10 ms·

Embedded Systems – Shape the World (2014)

53 points by Cieplak 3y ago

eschneider 3y ago
This looks like a good resource. Remember kids, embedded systems is like retro computing that people will pay you to do. :)
- MrBuddyCasino 3y ago
  pay very little to do
  - ChuckNorris89 3y ago
    >pay very little to do Not in Europe. Since FAANG level jobs aren't that numerous here compared to US so the difference between embedded and non-embedded work is close to zero. Pay is pretty similar. Add sure, but if your only target in life is working for FAANG & Co. then every other tech job in the world pays very little. You might as well also not want to be a doctor, teacher, lawyer, as well because hey, FAANG pays better. Everyone should abandon their passions and career dreams and become devs for FAANGs because pay is everything that matters in a career.
    - fest 3y ago
      Yeah, I work in a robotics company in Europe- our embedded people are paid essentially the same as web/enterprise software developers in the same market. Definitely a lot less remote Opportunities (though usually for institutional reasons- most of the work can be done remotely, if company is organized for that- which not a lot are). EEs are IMHO underpaid in both sides of the pond though.
      - i_am_a_peasant 3y ago
        If you're in Germany I'd like to get in on that. :P
      - tpmx 3y ago
        > EEs are IMHO underpaid in both sides of the pond though. Because of EEs in PRC and Taiwan with relatively speaking much lower salaries for about the same level of competency [who also have native-level access to the design/supply ecosystem over there].
      - AnimalMuppet 3y ago
        Embedded can be remote, if the hardware is small enough and cheap enough to ship one to each remote developer. I worked on telephones for the hard of hearing, and we had some remote developers. No problem. On the other hand, there are embedded systems that are for 3D printing spaceship hulls. I kind of suspect that remote isn't going to fly for working on those systems...
        eschneider 3y ago
        You can probably do (most) of that remotely. If you're writing embedded systems that can't be 95% tested off the hardware, you're living in unnecessary pain. I work on devices bigger than a car and sure, occasionally I have to visit the 'office', but almost everything can be debugged and tested remotely.
        fest 3y ago
        Yes, remote access to test units (if practical, I.e. more than one instance of system exists) or the real one, simulated h/w for higher level software (i.e. a GUI that connects to simulated device instead of real one). Obviously there will always be some physical component to embedded development, but in many cases it is possible to do a lot remotely.
  - synergy20 3y ago
    very true, though I like it technically, pay is not that great,also hard to scale up
  - eschneider 3y ago
    I don't know about that... I've done both embedded and non-embedded work and I've never had to take a pay cut to do the embedded work. It's mostly a question of finding a company where the embedded software isn't an afterthought.
  - sgtnoodle 3y ago
    I've always been quite happy with my compensation as an embedded software engineer. There also seems to always be high demand for the skill set.
kierank 3y ago
Seems to miss the part where it's ok to create terrible software in embedded, with custom forks of everything, binary blobs galore. But it's all fine as it shaves a few cents from the BOM.
- joezydeco 3y ago
  Those who love law, sausage, or embedded devices shouldn't watch any of them being made.
- spyremeown 3y ago
  We are slowly getting better at this. Most embedded Linux companies for example have shifted or are in the process of shifting to an upstream-first policy. A lot of the toolchains are open, busybox is used everywhere. A lot of the really good HALs are open source. We're getting there!
  - synergy20 3y ago
    which open source HAL do we have these days? thanks.
    - tonyarkles 3y ago
      I've been using Zephyr lately and have been really enjoying it. It still has the vendor HAL libraries available (e.g. Microchip/Atmel ASF), but the devicetree-based HAL has been fabulous to work with. I've tried a number of different embedded frameworks/RTOSes over the years, but it's the first that I actually feel happy using. There's a few sharp edges for sure, but it's at least an order of magnitude better than anything I've used in the past. One thing that's been particularly good is that the abstractions feel right, and it's allowed our team to collaborate on firmware without constantly tripping each other, because the module boundaries and abstractions are good.
      - tomhoward 3y ago
        This is good to hear. I'm just getting up and running with tutorials for a Nordic IoT SIP using Zephyr, with plans to develop a new model of an agricultural data monitoring/control product we've been offering for a few years. So far it seems really good. I personally have very little experience with embedded software (many years with web software) but am finding it easy to pick up.
PAPPPmAc 3y ago
I teach a sophomore-level class for electrical and computer engineering majors out of the Valvano materials, we put references to the chapters in both the free online and later inexpensive print version in our materials. It's great in many ways, and free-to-cheap which is a particularly nice quality in a textbook, but it is _quite_ out of date at this point: it's tied to the old "ARM Compiler v5" bespoke compiler rather than the "v6" LLVM-derived one, and even Keil doesn't ship with v5 anymore. v5 used a different assembly syntax (v6 is GNU style, being LLVM derived - fortunately the toolchain can still ingest v5 style assembly _files_ just not inline so we can keep the assembly the students interact with matching the book), different tooling for C/Assembly interop, and an assemblage of bespoke macros and extensions, so there has been some creative tooling work to keep the book and assignments consistent for the students. The EK-TM4C123GXL TivaC board that is the less expensive of the two targets the Valvano materials are built around is a pretty nice if slightly expensive embedded trainer, about $20-25 for a Cortex M4F with a fairly rich assortment of peripherals, a 2nd micro configured as a programmer/in system debugger on one end of the board, easy-to-interface male AND female headers on a bunch of broken out pins, pads to tap the USB-Serial traffic while it still looks like serial, high and low speed oscillators arbitrated by a pretty fancy clock circuit, two user buttons, and an RGB LED. Powerful enough and with enough Flash and RAM to get interesting things done, not so powerful or memory rich that you can get away with modern programming decadence, and simple enough that we can talk about the nuts and bolts of any particular feature. We construct from "Baby's first assembly program" through "Register-direct manipulation" and finally "Controlling external devices via the HAL functions" over the course of a semester. Because of the way the degree programs are structured, we're a little more focused on "This is how computers actually work" than on embedded applications, but I'm still pretty proud of the course and the students mostly seem to appreciate it.
- iMario 3y ago
  Very interesting... are your materials available?
  - PAPPPmAc 3y ago
    Unfortunately it's mostly locked up in a university CMS, and I haven't put copies out on the public 'net partly just to slightly obscure the semester to semester changes we make to throw off lazy cheaters.
freeman94 3y ago
Hah, Valvano was a professor of mine. Funny to see this here. It was a good class.

