it was it was bad.
• 9:27
  But then and then that laptop, like, this is kinda like this jalopy that is still making, like, I can still, you know, not totally hate my life with. That thing gets stolen. And I'm like, what the thief what was actually stolen from me was actually a, like, working Udev rules. That is what those thieves stole from me, and it is priceless. Like, working udev rules and a working wifi config, and working audio.
• 9:53
  And so I bought a a replacement XPS 13, and it is so enragingly bad. I mean, that that product has just gone absolutely into the ditch, and everything about it is awful. First of all, Adam, you know, like, when you got, like, like, the escape key is hooked up to, like, razor blades and voltage? You know what I mean? When the the escape key is, like, they're clearly, like, okay.
• 10:18
  I get it. Like, you're you're an emacs person, whoever designed this keyboard. Like like a dog bites me whenever I hit escape. Like, okay. Okay.
• 10:28
  But it was and everything about that computer is awful, and the keyboard's awful, the touchpad is awful, the touchpad also, like, ceases up where you know, because they they wanna be able to use the touch pad as as the mouse, and the thing just like it's like, you know what? I'm no longer a mouse. I just like and it it kind of, like, freezes for seconds at a time. And the worst thing is, and I'm sorry, we're gonna be done with this therapy session really shortly. But my audio was working on that thing.
• 10:53
  And it then stopped working. And it was a real I mean, it's such a hassle because I you can see other people have the same problem. I can go Google my problem, and I you know, I'm going into d message, and I can see that they that the the the kernel driver is tossing, and I'm trying to update my kernel. I'm trying to do all the kinds of the things that you do, and I just cannot get the audio working again. And you get to the point where, like, I can spend the rest of my life getting this audio working, and it's, you know, everyone's Adam's roasting me.
• 11:23
  Everyone's roasting me on the I was just like, this is awful. Of course, it's like, it's very bad, like, you're a computer company, you can't get your audio working. I'm like, it's not so anyway, when do when the I saw the framework presentation, I'm like, I'm buying a framework right now. And then, Adam, you were like, hey. You know what?
• 11:40
  I've got this framework that, like, we will get this extra machine that I basically have.
• 11:44
  Speaker 2: Basically, I've gotten in line for the 13 and then got in line for the 16 inch, but forgotten to take myself out of line. So I was able to pass that off.
• 11:53
  Speaker 3: Wow. That worked.
• 11:54
  Speaker 1: Yeah. It did well, it definitely works. Then and then the 13 that I had ordered at o at OSFC arrived basically the same time. So we have got a we've got a 13 that's kicking around, an oxide there. So we
• 12:04
  Speaker 2: know we're
• 12:04
  Speaker 1: gonna get yeah. We're we're we're some other lucky oxide engineers gonna get into framework. Anyway Thanks. Ralph, I finally assembled this thing over the weekend, and, it was great. I absolutely loved it.
• 12:15
  It was so much fun, and I I I I I I'm sorry that I tried to fly too close to the sun and get the the fancy audio working on it. I know I can't. I just I just couldn't quite in time. But, you've got a really, great product here, and it was just exciting to go. I I and I I love so much about this.
• 12:35
  So I got a lot of questions for you.
• 12:37
  Speaker 3: Alright. Yeah.
• 12:38
  Speaker 1: One is so I love the, the the orientation around community, because I will tell you that one of the most frustrating things it's always frustrating when you feel like you spend a lot of money on something, and you've got a problem, and you're kind of being blamed for the problem that you've got. And I love the orientation around community and allowing people to, like, comment on their experience on your you know, as people kinda going through the steps to configure your laptop, they can see the experience of others. That it feels like community is a really important part of of the the origin story of framework, but I'd I'd love to hear what that origin story actually is.
• 13:15
  Speaker 3: Yeah. It's true. No. I mean, the whole philosophy is that it's your product, and you can do what you want with it. And so the community is just this incredibly powerful aspect of the whole thing.
• 13:25
  And really from the start, you know, the part of making this product modular and upgradeable and repairable was opening up enough of it that we could actually get the community developing their own compatible parts or coming up with their own fixes or you know coming up with their own recipes for things that we would just never be able to do on our own. And you know of course, you know, Adam and Brian coming from the software world, you know, you know that this is incredibly common in software to develop an ecosystem around a set of open source software and have the, you know, the power of the community become this multiplying factor, and it really ends up being pretty magical for everyone. And in hardware, that's mostly absent, and it's, you know, pretty unfortunate that it's mostly absent. And so with framework, I really looked at that and thought, like, you know, we can actually do something really cool here. And just, you know, concrete stuff.
• 14:12
  If you look at the expansion card system, so for folks who are not familiar, expansion cards are these little modules that you can slot in to choose the ports on your framework laptop. Right from the beginning, we published a bunch of CAD reference designs, specs up on GitHub. And then almost immediately, we saw people in the community building their own cards, and there were things that we had never even thought of, let alone, you know, have the the bandwidth to go and build ourselves. And really exciting stuff starts to happen as we build enough of an install base ourselves that these people creating their own modules can actually start to build businesses against that, which of course is a win for everyone. You know, it's a win for us as a company, of course, but really more importantly, it's a win for the community and the end users of the product who are getting this really great ecosystem built around the product that they're using.
• 15:00
  Speaker 1: That was No. So feel so great. Yeah. Sorry, Adam. Go ahead.
• 15:03
  Speaker 2: Oh, no. I just gotta say the the expansion card was one of the things that really got me hooked initially. Just so, representative of the different philosophy that you were taking. You know, opening things up, inviting developers in, inviting you know, I had no idea what might go into that slot, but the idea that you were open to it was such a breath of fresh air compared with, you know, all of these extremely closed ecosystems. I thought that was great.
• 15:29
  Speaker 3: Right. And it's just that's very obvious thing in retrospective. Like, you know, everyone's got different ports that they need. And if, you know, if you look at laptop reviews or community comments from, like, you know, 2,005 through 2019, like, half the reviews out there call out as a negative, not enough ports, or it doesn't have the right ports. And so for us, looking at building a new laptop from scratch and thinking about, like, okay.
• 15:53
  How do we set ourselves apart? One very obvious thing was, why don't we just let people choose the ports they want? We're never gonna be able to get it right on our own.
• 16:00
  Speaker 1: Yeah. That is all I love it. I and I love it because I think I've got I mean, I think many developers have got idiosyncratic port needs. One of the things, you know, developing a microcontroller software firmware, like, USB a ports are actually super convenient. So and laptops don't generally have them.
• 16:19
  So, generally, you're having to go through a hub if you wanna get any any USB a. So it's, I am for the 4 that I picked, I definitely one of mine was USBA. I am I am a little curious if I can ask you a technical question. The apparently, if the USB a, if the USB a expander is in the rear, it draws more power.
