For 'roll-your-own' Industrial Control the re-hash (and amplifcation) of this information on Microcontrollers, Microprocessors, PLDs, ASICs, FPGAs could have been integrated into the earlier class had those classes been groomed in their presentation. The objective being to cover this information, then press on to bigger things ...
Good presentation - better slides. Still, verbal description exceeds by too far the bulleted text that would make the presentation stand-alone and self documenting. 28 slides in this presentation compated to 22 slides in the previous course shows that better slides means more info can be covered in same time -- efficiency and productivity -- this could be improved as stated above ...
About the slides 13 to 15: the duty cycle and the power-down and the peak currents (which are different for 8, 16 or 32 bit microcontrollers) are important in the power calculus. So the 32 bit microcontrollers 'supremacy' is not so obvious for any kind of applications. Thank you.
Could somebody help?... I have power point 2007(updated) in spanish. When I open, the downloaded slide I've this mesg: "there is no text converter for this type of archive (my translation from spanish). Do i need to switch some option( which i dont see it in "properties" for this program) or ask MS for an update to be able to read in english? I dont have this problem with Adobe reader 9.5.0 part of the whole MS suit in spanish. Thanks.
@gschmick "Which uC's have built-in RF sections?" I don't know off-hand -- but there's a company called Synapse Wireless here in town (synapse-wireless.com) who do low-power area network modules and I knwo their CTO and he was telling me about them -- they'v ebeen around for at least a year -- maybe more -- I'll try to find out more (but no promisses because I am up to my ears in alligators fighting fires without a canoe at the moment ... I never metaphor I didn't like :-)
I developed in forth a library of quotient arithmetic, that may have its use in "integer like" signal processing. If one could devise a MPU that implemented that sort of processing done in forth, it would be neat!
One of the things I wan tto do is to write my own BASIC interpreter. Forth is a threaded-interpreted language -- I always wanted to create my own Forth Interpreter also ... but there's so much fun stuff to do and so little time to do it all in...
@jjrochow "Max: do you know anything about System-C?" Not to program in, but I know a bit -- my understanding is that it's like regular C but with some extra data types and also with the concept of time. If you try to represent things at thr Register Transfer Level (RTL) it's not tremendouisly efficient ... but it gets more efficient as you move higher in abstraction. It's used a lot to generate Transaction-Level Models (TLMs) which are good for greating virtual prototypes and testbenches and stuff
I've played around with different Forth 80 implementations and also a bit with retroforth a while ago. I smuggled in a tiny retroforth app into some software running at a service provider for a "large german car manufacturer" :-D
@clia -- re your question about GPUs -- as you say, although they have lots (sometimes hundreds) of cores, the cores themselves are relatively simplistic -- having said theis they each have local memory and sophisticated fixed-point capability (I think some may even be floating-point these days) -- do it depemnds on your program -- if the program is ammenable to being partitioned across hundreds of processors then you can see orders of magnitude speed up -- but it woudl have to be the right sort of program
Re the questions about programming. Antifuse FPGAs are programmed one time off-board and cannot then be reprogrammed. Flash devices are programmed on-board (typically) -- they will power-up instant on, but they can be reprogrammed. SRAM-based devuces have to be reprogrammed every time the board is powered up. Thsi can come from an external flash memory chip or a number of other techniques. A hybrid-flash/SRAM device powers up and performs a massively paralled copy from th eflash to the SRAM -- so it appears to be instant on -- also the Flash can be reprogrammed while the chip is still running using its SRAM configuration .... Phew!
@Max - I heard that the reason you can have so many cores in a GPU was because of the simplistic nature of the logic they represent; instead using higher order math, they break this down into many more simple logic steps that they can then attack with the numerous cores. Does this require the programs to be rewritten in order to help the GPUs understand what they have to accomplish?
Max: Good presentation, however, there *are* quite a few ASICs that do not have a processor on them. There is a pretty good price hit for including a processor since it requires licensing and silicon area and often extra pins which means a larger package.
@tstout - I think that is the case. I can't help but wonder if that is why the new fuel system design for NASCAR Sprint Cup causes the drivers so many headaches trying to restart a stalled engine. They have power down, power back up (reboot?), wait (reload programming?), the try restarting the engine. Are they using FPGAs?
The streaming audio player will appear on this web page when the show starts at 2pm eastern today. Note however that some companies block live audio streams. If when the show starts you don't hear any audio, try refreshing your browser.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.