53 points by Cieplak 3y ago

This looks like a good resource. Remember kids, embedded systems is like retro computing that people will pay you to do. :)

MrBuddyCasino 3y ago
pay very little to do
- ChuckNorris89 3y ago
  >pay very little to do Not in Europe. Since FAANG level jobs aren't that numerous here compared to US so the difference between embedded and non-embedded work is close to zero. Pay is pretty similar. Add sure, but if your only target in life is working for FAANG & Co. then every other tech job in the world pays very little. You might as well also not want to be a doctor, teacher, lawyer, as well because hey, FAANG pays better. Everyone should abandon their passions and career dreams and become devs for FAANGs because pay is everything that matters in a career.
  - fest 3y ago
    Yeah, I work in a robotics company in Europe- our embedded people are paid essentially the same as web/enterprise software developers in the same market. Definitely a lot less remote Opportunities (though usually for institutional reasons- most of the work can be done remotely, if company is organized for that- which not a lot are). EEs are IMHO underpaid in both sides of the pond though.
    - i_am_a_peasant 3y ago
      If you're in Germany I'd like to get in on that. :P
    - tpmx 3y ago
      > EEs are IMHO underpaid in both sides of the pond though. Because of EEs in PRC and Taiwan with relatively speaking much lower salaries for about the same level of competency [who also have native-level access to the design/supply ecosystem over there].
    - AnimalMuppet 3y ago
      Embedded can be remote, if the hardware is small enough and cheap enough to ship one to each remote developer. I worked on telephones for the hard of hearing, and we had some remote developers. No problem. On the other hand, there are embedded systems that are for 3D printing spaceship hulls. I kind of suspect that remote isn't going to fly for working on those systems...
      - eschneider 3y ago
        You can probably do (most) of that remotely. If you're writing embedded systems that can't be 95% tested off the hardware, you're living in unnecessary pain. I work on devices bigger than a car and sure, occasionally I have to visit the 'office', but almost everything can be debugged and tested remotely.
        fest 3y ago
        Yes, remote access to test units (if practical, I.e. more than one instance of system exists) or the real one, simulated h/w for higher level software (i.e. a GUI that connects to simulated device instead of real one). Obviously there will always be some physical component to embedded development, but in many cases it is possible to do a lot remotely.
- synergy20 3y ago
  very true, though I like it technically, pay is not that great,also hard to scale up
- eschneider 3y ago
  I don't know about that... I've done both embedded and non-embedded work and I've never had to take a pay cut to do the embedded work. It's mostly a question of finding a company where the embedded software isn't an afterthought.
- sgtnoodle 3y ago
  I've always been quite happy with my compensation as an embedded software engineer. There also seems to always be high demand for the skill set.