• 16:40
  Speaker 3: This is that's correct. Actually, unfortunately, this is actually an artifact of the retimers that are in, the AMD system specifically. So we use these retimers from a a start called CANDU that's based I think they're headquartered in the Bay Area. It's a really great part. Like, it's this really excellent, you know, USB 4 retimer that's, you know, it's pretty compact package, and it's relatively power efficient, but it makes the assumption that there's going to be a device on the other end, a USB 3 device on the other end that can negotiate going into the low power USB states, like the u 2 or the u u 3 states.
• 17:16
  But if you have, you know, the termination resistor for a USB a port there, but no device telling it, hey, I'm ready to go into u 2 or u 3, the retimer just stays and waits for a device to show up, which is pretty unfortunate. And so we do basically just share some guidance for folks on, on our AMD systems to not put empty USB a expansion cards in those rear ports because it'll make the retimers not go fully to sleep and stay in a higher power state.
• 17:44
  Speaker 1: And that's so then the retimers are not on those front ports. The front ports must not don't don't have the issue?
• 17:49
  Speaker 3: Yeah. So that's right. Right. Or it has a different retimer that doesn't have that issue, that is not a USB 4 retimer, and then the front left, as well does not have that that retimer.
• 18:02
  Speaker 1: That's great. Well, I I mean, I I love that. I I mean, just just getting that transparency is such a and, you know, you you would spend, like, years of your life being, like, I think I finally figured out what's going on with my battery issue. Like, if I you if I plug something, anything into the USB a port, it will consume because, I mean, battery life is a huge problem. I mean, among the many, many, many problems, I have this Dell piece of junk.
• 18:35
  It it will be it will consume power when it should not be consuming power. And that was few I mean, it's it's an added engineering challenge, I'm sure, in a laptop.
• 18:45
  Speaker 3: Oh, for sure. Yeah. And this and it's not that, like, stuff like this is that uncommon. I mean, you're right that, like, if you look at pretty much any consumer electronics device, they're gonna have these weird quirks just because of the inherent complexity of the things that we're building. But most of the time, you will see, you know, the 1,000 post forum thread or, you know, you know, ancient Reddit topic where people are being gaslit into not knowing if there's a problem with their product.
• 19:10
  The thing that is unique about us is that we will just go ahead in our community thread. And if someone asks a question, we'll go in and just answer, like, hey. Here's, like, the reason it behaves this way.
• 19:18
  Speaker 2: I well, as you were doing as you were doing bring up even, I saw you sending out emails to folks who are waiting in line for these devices explaining the state of that, which I thought was amazing transparency along those lines and was very evocative of some of the same stuff that we were going through at Oxide about, you know, kind of following all of these these problems with different parts of the system and so forth. Yeah. I'd love to hear more about just the journey of of bringing up some of these components in your and some of these new systems and working with some of your partners and vendors on navigating through some of these hairy issues.
• 19:57
  Speaker 3: Yeah. Definitely. And just on the on the open s, which I, yeah, I love about oxide as well. It's this weird thing that I'm sure you've also seen that, like, you know, in companies who've been in the past, the default was always, well, you know, we should be careful about what we're sharing. We don't wanna, you know, send out something that might give people the wrong message or make people feel bad about the product or be afraid of partnering with us or be afraid of buying the thing.
• 20:18
  But in every instance in framework, and I'm sure in Oxide as well, in every instance that you go out there and just actually be open and honest and transparent about a thing that's happening or a thing that's gone wrong or a decision that you've made, people receive it positively. Like, they appreciate the transparency, they appreciate the understanding, and even more than just that specific instance of you sharing that, it builds long term trust. People Yeah. Basically start to give you the benefit of the doubt.
• 20:45
  Speaker 1: Amen. And that long term trust, I think, is so important, and it's tragic to me how many companies don't seem to value it because I think that trust is everything. I mean, if you if you trust someone, you're willing to endure difficult times with them. You know? And you you can hear that, like, hey.
• 21:04
  You know? Here's what's going on with the the, you know, this is why this batch has been delayed. It's like, hey. I'm I'm great. Thank you for letting me know.
• 21:12
  Like, here's what's going on exactly. Or, you know, here's what's here's what challenge we're having with or and giving people that full context, I think is so important and so appreciated.
• 21:25
  Speaker 3: It works. Everyone should do it. Highly recommend it.
• 21:28
  Speaker 1: Yeah. Absolutely. And I'm I'm sure there there there have been, like, a lot of of bumps along the way. Actually, one question I definitely have, and sorry, Adam. I didn't mean to short circuit.
• 21:38
  Any question you might have around, kinda navigating the ecosystem, but I am dying to ask about DDR 5 because the, this is actually the first DDR 5 machine I've I've had, which is exciting. I I loved your transparency such as one can be transparent because it's a very opaque process in the part about, like, look, it's gonna take a couple of minutes to boot this for the first time. And it was interesting to to read the community forums because we that's what we see with with our general based systems. They take a long time to train. Was that at all a to what degree was that a surprise?
• 22:13
  I I love, again, you're kinda being as transparent as you can be, with your own users about, like, this is something you gotta expect. It's only gonna happen the first time. Presumably, it's because you, will then encode, some starting parameters to really speed up training on on subsequent boots, I assume.
• 22:29
  Speaker 3: Yeah. That's just, I mean, it's part of the platform behavior for both Intel and AMD on their on their mobile processors. You know, normally for most computers, you probably wouldn't ever see it happen because they're fully assembled out of the factory. They go through memory training, and then they ship to you and you boot it, and it boots like it would normally. The thing that's unique for us, of course, is that most of our customers are buying the DIY edition, and that means that memory and storage aren't installed and OS isn't preloaded.
• 22:59
  And so that installation process, the out of box process, you're putting in memory, and so it has to train the first time around. And so we had to flag to people, like, hey. This this is gonna take a bit longer, and it's a bit unusual, and this is the the reason that it's doing that.
• 23:14
  Speaker 1: Yeah. Would, again, very helpful, and DDR 5 is so, it it just takes a long time because it's very what it's trying to do is very aggressive in in many ways. And, so it was just kinda interesting to see that and and fun to actually have DDR 5 in my hot little hands in terms of my own computer. So in terms of the, the attacking into some of the open source firmware stuff, with what's been your your take on that? Because I know, I mean, the the lowest level platform enablement, I think you you probably have still had to use some vendor specific code, I would assume.
• 23:49
  I know you got some insight I I think I've got an inside bias. Yeah. We
• 23:52
  Speaker 3: use that's right. We use inside so our embedded controller is actually forked off of Google's Chromium EC embedded controller firmware, which has actually been really great. You know, a lot of, I think a lot of you know, notebook brands basically license embedded controller firmware, which there really isn't a a strong reason to get proprietary embedded controller firmware. You know, there's nothing, you know, secret or magic in there. It's just good solid general purpose housekeeping, microcontroller firmware.
