What will happen "pic programming" in the future?

[lokken]

I know this is not technical question but I wonder the future of this work. Because new processors like ARM7 will be more popular and they are faster than most microcontroller like PIC. At this point, I think CCSC should add new microcontroller to their library (like Arm based microcontroller...)
What do you think about this question? Is this question not realistic or silly?

[Ttelmah]

You are speaking as if the PIC, was one generation of processors. It isn't. You have PICs ranging from the PIC10 family, through PIC16, PIC18, PIC24, PIC30 and PIC32 etc..

However, you have to remember though that in many applications 'performance', is _not_ the key equation. Power consumption, and cost are two other primary factors. I have a couple of projects using really low power members of the RFPIC family, where shipping totals are over 10^6 units per year. As with things like nuts and bolts, people will use much more sophisticated fastenings if needed, but will still be buying basic nuts for the forseeable future. Some models of the PIC, are exactly this in computer terms. The basic architecture at the core of the original PIC, allows reasonable performance using low die sizes, and without the overhead of memory caching etc.. Once you go faster, other architectures are 'better', so it would be foolish to try to force the original PIC architecture to grow into something much more powerful. The fastest 'PICs', are not really based on the the original architecture at all, but are MIPS based.

Now, the big pressure coming at present is not PIC versus ARM. Both are good at some things and have weaknesses in others, but the realisation that the biggest development environment in existence, is the PC, and as such for larger projects needing this sort of level of computer 'power', the Arduino concept is gaining ground fast, and support for this for the ARM, has appeared, giving this family a real 'leg up' in terms of acceptance. Unfortunately the same development from Microchip, is still 'unseen', so the larger PICs look to be falling yet further behind....

Microchip already has a competitor for the ARM (some parts of the PIC32 are faster than the ARM equivalent), but are losing ground, because the support is not 'there' in the way it exists for the ARM core.

Best Wishes

[jbmiller]

A couple of other things to consider is the developement costs in switching to 'yet another processor'. There can be a tremendous learning curve($$$) to the new 'chip', and if you're an old guy like me it gets harder everyday ! I've settled on PICs after 1802, Z80, 6800, 68HC11, and a few more. The HC11 was the killer. By the time I'd gotten the code cut, boards laid out, R&D,etc. just ready for production levels (1,000) I was told by Motorola I wasn't on the 'blue list' and could NOT order parts. Hmppfh! Called Microchip, went to a seminar and haven't found a thing I can't do with a PIC. Almost 20+ years worth of assembler and C code has convinced me to NOT go to another manufacturer.
As well I can get DIP packages ! Something I really appreciate, great for R&D, now that I have bifocals !! I've never understood the 'make it smaller, no smaller, I mean really, really smaller' push. But that's another story.....

[SherpaDoug]

[newguy]

When I was in university one my profs baited us by asking "what is the best processor available now?" A lot of people said the latest & greatest Intel offering along with a few others. Then he slammed us with "Is it suitable for a remote control? How about a programmable thermostat?"

That made us realize that "latest & greatest" doesn't mean squat for a small, cheap, and battery powered application. From then on we all learned that we have to ask "what's the application?"

Same sentiment applies here.

[mkuang]

[Jerson]

I was forced into a MSP430 project about a year ago. I found the learning curve short and sweet, but, once stumped, even TI was no help. I was toying with the 16bit Delta-Sigma ADC being oversampled to 20bit. It refused to work as advertised in the app note. Finally, the customer just upped and went away. MSP430 is neat in its own way, but, support is very very important at least when starting fresh.

I am a freelancer and from the kind of work that is being asked for, I can say very confidently the PIC rules today. ARM has a bright future and is slowly covering territory once dominated by the 8051. Advantage of ARM is it is multisource just like the 8051.

The opinion expressed above is my perception and you most certainly may disagree with it.
_________________
Regards
Jerson Fernandes

[bkamen]

I did an SCH/PCB only gig for a client recently that use an AVR USB IC that I found limiting.

To get the memory they wanted, they had to use a 64pin Exposed Pad 64QFN-EP.

They also wanted the design to be SMALL and cheap (think 2 layer PCB). THis can be hard with the EP that wastes all that real-estate under the IC.

When I finally got to making some proto's.. I found the version AT90USB to be hard to find, low quantities and long lead times.

I reported this to the client and he asked if there was any solutions.

I looked back at trusty Microchip and they made a part with the same specs in various footprints (as they usually do in the PIC18F line) that had all the memory, better low power (project was running on 3V Li coin), and a smaller footprint.

Smaller footprint == lower cost.

If we changed the design to MCHP, not only could we get parts, we could save space and power and eliminate the 3.3V DC-DC boost converter that was needed for the AT90USB. A total savings of about 61% of the Atmel version when it was all calculated.

Every time I venture to look at Atmel (about every 3-4yrs), I'm just not impressed.

Anyway -- this was the perfect example of how smaller, more efficient and lower cost won the day. No 32bit MCU needed here.
_________________
Dazed and confused? I don't think so. Just "plain lost" will do. :D