kierank 3y ago

Seems to miss the part where it's ok to create terrible software in embedded, with custom forks of everything, binary blobs galore. But it's all fine as it shaves a few cents from the BOM.

joezydeco 3y ago
Those who love law, sausage, or embedded devices shouldn't watch any of them being made.
spyremeown 3y ago
We are slowly getting better at this. Most embedded Linux companies for example have shifted or are in the process of shifting to an upstream-first policy. A lot of the toolchains are open, busybox is used everywhere. A lot of the really good HALs are open source. We're getting there!
- synergy20 3y ago
  which open source HAL do we have these days? thanks.
  - tonyarkles 3y ago
    I've been using Zephyr lately and have been really enjoying it. It still has the vendor HAL libraries available (e.g. Microchip/Atmel ASF), but the devicetree-based HAL has been fabulous to work with. I've tried a number of different embedded frameworks/RTOSes over the years, but it's the first that I actually feel happy using. There's a few sharp edges for sure, but it's at least an order of magnitude better than anything I've used in the past. One thing that's been particularly good is that the abstractions feel right, and it's allowed our team to collaborate on firmware without constantly tripping each other, because the module boundaries and abstractions are good.
    - tomhoward 3y ago
      This is good to hear. I'm just getting up and running with tutorials for a Nordic IoT SIP using Zephyr, with plans to develop a new model of an agricultural data monitoring/control product we've been offering for a few years. So far it seems really good. I personally have very little experience with embedded software (many years with web software) but am finding it easy to pick up.

PAPPPmAc 3y ago

I teach a sophomore-level class for electrical and computer engineering majors out of the Valvano materials, we put references to the chapters in both the free online and later inexpensive print version in our materials. It's great in many ways, and free-to-cheap which is a particularly nice quality in a textbook, but it is _quite_ out of date at this point: it's tied to the old "ARM Compiler v5" bespoke compiler rather than the "v6" LLVM-derived one, and even Keil doesn't ship with v5 anymore. v5 used a different assembly syntax (v6 is GNU style, being LLVM derived - fortunately the toolchain can still ingest v5 style assembly _files_ just not inline so we can keep the assembly the students interact with matching the book), different tooling for C/Assembly interop, and an assemblage of bespoke macros and extensions, so there has been some creative tooling work to keep the book and assignments consistent for the students. The EK-TM4C123GXL TivaC board that is the less expensive of the two targets the Valvano materials are built around is a pretty nice if slightly expensive embedded trainer, about $20-25 for a Cortex M4F with a fairly rich assortment of peripherals, a 2nd micro configured as a programmer/in system debugger on one end of the board, easy-to-interface male AND female headers on a bunch of broken out pins, pads to tap the USB-Serial traffic while it still looks like serial, high and low speed oscillators arbitrated by a pretty fancy clock circuit, two user buttons, and an RGB LED. Powerful enough and with enough Flash and RAM to get interesting things done, not so powerful or memory rich that you can get away with modern programming decadence, and simple enough that we can talk about the nuts and bolts of any particular feature. We construct from "Baby's first assembly program" through "Register-direct manipulation" and finally "Controlling external devices via the HAL functions" over the course of a semester. Because of the way the degree programs are structured, we're a little more focused on "This is how computers actually work" than on embedded applications, but I'm still pretty proud of the course and the students mostly seem to appreciate it.

iMario 3y ago
Very interesting... are your materials available?
- PAPPPmAc 3y ago
  Unfortunately it's mostly locked up in a university CMS, and I haven't put copies out on the public 'net partly just to slightly obscure the semester to semester changes we make to throw off lazy cheaters.

freeman94 3y ago

Hah, Valvano was a professor of mine. Funny to see this here. It was a good class.