• 24:19
  And so we just took what Google developed, made it work with Windows. Obviously, you know, I'm trivializing a bit, but made it work with Windows, you know, added the things that we needed for our own, and then kept on that. So we've done that with, you know, 11th gen, 12th gen, 13th gen Intel. We did it again with AMD. We're gonna continue to use it in the future.
• 24:38
  But for BIOS, there's quite a lot that's, darker, more challenging, especially as it comes to doing, doing notebooks, and, you know, supporting Windows and Linux. And so basically, like, right from the start, we wanted to minimize the number of of miracles that we had to pull off to to ship a consumer product, especially one that's on fairly recent platforms, you know, for example, like 13th Gen Intel and also the 70 40 AMD that we just, started shipping. We're one of the earlier OEMs to actually be shipping either of those things. And so for us, you know, really just take taking a step back and zooming out a little bit. When we ship our products, the most important thing for one of the most important things for us is making sure it's not a trade off to get the repairability and upgradability and customization.
• 25:27
  That that, you know, we don't want someone to have to look at our product and think, like, oh, yeah, that's really great. I love the repairability there. I wanna get that product, but it's on last year's platform. I I understand. Don't wanna make that trade off.
• 25:39
  And so really critical for us was making sure that our products are performance competitive, form factor competitive, price competitive, and as much as, you know, we would love to be able to take, you know, right from the start and extended period of time and a massive investment to try to get to open source, you know, BIOS alternatives, you know, coreboot based or whatever else. It would have meant we were shipping on. We were, you know, probably always going to be shipping on platforms that are old. And that's not a trade off that we thought was the right one to make with our with our mission at the outset. That's not to say we won't get there.
• 26:15
  That's something that, obviously, we see that there is strong interest and demand for. We do wanna get there. But for us, for now, the priority is making sure that we can ship competitive products earlier.
• 26:26
  Speaker 1: Totally. I think and I I think it's the right trade off because I think it is, and you you're right to, want to have the repairability on top of a competitive laptop that that one does not feel like they're having to give up things for the the repairability or the transparency, which I could tell you as a framework user, like, you definitely don't feel like you're giving anything up. You feel like you've got, like I've got a laptop that is just better across the board and, like, oh, and and by the way, it's also repairable, and it's not and it's like, oh, great. But that is that's really terrific. You you said that most of your customers do the DIY edition.
• 27:00
  Obviously, they're not a DIY edition. I actually I I actually didn't realize that there was a somebody else's does does it for you edition.
• 27:07
  Speaker 3: Yeah. There there is actually a prebuilt version. We call it prebuilt. That's preassembled, has Windows 11 preinstalled, basically just ships the way a normal Windows laptop would. You take it out of the box and press the power button.
• 27:21
  There are people who buy that, and we understand why people would buy it, but the majority of our customers are going down the DIY path because, honestly, it's it's fun to build your own computer.
• 27:34
  Speaker 1: It is, and it's fun to actually, like, see these components. And I've I I definitely have got new appreciation for the bezel. I'd like to say I have underappreciated the delicacy of the bezel. Bezel, I will never underappreciate you ever again.
• 27:48
  Speaker 3: That's right. Yeah. So the bezel is magnetic attached and color customizable. It was actually definitely one of the harder things to accomplish from a mechanical and industrial design standpoint. But, really, it started with this idea of how do we make the display easy to replace and not have it be something that you have to take, you know, heat gun or heat pad and try to, you know, melt some adhesive to try to loosen the display and replace it, but actually make it something that you can just take some fasteners and unscrew them and swap out the display if in case you ever break yours or need to do an upgrade.
• 28:20
  Speaker 1: Okay. That's interesting. So that this is an example where the repairability really was intentioned with some other aspects of the system. Did that come up a lot where, like, okay. In order to make this thing repairable, we're gonna have to this other element gets more difficult.
• 28:34
  Speaker 3: Well, that's the entire discipline of the I would say. Yeah. I mean, certainly well, I guess, just broadly, I think most people would be surprised or actually, I think most designers and engineers, but also people who are not designers and engineers, be surprised by how little we had to trade off to get get that repairability in. Because, you know, when you when you hear repairable laptop, I think, or customizable or modular laptop, the the image that appears in your mind is something that's as as thick as a phone book and, you know, looks crazy and has, like, stuff hanging off a bit. And, of course, our products don't look like that, but they are still extremely modular.
• 29:11
  And so it's really just thinking through pretty carefully the trade offs that you're making and especially really thinking through the frequency of different types of repairs and thinking about, you know, based on how often it is that you'll need to replace or upgrade a main board or a battery or a display, how much time is it reasonable to require a user to have to go through to to do that, that replacement. And so we really think about, like, making sure that the most frequent replacement items are pretty fast to replace. But then things that aren't that frequent, maybe we take some trade offs where it takes, you know, 15 minutes instead of 2 minutes to replace. And by doing that, we can make it mechanically simpler or more robust.
• 29:54
  Speaker 1: Interesting. And so naive question, what are the components that are I mean, I know it's like as far as I'm concerned, it's the battery and the t key, but, in your experience, what are the things that need to be, replaced earliest?
• 30:07
  Speaker 3: Yeah. I think in terms of, like, upgrade behavior, people doing memory and storage upgrades often ends up being bring, something that happens pretty readily. Obviously, for us, we have a a large enthusiast space, a lot of people who are power users who always want latest and greatest performance, and so they're doing main board upgrades. But, you know, you're not upgrading your main board every month or every 2 months, you're upgrading it every 2 or 3 years. So if you have to spend 15 or 20 minutes doing that main board upgrade, that's a pretty reasonable trade off for something that happens that infrequently.
• 30:38
  Yeah, I guess if we think about like the expansion cards for example, if you have you know there's 4 slots on a framework laptop 13, but occasionally you might have other peripherals that you want to use and being able to swap those out pretty quickly without needing to pick up a tool or anything, that's something that made sense to be a, you know, 5 second swap instead.
• 30:57
  Speaker 1: Makes sense. And, Adam, have you physically had a framework 16 in your hands?
• 31:02
  Speaker 2: No. I Dharam, do they are they out in the wild, like, to to normal people?
• 31:08
  Speaker 3: No. We haven't shipped them to customers yet. We're actually almost there, and we actually just sent out an email last week where, we shared that we had started main board mass production, started main board SMT, and we're on the path to shortly start full system FATP final assembly test pack out.
• 31:27
  Speaker 1: Yes. It is only It It out of the cool box.
