| View previous topic :: View next topic | 
	
	
		| Author | Message | 
	
		| dmendesf 
 
 
 Joined: 31 Dec 2005
 Posts: 35
 
 
 
			    
 
 | 
			
				| dsPIC33A support? |  
				|  Posted: Sat Aug 10, 2024 4:20 pm |   |  
				| 
 |  
				| Anyone knows if dsPIC33A support is planned for future releases of CCS compiler? |  | 
	
		|  | 
	
		| Ttelmah 
 
 
 Joined: 11 Mar 2010
 Posts: 19967
 
 
 
			    
 
 | 
			
				|  |  
				|  Posted: Sun Aug 11, 2024 7:42 am |   |  
				| 
 |  
				| Ask CCS. I'd say probably a while. Understand these are 32bit CPU's with a
 FPU. However since the core is compatible with the PIC32, It should be
 possible for CCS to support these. I wouldn't be at all surprised though
 if this becomes a new 'PCE' compiler.
 It is not just a small change to support these.
 |  | 
	
		|  | 
	
		| empty 
 
 
 Joined: 13 Jan 2018
 Posts: 16
 
 
 
			    
 
 | 
			
				|  |  
				|  Posted: Sun Aug 11, 2024 10:33 am |   |  
				| 
 |  
				| It may take a while if they decided to support this new family. CCS doesn't give any support to PIC32 devices as they're MIPS or ARM architecture
 based microcontrollers.
 The dsPIC33A is different, it has a 32-bit CPU using Harvard
 architecture, basically the same thing with other PIC/dsPIC devices.
 I hope we'll see a full support of these devices.
 |  | 
	
		|  | 
	
		| Ttelmah 
 
 
 Joined: 11 Mar 2010
 Posts: 19967
 
 
 
			    
 
 | 
			
				|  |  
				|  Posted: Sun Aug 11, 2024 11:28 am |   |  
				| 
 |  
				| Agreed. It is actually even more compatible than that, the instruction core matches the 16bit units. However a lot of the peripherals are different, which
 is why I suspect they may launch it as an alternative, rather than just an
 extension to PCD. Crossing things that they do decide to support these
 on a new version of the existing compiler.
   However given how radical the changes are, I'd be very careful for a few
 versions if they do do this.....
 
 Asking them is the thing to do.
 |  | 
	
		|  | 
	
		| dmendesf 
 
 
 Joined: 31 Dec 2005
 Posts: 35
 
 
 
			    
 
 | 
			
				|  |  
				|  Posted: Tue Aug 20, 2024 11:58 pm |   |  
				| 
 |  
				| Asked them a few days ago. Still no answer... Probably a tough decision at this point. |  | 
	
		|  | 
	
		| Ttelmah 
 
 
 Joined: 11 Mar 2010
 Posts: 19967
 
 
 
			    
 
 | 
			
				|  |  
				|  Posted: Wed Aug 21, 2024 5:04 am |   |  
				| 
 |  
				| Yes, They are probably asking themselves just how hard/easy it would be.
 Touching timber things....
  |  | 
	
		|  | 
	
		| temtronic 
 
 
 Joined: 01 Jul 2010
 Posts: 9589
 Location: Greensville,Ontario
 
 
			    
 
 | 
			
				|  |  
				|  Posted: Wed Aug 21, 2024 7:53 am |   |  
				| 
 |  
				| As in any business decision about a product, you have to consider time and effort vs actual ,probable sales. I'm sure someone, somewhere in the world could do the math. Take sales of that series PICs then look at their market share for say the PIC18F series. If similar, then a compiler my be a worthwhile venture.
 One problem is do you hire more coders or take existing ones away from say 'support' or 'bug detail' ?
 I don't know if any of the current compiler code could be used 'asis' for the new version or a total 'from the ground up' is necessary ? Hate to think of the manhours it'll take !
 Those in active business know how long it took them to get their product into the market and true costs. So was it worthwhile ??
 Frankly I don't see how CCS stays up to date considering all the new PICs !!
 |  | 
	
		|  | 
	
		| Ttelmah 
 
 
 Joined: 11 Mar 2010
 Posts: 19967
 
 
 
			    
 
 | 
			
				|  |  
				|  Posted: Wed Aug 21, 2024 9:14 am |   |  
				| 
 |  
				| The key on these is that the core instruction set is fully compatible to the existing 16bit chips. So the compiler needs no changes at all. The
 differences' are the peripherals.
 Makes it a lot easier.
  |  | 
	
		|  | 
	
		|  |