Home > On-Demand Archives > Q&A Sessions >
Live Q&A - Embedded Design During Chip Shortages
Dave Hughes - TUXERA - Watch Now - Duration: 27:03
Excellent!
Thank you for your talk, Dave. An excellent discussion. I'd appreciate the slides for reference as well.
thank you for your kind words - i have forwarded your request - you should get a response soon...
Dave
Abstracting the RTOS was the most interesting section, that is one area I didn't consider for chip shortages! Thanks for sharing your knowledge in that area.
Thank you Dave, for your presentation. Will you be publishing your slides here?
I've pitched a similar framework before -- before 'Supply Chain Issues' was a thing -- so your talk is resonating well with me. Thank you for your consideration.
Hi Dan, Thanks - our marketing team will contact you soon about how to get the slides.
Dave
Hi Nathan, I really liked the talk! Thank you very much!
One initiative achieving a level of abstraction for ARM is CMSIS - what is your experience with /take on CMSIS?
Andreas
Hi Andreas, I think we covered this a bit in the live Q&A - but in short - if you are wedded to ARM (or more accurately the supported Cortex types) - then it is a very comprehensive abstraction and as long as you can get conformant drivers for your target it must be an important option. For us it did not work because we needed our components to run on any architecture - but we have abstractions to CMSIS (and in particular the RTOS) for our components.
Thanks for the presentation, very interesting.
At my company we are seeing more and more customers asking us to port their firmware to new microcontrollers. The difficulty of those migration projects highly depend on the abstraction you describe in your video.
Also, could you publish the slides ?
hi Nathan, I have forwarded this request to our marketing team - they should be in touch soon.
Thanks for listening!
Dave
Very interesting talk!
I think pretty much everybody today is minimally aware of the risks associated with non-portable code and some basic strategies like a set of primary software design principles, and the use of language usage constraints like MISRA. Today, however, with the increasing shortage problems, these risks have become more probable than ever,. and developing without minimum portability in mind is pretty much suicidal. In addition, the rise of embedded software complexity and off-the-shelf software and development ecosystems provided by silicon vendors with a certain degree of lock-in provides a lot more of coupling surface increasing the effort of making it portable, making careful planning and designing more necessary than ever.
What I enjoyed most about the talk though is the ability to structure and summarize the main challenging aspects and tactics to overcome them with success that the talk is able to distill.
Thank you or your kind words and your thoughts - much appreciated!
Dave
Well-said Dave. Very good guidelines to follow that could save many projects from all the different kinds of crises that could cause a need to change MCUs to meet demand.
10:02:58 From Jacob Beningo : Feel free to enter your questions into the chat box or raise your hand to unmute your microphone and chat directly with the speaker 10:03:37 From Paul Vittorino : So do you not like CMSIS? 10:04:36 From Shilpa Anbalgan : Can you please put in the chat, on reference books 10:05:07 From Max Maxfield : Embedded C Coding Standard by Michael Barr 10:05:46 From Shilpa Anbalgan : Thank you 10:10:27 From Ross : MISRA C does an excellent job of making C safer. There doesn't seem to be a similar standard covering modern C++ though? 10:11:36 From Ross : The Misra C++ standard is very old and doesn't cover C++11. 10:14:07 From Max Maxfield to Stephane(Direct Message) : An other book is "The Hacker's Delight" by Henry S. Warren, Jr. 10:14:41 From Max Maxfield : Another book is "The Hacker's Delight" by Henry S. Warren, Jr. 10:19:30 From Paul Vittorino : How would an engineer who has not seen many different microcontrollers know the appropriate level of abstraction? 10:24:08 From Nathan O. : How would you justify the additional development time needed for building those abstraction when asked by management to "just get things done" ? 10:25:24 From Gillian Minnehan : Good question Nathan, I'm also interested in that 10:25:54 From Max Maxfield : Tell them it takes less time to add these abstractions than to sort things out later -- and it costs less money to sort things out upfront than when they are out in the field. 10:26:11 From Andreas J. : Thank you very much Dave! 10:27:30 From Nathan O. : Thanks, I'll try to show them your presentation, it might help educate them ;) 10:27:42 From Phillip Kajubi : Thank you so much!! Fantastic session 10:28:04 From Brandon2022 : Thank you
Hi Dave, it was a real pleasure listening to your talk, getting insights and and hear about your experience on SW Portability/Abstratction/... . I really appreciate you taking the time to share all this and giving us those really valuable tools and thoughts for our engineering tool-box.
There's on more thing you made me curious about and it would be great to hear your opinion/experience: During the the Q&A you made a very interesting statement about communications protocols - that THE critical part of the design is how you move the Data.
Do you have some more tipps for designing/developing a performant communication protocols? What to watch for? What are the trapps and pittfalls to avoid? Abstraction best practices?
Since I am currently working on designing a specific middleware, so to say a highlevel protocol, you hit a sweet spot with that comment :-)
I am really looking forward to hear some more of your ideas!
Thanks, Andreas