• 31:30
  Speaker 2: It oh, it's a really cool box. Nirav, can you talk about the the GPU and and how and, like, the integration there and kinda how you came up with those concepts. Yeah.
• 31:40
  Speaker 3: Yeah. So the holy grail for gaming or really other high performance notebooks is the ability to upgrade graphics. You know, if you're a gamer or you're doing CAD or you're doing simulations, the thing that is constraining you most of the time is going to be your GPU. And in desktops of course or workstations you can do those upgrades, and it's maybe even the most common upgrade behavior if you're a gamer to upgrade a GPU in a desktop. But if you are buying that fancy $2,000 gaming laptop in 5 years, it's going to be sad when you throw the latest games at it.
• 32:17
  And so we looked at it and thought, like, well, obviously if we're gonna build a high performance notebook that's aimed at gaming like we did with, Framework Laptop 16, of course, along with any other high performance use case, we better solve that problem. And so we looked at how everyone had solved it in the past. There have been several attempts, actually, a few that ended in class action lawsuits because they didn't go so well Oh, wow. And and thought, okay. How do we not end up with a class action lawsuit when we make our version of this?
• 32:43
  Speaker 1: It's honestly good.
• 32:46
  Speaker 3: Yeah. It's a good place to start when you're Yeah. When you're starting to Yeah. And it because they not get sued.
• 32:50
  Speaker 1: And this is because the products ended up being defective because it was too complicated. Right? Or
• 32:56
  Speaker 3: It was, I I think I have to look at the details of it, but it's basically false advertising of claiming that, the product is going to be have upgradeable graphics and then there never being upgradeable graphics.
• 33:07
  Speaker 1: Interesting. Yeah.
• 33:09
  Speaker 3: And so and so for us, it was really looking at that studying what went wrong, and the obvious problems were that the graphics cards followed you know either the old school MXM format or custom formats where they were cards or basically buried inside of the system, mounted to the main board, buried inside, and so they're constrained in x, y, and z. They're constrained thermally, and they're constrained in terms of power, of course, based on x y z envelope. And so when, you know, you design around that specific x y z and power and thermals. And then the next year, NVIDIA comes along and, you know, start to be a few months as NVIDIA tends to do before they launch something new. They finally share with you what it is that they're about to launch, and it turns out not to fit in that x y z, you're stuck and you get sued.
• 33:55
  And so for us, we looked at that like, well, we better make sure we design this in a way that we're not constrained. We're not constrained by anything that any of the GPU vendors are gonna throw at us when they move generation to generation. Even if it's just shifting stuff around, but also things like needing a little bit more x y or needing a little bit more z. And so what we did is we actually designed the back section of the framework laptop 16, basically, the the the back of the base section to be something that you can unscrew and slide out, and there is an internal interposer that we can put, about, you know, a little over a 100 watts of power over along with, 8 lane PCIe. And, basically because that module slides out and is, is flushed with the surface or exposed on the surface on the bottom and on the back of the computer, we can extend the depth of that module and we can extend the thickness of that module.
• 34:52
  And because it's also a module that slides out, we can basically decouple what happens inside of that thing with what happens inside of the rest of the laptop. And so we've given ourselves just an immense amount of design freedom to be able to make that expansion bay module and or graphics card that sits inside of it do whatever it needs to do as NVIDIA or AMD or others throw things at us in the future.
• 35:17
  Speaker 1: That's awesome. And was that modularity helpful for development as well in terms of of testing and validation?
• 35:23
  Speaker 3: Not really. No. It added quite a bit of complexity actually in part because we we actually do use a different supplier for the main board and for the graphics card, which is something that is pretty common, of course, in something like a desktop. It's not something you would normally ever have to deal with in a laptop, and so that's something that we had to navigate with our main ODM partner for the laptop and, the module maker for the graphics card, of course, along with AMD who we partnered with on, on the, the CPU and GPU for the framework laptop 16.
• 36:01
  Speaker 1: And could you I mean, especially when you're talking about consumer quantities, I mean, could you, like, speak a little bit about the actual, like, physical manufacturing of this stuff and working with your RDM partner? And how do you I mean and validation testing. How did how do you, there's a lot of validation to try to pull off.
• 36:20
  Speaker 3: Oh, absolutely. The great thing about one of the the nice simplifying things about what we're doing broadly as a company and framework is that we very deliberately go into mature product categories. And so for us, it's not that we have to treat an entire or teach an entire supply base how to build a laptop. They know how to build laptops. It's actually just teaching them the unique things about what our product is and and what is different about ours versus the things they've built before.
• 36:48
  And so as we started to build the framework laptop the first time around in 2020, we really actually went to every top tier supplier for all of the big notebook brands and pitch them on this thing that we're building, you know, both the the final assembler, you know, main ODM partner, but then also all the key modules like battery vendors and keyboard vendors and so on. And so we're able to take, you know, pretty substantial shortcuts by being able to leverage all of their existing knowledge, design knowledge, but also validation knowledge to build what ended up being just an an 18 month process from starting the company in January 2020 to shipping the first units out the door in July of 2021, of course, you know, during a a pretty challenging period of time as well.
• 37:35
  Speaker 1: Yeah. That's amazing that you were able to and and it was that because you were able to kinda leverage what these these kinda extent and how how much of those those designs are your own designs versus leveraging kinda Exton ODM designs?
• 37:48
  Speaker 3: Yeah. I mean, the design itself technically is is entirely ours. Like, the the the tooling is all ours. The industrial design is all ours. Like, we own all all of the design ultimately, but but it's still like that's kind of like an oversimplification where if you think about, you know, the the the schematic for modern x86 main board in a consumer product, that's like a 100 page schematic, and a lot of it is pretty much the same across every notebook.
• 38:16
  And so, you know, having a partner who's built, you know, of 50 other notebooks that use the same form, it means they they know the tips and tricks to be able to make, let's say, 50 of those 100 pages go really quickly, and be able to, you know, basically bring in that design knowledge, but then also the the corresponding validation knowledge that goes along with that.
• 38:37
  Speaker 1: So leveraging kind of their their own internal reference designs. They're kind of they're able to mute does permute designs to to kind of speed the the development and validation.
• 38:46
  Speaker 3: Yeah. That's right. And we work so CommPilot is our main manufacturing partner, who we've actually used now for every notebook that we've built built so far. And so, you know, Intel and AMD have their own, you know, customer validation boards, reference validation platform boards, and then compile internally builds their own internal reference boards or customer boards, and then basically we just take, you know, the next level down from that, which is building our own main board leveraging the learnings off of both of those.
• 39:14
  Speaker 1: Yeah. Interesting. And then are you because the, so, Adam, I actually saw I did see a framework 16 in the flash at the OSFC conference. The engineer yeah. And it's you're in for a treat.
