Let's Make Robots!

MCUs a dying Breed (eg.Arduino...

Are MCUs a dying Breed (eg.Arduino......Propeller .... Picaxe)


How would you react if I were to predict that by the end of this year our favourite microprocessors would only exist as a Ghost in the works.......


Background :-

Parallax (news) has released   on the 15th August, 2014 the code for a verilog version of the Propeller P1 (Open source Details here)

This allows you to program a FPGA chip to run as a Propeller.....

Parallax are very bold with this move ... so come on Atmel and Picaxe when does your verilog code hit the scene!!!.


Below is my working example ...... Its my Deo Nano emulating a Propeller (P1) mcu .. yes including the Video coding.

Yes it emulates 8 parallel cogs (why stop at 8 you can easily add more...only the number of gates limit )

They have so far released the verilog coding for three emulation boards

  1. De0 Nano
  2. BeMicroCV
  3. DE2_115

What does this mean for our favourite mcus .......

With the larger FPGA chips you could conceivably put a couple of Arduinos,Propellers and Picaxes ....all together on one chip, yes all your favorite micros on one platform... now wouldn't that be awesome.


What is a FPGA In a nutshell :-

FPGA  (Field Programmable Gate Array)

Imagine a box full of "floating" logic gates all separate from each other.....

Your program picks up these logic gates and connects them together to form program chains (modules).

These program modules run through the chip, each module can be separate or linked to others, also modules run in parallel with each other (ie they all run at the same time)

The chip is accessed with basic I/O pins and the ouside world is contacted. ie as an Arduino/Propeller/Picaxe system.

In essence its a programmable circuit, aimed for rapid prototyping and deployment (ie make your board first and think of a product after).

Here is the Deo Nano representation of my Emulated Propeller chip :-

and here is how it actually places it on the chip :-

If I squint my eyes I can make out the 8 cogs......can you ?


So my Question is :-

What is the future for our favourite mcus and will they only survive as "Ghosts" on an FPGA.?

 


Update :- 20140831

 

FPGA Tracking my HackaDay prize Tee_Shirt

I have used OpenCV via MyRobotLab to send controls to a camera mounted on a X/Y servo gimbal.

that tracks and follows me around ...in this case the nose of the skull (if you look closely you can see the cross hair)

 


 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

I also don't think that FPGAs will not be replacing standard microcontrollers anytime soon.

However, they are really great where you need programmable logic.

  1. When designing a new processor you can do this on an FPGA development board instead of simulating it on a mainframe. Now, in fact, pretty much anybody can design their own processor if they are willing to learn the language to program the FPGA in.

  2. If you need fast programmable logic. For instance, many of those LED signs and billboards are run by FPGAs because normal microcontrollers aren't fast enough.

  3. If you have a specific thing that needs to run faster than it would on a microprocessor. For example, I'm building a balancing robot. Perhaps I could put the basic balancing/motor control algorithm into an FPGA so I can run the control loop faster and perhaps smoother. (I doubt I'd actually go this route because I've seen videos of balancing robots that ran smoothly over 10 years ago.)

Thanks for your input and giving a positive grasp on the ways these chips can blend in..

.... also gives me the thought now that even low logic count FPGAs have a chance to fit in.

regards G

You're welcome.

I do think that FPGAs are useful in certain areas. They also have a "cool factor" that is hard to beat.

If somebody like OB1 wants to make his ideal microprocessor, then the FPGA is the way to go. When I was in college we wrote simulators in C that were fairly slow. Of course sometime this was because tool used were occasionally wrong and used for the wrong reasons.

There was an old editor called, I think, teco. If you want to take a look at how both horrible and powerful it was, there is a teco emulation mode in emacs. Which is strange because emacs was originally programmed in teco. The horribleness of teco came about because the commands are single characters followed directly by a parameter such as a file name, followed by other characters. One fun thing to do is to load a file and type your name in the command line. You were smart enough to make a backup of that file off line, weren't you?

Sysadmins would have contests with teco as to who could make the best hack. To my knowledge the best hack was a full Fortran compiler. However, a few days later, possibly before the first "hello world" program was created, the hacker confessed: he did not really write a fortran compiler. He wrote a simulation of the computer in teco and used the PDP11's fortran compiler!

So it is possible to write simulated processors in anything that can be classified as a programming language.

Isn't asking AtMel or RevEd to release a verilog for their chips kind of like asking Dodge to come up with a way for a Bugatti Veyron to emulate the driving experience of an Omni? ;-)

This discussion is way over my head, but not for a very long time, the arduino is here to stay, and stay this article talks about why

http://makezine.com/2011/02/10/why-the-arduino-won-and-why-its-here-to-stay/

It boils down to even newbies like me can learn to use it, there is so much support, libraries galore, price point on uno is $10 shipped a pro mini $3 shipped the masses can afford to play, for most of us not techno savy its way over kill for our projects anyway.

$7 uno r3 http://tinyurl.com/o84c8ku

$2.25 pro mini shipped http://tinyurl.com/kwdnngb


I love ya Dude!, and this is fantastic news for people with your incredible skills, but for those of us in the hoveling mass hobby crowd this is years away from making a dent in arduino's market share.


For those following along You could also try out the FPGA Propeller II  ..... It gets better and better ....

Propeller II on the Deo Nano and DE2_115

Qoute :-

Chip has released the binary for:
- a 1 cog Propeller II emulation (with restrictions) that runs on a Terasic DE0-NANO FPGA board.
- a 5 cog Propeller II emulation (with restrictions) that runs on a Terasic DE2-115 FPGA board.

Hi Gareth, I just received my BeMicro CV kit. I'm ready (I hope) to jump aboard the FPGA bandwagon but I also doubt FPGAs will be replacing microcontrollers anytime soon.

I see the FPGA as another tool in the toolbox which I will hopefully learn to use. Having my beloved Propeller to play with on the FPGA is what made me take the plunge but I doubt I'll be replacing many of my Propeller (or (don't tell anyone) Arduino) boards with FPGAs anytime soon.

BTW, I think it was bdk6 who first pointed out the Verilog code for the Propeller here on LMR.

I am also worrying about my stock Propellers getting "Dusty".

The BeMicro CV kit will be my next one.....

I am well pleased with the Propeller emulation... the Propeller DL is mega fast, you will not be disappointed.

You can guess the first thing I did was to test out the video.... :-)  ..... need I say more.

tnx for the bdk link......I am checking it next.....

BTW :- I also have some news with regards to a Propeller Project I am collaborating on ... I will appear somewhere shortly.... stay tuned.

Sorry Gareth but I think your prediction is out by at least 20 years. BDK6 has hammered you pretty good.

The only reason the current MCUs will stay is because they have a "Comfort Zone".

Anyone trying to relieve me of my Picaxe 08M collection will have a hard job I tell you .....

regards G