• 39:24
  It was really good looking. It was it was Yeah. Fun to kind of and it was showing off some of the configurability of the keyboard and so on. And he was based out of Shenzhen. So it it sounds like you've got a lot of engineers actually in Shenzhen.
• 39:37
  Is that is that right, Ralph?
• 39:39
  Speaker 3: In Taiwan. Yeah. No. So, yeah, so, actually, Daniel and, we've got about, let's see, about 18 people now.
• 39:44
  Speaker 1: Oh, in Taipei? Yeah.
• 39:45
  Speaker 3: In Taipei. Yeah. So, actually, half of Oh, I
• 39:47
  Speaker 1: would say. Okay.
• 39:48
  Speaker 3: Yeah. We've got 2 offices. We've got got one here in San Francisco and one in Taipei. Actually, they're both that we're just before you started, I think we were chatting about this, but in both locations, we've got a cheapest office in the most expensive neighborhood, which has worked out really nicely for Grand Rapids. So, yeah, we're in a cool neighborhood in Pepe and a cool neighborhood in San a Scott.
• 40:07
  Speaker 1: Are we is that in Emeryville? Are we are we is is that a Are we talking about Emeryville there? Are we talking about the the Oxide office?
• 40:13
  Speaker 2: We are we are definitely the cheapest office in maybe not the most expensive neighborhood.
• 40:19
  Speaker 1: We are. Yeah. You know, I hadn't really thought it was that way, but we kinda are, really. Yeah. We are.
• 40:23
  Hey. Listen. Yeah. Better. Hey.
• 40:26
  I was in an office. It was the most expensive office in the worst neighborhood. Yeah.
• 40:28
  Speaker 3: That's the only thing. That's right. That's the opposite.
• 40:32
  Speaker 1: You you can go the the the wrong way. Okay. So there is service at at a time that you have. I'm sorry. Yeah.
• 40:36
  This
• 40:37
  Speaker 3: no. No. And this was really, really critical for us. And, actually, this is a big part of the reason we're able to ship in 18 months during, you know, global pandemic and lockdowns is that right from the start, we started building a team in Taiwan. And so, Nick, who's our lead industrial designer and head of industrial design is, is based in Taiwan.
• 40:58
  And, he was the 5th person into the team, and that was just incredibly, incredibly important. And so we started building an industrial design and engineering team and part of our operations team in Taiwan, which is excellent because basically every manufacturer in every part of the notebook supply chain is headquartered in Wendy kilometer radius from our our office there. And so, you know, no matter what we need, all the engineers are a short, metro ride away, basically. And the other really critical thing that we did very early on was we we have the benefit of starting our infrastructure from scratch. Every electronics brand in the world, if they could start their infrastructure from scratch today, would build as much of it as they could outside of China.
• 41:46
  You know, we all saw that trade tensions are only going to get worse. In 2020, of course, you know, they were already in a in a pretty challenging spot, And so we looked at and thought, like, well, we better start our infrastructure in a place where we know we're gonna not run into potential risks around, around trade issues. And so we actually do right from the beginning, we've done final assembly in Taiwan as well. And so the really convenient thing there is that our factory is, about an hour and a half drive or also about an hour by, high speed rail from our office. And so, like, engineers and operations folks can go from our office to the factory and go back within the same day, which is just incredibly useful.
• 42:29
  Speaker 1: That is huge. Yeah. We have found that that same phenomenon in Rochester, Minnesota. So the but the similarly, kind of having that ecosystem of suppliers and the ability to get I mean, boy, the ability to actually, like, walk the line and be on on-site in your manufacturing facility, boy, you really need it if you wanna develop something correctly, let alone fast. But you you really need to have that physical presence.
• 42:57
  Speaker 3: Absolutely.
• 42:59
  Speaker 1: Yes. And and it the interesting foresight in terms of, like, realizing that, like, hey, the tensions with given
• 43:05
  Speaker 3: kind of
• 43:05
  Speaker 1: the trade tensions with China. And, of course, the, the all of our machines all start in Taiwan at TSMC. So Right. I mean, all of humanity has a dependency on on Taiwan manufacturing. So it makes sense to to just locate the the the the whole kit and caboodle there.
• 43:24
  Speaker 3: Yeah. The other very specific and annoying thing for us as well is, section 301 tariffs. So for folks who are not familiar, now, yeah, part of the trade tensions ways that the way that that's played out is a new set of tariffs on a pretty vast array of different types of products coming from China importing into the US, typically a 25% tariff. And there are exemptions on specific types of products, and one product that there is actually an exemption on is laptops. But unfortunately, there are not exemptions on basically any of the modules and components that go into a laptop.
• 44:03
  Actually, if we wanted to build the framework laptop in the US, it would be and and manufacture components in China would be substantially more expensive than just actually building a laptop in China. But for, for us, it also meant that because the product's upgradeable, you know, things like main board upgrades, we wanna make sure we don't have to bake in a 25% tariff on the price of a main board upgrade. And so, it's really critical to make sure that we're doing main board manufacturing outside of China.
• 44:31
  Speaker 1: Interesting. Interesting. I mean, this is so much to navigate when you are trying to make a physical product. I mean, there's so much that is complicated, but so, again, in terms of financing, I mean, did you, because, I mean, like Oxide, what you're doing is very idiosyncratic and yet very sensible. Did you, and if you could excuse me prying for a moment, I mean, how did you financing this is an effort like this is a real challenge.
• 44:59
  How'd you get it off the ground?
• 45:01
  Speaker 3: Yeah. It's as as I'm sure there there are folks in Oxide are very familiar with it. It's not easy to fundraise for hardware. You're built basically instantly by being a hardware company, you're filtering down the list of possible investors substantially. And then for us also we have the added challenge that we're consumer, and so you have to take the intersection of people who are not afraid of consumer and people who are not afraid of hardware, which results in a much smaller list of, possible investors.
• 45:31
  We did we were able to bring in some really great investors from, our seed round. We got in, a couple of really great small, small funds, Pathbreaker, Anorak, and then a bunch of folks from the Oculus ecosystem. So basically, the the whole Oculus founding team. And then, we were able to bring in Spark for our series a, and Spark is great. Spark actually also invested in, in Oculus relatively early on, but they are one of these great venture capital firms that's not afraid of consumer or hardware or consumer hardware.
• 46:05
  Speaker 1: So that's so that's interesting that you're because you're tapping I mean, Oculus, the I mean, from a venture perspective, the most important thing about Oculus is it made everyone a lot of money. So it's a it it it's kind of a it's a contrarian play that had a big outcome. And so, I mean, that's helpful to be able to kinda tack into that. I mean, that's kind of the say because the way you describe it is, like, yeah, we're taking the intersection of a consumer products and hardware. It's like, yeah, that's an empty intersection.
• 46:32
  We don't have a to to be able to point to a big outcome. And the thing that I think I'm sure you felt that frustrating as well. I mean, it's great that you've got the the immediacy of that Oculus outcome, and kudos to the Oculus founders for for, you know, helping kind of the next generation of start ups and their former colleagues in terms of of starting the air things. But, god, it is frustrating because, you know, so many of the most valuable things that we have built are a combination of hardware and software. And and yet you find again and again and again, you find venture firms that simply do not want any hardware component whatsoever.
• 47:09
  It's like, oh, okay. So you would not fund NVIDIA? You would do that? You would fund AMD? You would fund Tesla, SpaceX?
• 47:16
  Like, you just say nope on all that? Like, okay. Yeah. I get it. So you wanna fund, like, Groupon and Foursquare.
• 47:22
  Awesome. Yeah. And one of the
• 47:24
  Speaker 3: big challenges with this is that if you look at, you know, the portfolios of a lot of these of a lot of these firms, you'll see that there was some single high profile consumer hardware failure where someone stuck their neck out and decided, like, okay. This Juicero thing sounds great. We're we're gonna get a call again. And then a couple years later sorry to anyone who worked at Juicero. Nothing against you personally.
• 47:49
  I bet, but you did ruin consumer hardware for every Apple. But yeah. But it it is that. And then they looked then and think like, okay. We're never trying this again.
• 47:59
  This is that was it. We're 1 and done. But really though, it is you need to see companies like Oxide and companies like Framework and hopefully many others out there go and succeed and, you know, get traction and and gain growth, to actually make these categories credible for investors again.
• 48:17
  Speaker 1: Yeah. And I think, again, I mean, kudos to the Oculus folks. I just I think that is so important to be kind of paying that forward and and knowing it's like, no. You actually not only can you do interesting things at this junction, but you can develop uniquely valuable things and then things that, like, honestly, like, have a deep moat. I mean, I love that you and then you like Oxide, you've got the comfort of being totally transparent because it's not like someone is gonna look at the design of your expander ports and be like, I can go start a laptop company.
• 48:49
  It's like, yeah. It's a little harder than that. It's a term. Sure.
• 48:52
  Speaker 2: Or if Dell wants to embrace that, go go for it. I'm sure.
• 48:55
  Speaker 1: Go for it. Love that.
• 48:57
  Speaker 3: Yeah. The key thing for us, like, it looks like the thing that we're doing is building these, like, really innovative novel products. But actually, like, fundamentally, the novelty in framework is actually the business model. So, like, the the thing that that we hope people knock off more than anything is actually the business model, but the thing that's the hardest to knock off. This model really, at the end of the day, is about aligning incentives around longevity, that, you know, a product lasting a long time is good for the end user and actually good for us as a business.
• 49:28
  Speaker 1: Yeah. That it that's really interesting in terms of that of the longevity of the relationship as a business model. And, yeah, and and how do you get people to kinda replicate that? How do you get people to to see the power of that? Because, I mean, do you get resistance from folks being like, wait a minute.
• 49:44
  If you make this framework thing easy to repair, someone won't have to buy a new one from you every 18 months. And you're like, no. No. No. That's that's, like, that's a good thing.
• 49:52
  That means I'm making a robust product, and I can go find more customers for this thing.
• 49:56
  Speaker 3: Right. Yeah. In general, like, you know, obviously, our customers, like, totally get it. And I think consumers and, you know, journalist tech press, they get it. Like, they they see the pain points.
• 50:04
  They understand why it's important, why this makes sense. And, you know, even investors, I think, like, the ones who are not afraid of consumer hardware, like, this and think, like, yeah, this feels inevitable. And when something is inevitable, you better get in on it early rather than late. Even suppliers, I think the same thing plays out that, like, you know, in theory, if we if we succeed, when we succeed, we're gonna shrink the size of certain categories. But for suppliers looking at that, they can see that, well, if, you know, if it fails, it fails.
• 50:32
  But if it succeeds, we better get in on this thing early and make sure that we're the ones building that other you know, rather than being late to it.
• 50:40
  Speaker 1: Well and I think, I mean, this is this is classic innovator's dilemma and disrupting innovation where it's like, yeah. This is why the existing players are not going to do this, and why this I mean, and I guess I should ask you as a question. I mean, did you when you had this idea for what you wanted to go do, was it always gonna be a start up? Or did you contemplate, like, can I get this to happen inside of the existing players? Or was it always just clear that, like, no, this has to be new company formation?
• 51:07
  Speaker 3: No. It was yeah. It was clear that it needed to be a startup from the beginning. And it's really it's really yeah. It's it's the counter positioning of the business model.
• 51:16
  You know, if you're an executive at, at I'm not gonna pick a name, I'll say just any electronics maker, and you go to your board and say, hey, we've got this new strategy. We're gonna cut our unit sales and revenues in half over the next 5 years, the board's gonna find another another CEO. But, obviously, for us, we can do that. We can start from 0. It's all growth.
• 51:37
  We're just capturing more of the pie even if the pie is getting smaller as we do that.
• 51:42
  Speaker 1: Totally. Yeah. And I when I I think it is it it is very hard to take a a company that's built on the wrong way of doing it or a way of doing it that that you're trying to obviate and trying to actually modernize. It's just excruciatingly difficult. And, I I mean, certainly, we came to the same conclusion that, like, now this there's a bunch of reasons why this necessitates new company formation, and why the existing players just simply aren't gonna do it.
• 52:10
  And also, I mean, I think the I mean, I'm not sure. You're obviously at Facebook or and or Oculus. It's like when you're in a large company, you're also there's a lot of you're not having to raise money. Great. But you are having to do a bunch of other things that are a huge waste of time.
• 52:25
  Having to to fight rival organizations, everything else. So there's a lot of as challenging as it is to be a start up, there are also a lot of reasons to do it that way.
• 52:33
  Speaker 3: Right. Yeah. And in in Facebook, building the framework laptop would have been 18 months of strategy discussions and then another maybe, you know, 2 years to build the product.
• 52:45
  Speaker 1: Totally. And and then so I mean, obviously, in terms of the future, you've got the the the framework 16. It and I is the the the target demographic for that, I assume, is is is you're really getting into the the laptop power user. I mean, Adam is the target
• 53:01
  Speaker 3: Yeah. We're getting a little bit broader. That's right. And it's that's really how we think about generally, how we think about our roadmap. We think about it in terms of expansion both in categories and in audiences, and obviously those two things go together.
• 53:13
  So we build products that solve more problems for more people, and we continue to add on to that set of problems that we're solving with each one. And so with the framework laptop 16, a big one was gamers would love to be able to upgrade graphics. People who do heavy CAD workloads would love to be able to upgrade their graphics, and they can't today. So let's go and try to solve that.
• 53:33
  Speaker 1: Interesting. So that it really focused on that upgradeability being a real differentiator for those Right. In in those markets.
• 53:41
  Speaker 3: Yeah. That was that was the key. And then the other one was, kinda going back to, like, solving these obvious problems, like, with the expansion cards. Everyone's got different ports. Everyone needs a different set of, peripherals that they plug in.
• 53:52
  For framework laptop 16, we looked at, your big screen laptops basically, and we found that people were making decisions oftentimes based on whether or not a num pad was present on this, you know, $2,000 machine. This, you know, $15 item dictating a $2,000 purchase felt felt a little weird. And so we thought, well, okay, why don't we just let people choose the input that they actually wanna have? And so similar to expansion cards, we built this input module system that lets you reconfigure input, which then also brings us back to this idea of community development where we're never gonna be able to build every type of input device that anyone in the world could want. And, you know, if we look at people's desks, we see their desks are covered in this, like, vast array of different devices that are pretty personal to them.
• 54:38
  And so by opening input in a laptop up as something that is this customizable surface, so we open an ecosystem around, we got to bring in the entire creativity of the community to bring all of that uniqueness and personalization into the laptop itself.
• 54:57
  Speaker 2: And you talked about the the community and and some of their creativity. I've seen some folks, like, building custom cases for the boards that you're making and kind of building, you know, Nintendo Switch like devices. Have there been examples that you've seen from the community that were surprising or particularly delightful?
• 55:15
  Speaker 3: Oh, yeah. Probably the most shocking thing that we've seen is a braille laptop, and this was it's actually a small company called Herbert Research. They're based in Massachusetts, but they've been making, these braille input and output devices for quite a while actually, but they've never been able to make a standalone entire computer that is just a braille laptop effectively. And so they were able to take the main board and build it into this braille device, and we were we actually got a demo of it. They they shipped us one to test out, and it was just this, like, mind blowing experience to, you know, know that there's a framework laptop main board inside of that thing and powered on and see the the the braille lettering start to activate.
• 55:57
  Speaker 1: Wow. And that I mean, that's what what a I mean, it's a it's a terrific feeling just in general to unlock the creativity of other people. Feels great. I mean, there's something I I don't know. It's, like, uniquely human about it of where we that we truly the sum is greater than the parts.
• 56:13
  But to be able to do unlock someone's creativity that that is unlocking for for a third party is, I mean, potentially life changing for them, to be able. That is really amazing. I don't know if you've listened to, Matt Campbell, who's a a a regular listener and and contributor to Oxide and Friends. We had a great discussion with with him. Like, Adam, when was that?
• 56:36
  How many how many years ago was
• 56:37
  Speaker 3: that?
• 56:38
  Speaker 2: That was back in the, in the Twitter Spaces era. So
• 56:40
  Speaker 1: In the Twitter Spaces era.
• 56:41
  Speaker 2: Year and a half, 2 years ago. Yeah.
• 56:43
  Speaker 1: Yeah. Which means you only have to suffer through about, like, 15 minutes at the beginning of the episode where we figure out the audio.
• 56:50
  Speaker 2: It's a theme.
• 56:51
  Speaker 1: It is definitely a theme. I I you know, I gotta say, like, Discord, you know, whatever we are now, a year plus in, Discord has got very few audio issues all all cold. It's probably been great. But, I mean, I can just imagine that they and I if if you didn't didn't catch this one, I didn't know if it was really talking about also the segue to it was actually, about the PLATO system and how, how that just opened up whole new vistas for people back in the day. And it's really neat that you can kind of unlock this ethos where people don't have to just accept these computing devices as kinda falling falling from space.
• 57:26
  They can actually understand how they work. They can actually augment them. They can buy something else from a third party that does something that's neat or interesting or valuable or life changing. I mean, I think it's just that's extraordinary.
• 57:39
  Speaker 3: Yeah. And and we're still really only, I guess, less than two and a half years in from shipping the first product. And so these are still mostly new laptops that are circulating that people are still using as their their primary computer. And I think some of the really interesting and creative stuff is gonna start to happen as those things age even further and, you know, someone picks one up off of, you know, out of a drawer that that's not being used because it's a decade old, but all the documentation's out there, and so they can do anything they want with it.
• 58:10
  Speaker 1: Yeah. That's that that'll be great. And so I guess, you know, so far, what what have been some of your big surprises, I guess? I mean, you've had a a a clear vision for what you wanted to go the the part of the market that you wanted to go serve. And, speaking as part of that market, it feels pretty great, but it I'm sure there have been some surprises along the way.
• 58:31
  Speaker 3: Oh, for sure there have been surprises. One of the pleasant surprises I would say is, like, in a few different open source and specifically, like, Linux communities, we've we've kind of ended up becoming almost a default. Like, if someone, like, asks in a in a specific community around a specific Linux system, like, hey, which which laptop should I pick up? What's got good support? That the answer will just, you know, oftentimes rapidly be, hey.
• 58:54
  You should get a framework laptop. We've, you know, tested it on every generation. It works smoothly. The laptop's great. And, you know, becoming the default is just this very, very, very powerful place to land at, and something that we wanna try to do for more communities so we can out basically how to solve the community's problems.
• 59:13
  Speaker 1: Well, so, actually, that helps to answer a question that came up in the chat. Support, any plans to support previous state in terms of of, another alternative operating system?
• 59:21
  Speaker 3: Yeah. This is an interesting one. So we don't, like, do any anything officially around it, but we have seen a lot of folks in the FreeBSD community using using framework laptops. We've seen a few good write ups. And so, yeah, at some point in the future, that's something where maybe we would do a do our own write up with just some condensed instructions around it.
• 59:38
  Although, I think you'll also find that for and, you know, I think I'm sure I'm sure both both of you are very familiar with. You'll find that in some of these communities, the the more, like, hoops you have to jump through and, like, the more obscure the setup process is, the more the more enjoyable it is to specific communities. We don't wanna make it too easy.
• 59:57
  Speaker 1: You know, I feel attacked. The, well, actually, I was actually the the the so there was an era. So, I mean, of course, we I I used to run Solaris and open Solaris on on laptops. Adam and I both did. And there was a a tremendous value.
• 1:00:13
  And I still feel this way about just running Linux, running Ubuntu on a laptop. It's like, I really it is so valuable to know that other people in the community are using this thing. Like, just getting a there's a there it is an enormous kinda economy of scale of getting the soft because, like, I need to know that I'm not gonna be alone trying to get this thing working. And we all were running the the the same laptop, this Acer Ferrari branded laptop. Adam, you still have your Ferrari?
• 1:00:40
  Speaker 2: Still have it. Still running bits compiled out of my home directory that were let prototype
• 1:00:45
  Speaker 1: Yeah.
• 1:00:45
  Speaker 2: For DTrace on, AMD 64.
• 1:00:49
  Speaker 1: Fast Trap x, baby. I think it's I think that Fast Trap x being the name of your what we now call a a branch, but was, what do we what do we call? What what do it was the the it's teamwork. The the teamwork rather, the, the gate your your framework. Your FastDrop x gate.
• 1:01:08
  But I, you know, I think my Ferrari might be running Fastrack x. I saw mine too.
• 1:01:12
  Speaker 2: The last operating system we ever installed on this.
• 1:01:15
  Speaker 1: It was the last operating install, but, I mean, listen. If we're gonna get free BSD working on the framework, it may be time to read the dream. Alumos on the framework. You heard me here first. Sure.
• 1:01:27
  Why not? The dreams alive.
• 1:01:29
  Speaker 2: I mean, many reasons why not, but sure.
• 1:01:31
  Speaker 1: I mean, many reasons why not. Only some of them good. That's right. The but I actually the that I mean, you have I mean, so much of that is actually so much of that is having the just having the hardware stay still long enough to get driver support. And and then having the transparency about, like, what is the actual, like what what is the Wi Fi NIC?
• 1:01:53
  What is the what is the NIC? What, you know, what what are these various elements that don't tend to stand still and don't and people aren't transparent about. And so it's really hard for it's hard I mean, what you read is, like, it's also hard by the way for it's even hard for Windows to keep up with some of this stuff. So, there's there is enormous value in getting a a kind of a a reference that we all and especially, I mean, you know, then it get that the point that you made early on becomes really important of your dedication to, like, no. We are gonna be also gonna make this thing high performing, and we're gonna hit the right price point, and we're gonna hit some of these other things around around repairability and around transparency.
• 1:02:33
  And it's, I mean, that's that's a lot of constraints on the problem. But, boy, the payoff is huge if you can get there.
• 1:02:39
  Speaker 3: For sure. Yeah. And the other thing for us is that, you know, we try to document as much as possible, but if we fail to document something and someone's trying to develop, you know, a kernel driver or, you know, use a space utility, whatever else, like, they can just post in our community, and someone's gonna show up and answer
• 1:02:54
  Speaker 1: them. That's pretty great. Yeah. So so where should people go for, for kinda next? Obviously, I'm I I I would imagine that you got at least some number of listeners that are configuring their framework framework 13 or 16 right now, but, where where should people go to kinda learn more about framework and the the community and so on?
• 1:03:13
  Speaker 3: Frame dot work. We actually picked up the domain name before figuring out what the company was gonna be, but it is frame.work.
• 1:03:20
  Speaker 1: Do you have someone do you have a framework.com that's trying to sell itself to you over and over again the way we do? We do. The oxide.com?
• 1:03:27
  Speaker 3: We do, unfortunately. I'm sure the the story is the same in in every scenario, but it's some delusional individual who's got, you know, too many zeros at the end of the number.
• 1:03:36
  Speaker 1: Way too many zeros at the end of the number. And so and, like, we approached the oxide.comowner, and we're like I think I'm like, I don't know. I pay, like, I don't know, $10,000 for it. And they're like, yeah. The the we work with a broker who's like, I'm able to get a real deal on this thing, like, 420 k.
• 1:03:52
  Speaker 3: You're you're lucky that you have a broker. Ours is, is an individual who doesn't who doesn't even wanna go the broker path.
• 1:03:59
  Speaker 1: Oh, did you do you wanna go the broker path? It is like and I'm almost like, hey. Sorry, oxide.com. We're oxide.computer, and I don't care about oxide.com. Oxide.com.
• 1:04:07
  I'm gonna go jump in a lake. And, hey, owner owner of framework.com. Get bent. It's frame frame dot work. So, the, it's well, good good on you frame dot work, and, yeah, I think you got I know we got a lot of, simpatico between our 2 companies for sure and a lot of of shared ideas about what the what the future of computing should look like.
• 1:04:36
  And, really, it it is about empowering the people who are using computers. And that is what computers that was the and I think you look at every big revolution in computing, and it was empowering for people. It was empowering for users, and it was empowering for the people who would develop who would harness their own creativity. And, I feel and you must feel too that we we we lost that for the last couple of years here in for both the the the client side and the server side. So, let's get it back.
• 1:05:10
  Speaker 3: Right. And, you know, the the interesting thing about that is, like, we point back at success cases and, like, actually looking at Oculus, that was one of the success cases that we started by empowering developers. So we made sure that we, you know, very deliberately and carefully positioned it as a dev kit knowing that consumers were gonna buy it. But most importantly, you know, if you thought, like, 18 months for a framework launching our first product was fast, we you know, in Oculus, we shipped the first dev kit in, like, 6 months and then
• 1:05:35
  Speaker 1: Oh my god.
• 1:05:36
  Speaker 3: The second dev kit in the year after that. So it was really, like, the first few years of Oculus were almost entirely focused on developer enablement. And, and, you know, as you said, like, it really, really is important to get developers bought in and enable and empower them.
• 1:05:51
  Speaker 1: Yeah. That well, I and that that's that's what it's about, and it's about empowering the people that are actually using this stuff. So, I I would what's, I I'm super excited to have my my framework laptop, and I know that we're gonna, I I would like to believe there's a framework laptop user out there that can't wait to have a rack full of oxide. So you don't have to
• 1:06:14
  Speaker 3: And do a bundle deal?
• 1:06:16
  Speaker 1: You know
• 1:06:16
  Speaker 3: what I
• 1:06:16
  Speaker 1: wanna do? Actually, we need to do, like, a framework oxide. We need to do, like, an oxide edition, like, an oxide bezel or whatever Yeah. That we do as, like, a giveaway. Like, buy a rack, and you get a get a free framework 13 or 16.
• 1:06:29
  Speaker 3: Yeah. We can come up we can come up with a, yeah, a one new module that's got expansion card slots or something.
• 1:06:36
  Speaker 1: Yeah. I I'm happy to give away the laptop. If you have to do something, we could,
• 1:06:40
  Speaker 2: that's right. All USB a across the front.
• 1:06:43
  Speaker 1: All USB a. Exactly. Exactly. Awesome. Well, Narayan, thank you so much for for joining us, and, it's we're we're we're huge fans, and we should I you you have to come back again.
• 1:06:57
  We can compare notes on, the the hardware software journey, and, I think inspire the next generation the way we ourselves have been inspired. So thank you so much for being here.
• 1:07:09
  Speaker 3: Thank you for having me. It's great to chat.
• 1:07:12
  Speaker 1: Alright. Thanks, everybody. Take care. Talk to you next time.