Author Topic: Microchip to acquire Atmel - how do you feel about this?  (Read 37353 times)

0 Members and 1 Guest are viewing this topic.

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 7373
  • Country: us
    • SiliconValleyGarage
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #25 on: January 15, 2016, 08:22:07 pm »
I think they should prefix all Atmel part numbers with PIC, just to piss off the fanboys.

+1
and relabel all PIC with AVR.

AVR18F2450
PIC328

Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 19369
  • Country: nl
    • NCT Developments
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #26 on: January 15, 2016, 08:45:36 pm »
I'm not feeling anything since both companies make crappy, bottom of the barrel products and therefore I don't use their parts in any of my designs. Over the years (decades) Microchip and Atmel products have proven to me to have higher (hidden) costs in the long run than the low price of the devices make up for.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline c4757p

  • Super Contributor
  • ***
  • Posts: 7805
  • Country: us
  • adieu
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #27 on: January 15, 2016, 10:32:34 pm »
The PIC16 architecture is getting a little silly, what with them now adding USB to a microcontroller that has only one working register and crazy paging rules. I wouldn't mind seeing a PIC16-style device (with all of the great PIC peripherals) but with an AVR single-cycle core at the heart & 16 working regs.
AVR CPU & PIC peripherals would be a top-notch 8-bit micro family.

That's an interesting idea. I doubt they'd actually do that, but I'd not mind that at all. PIC's CPU is the single most shite 8-bit CPU I've ever had the displeasure of using, and Atmel's peripherals are, well, boring. Swap 'em around the other way....
No longer active here - try the IRC channel if you just can't be without me :)
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4829
  • Country: gb
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #28 on: January 15, 2016, 10:43:23 pm »
The PIC16 architecture is getting a little silly, what with them now adding USB to a microcontroller that has only one working register and crazy paging rules. I wouldn't mind seeing a PIC16-style device (with all of the great PIC peripherals) but with an AVR single-cycle core at the heart & 16 working regs.
AVR CPU & PIC peripherals would be a top-notch 8-bit micro family.

That's an interesting idea. I doubt they'd actually do that, but I'd not mind that at all. PIC's CPU is the single most shite 8-bit CPU I've ever had the displeasure of using, and Atmel's peripherals are, well, boring. Swap 'em around the other way....

Indeed, but let us not forget that this shite architecture is no less than 40 years old now and still selling like hotcakes.

But you're right, it's all about the peripherals.
 

Offline c4757p

  • Super Contributor
  • ***
  • Posts: 7805
  • Country: us
  • adieu
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #29 on: January 15, 2016, 10:52:22 pm »
I don't really want to get a PIC/AVR flamewar going, but really, has anyone ever chosen PIC for the CPU? Nah... Microchip does decent peripherals. It's either that or familiarity.

Anyone who thinks the CPU is good is invited to address the four-clock instruction cycle plus relatively long cycles-per-instruction for many common opcodes, and the ghastly memory access methods that make compiler writing extremely difficult. >:D

I'd love to scoop out an AVR CPU and wrap it in some Microchip peripherals. Maybe an M0 core too. Sounds awesome. :-+

</dream>
No longer active here - try the IRC channel if you just can't be without me :)
 

Online VK3DRB

  • Super Contributor
  • ***
  • Posts: 1710
  • Country: au
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #30 on: January 16, 2016, 01:16:44 am »
I'm not feeling anything since both companies make crappy, bottom of the barrel products and therefore I don't use their parts in any of my designs. Over the years (decades) Microchip and Atmel products have proven to me to have higher (hidden) costs in the long run than the low price of the devices make up for.

That's why they are so successful.

If you write your code properly and do a robust hardware design around the chips, what's you problem? The only thing I dislike is the price of Atmels debuggers, and Microchip's mediocre IDE. Their chips are fine. Their documentation is pretty good.


 

Online VK3DRB

  • Super Contributor
  • ***
  • Posts: 1710
  • Country: au
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #31 on: January 16, 2016, 01:28:41 am »
Want to design a simple low cost high volume device? Microchip. Want CPU performance but higher cost is not an issue? Atmel. Want to be able to source the same chip is 15 years time? Microchip. Want decent support? Microchip.
Its horses for courses. You use the chip most appropriate for the application, cost, availability, support etc.

Anyone who can program one brand in C can program the other brand in C. Same with the plethora of other brands/architectures around. But the designer should consider what quality of tools there are.

For Atmel ATxmega etc, I really like Codevision AVR, a brilliant IDE written by a brilliant engineer in Romania. The registers are SO EASY to set up within your code, to get you going quickly. So for the cost factor, Codevision AVR can save you heaps of labour by avoiding sifting through a ton of register settings. It would be nice if all chips had such a nice IDE.

 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 31232
  • Country: au
    • EEVblog
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #32 on: January 16, 2016, 03:22:39 am »
For Atmel ATxmega etc, I really like Codevision AVR, a brilliant IDE written by a brilliant engineer in Romania. The registers are SO EASY to set up within your code, to get you going quickly. So for the cost factor, Codevision AVR can save you heaps of labour by avoiding sifting through a ton of register settings. It would be nice if all chips had such a nice IDE.

Yes, Codevision is very nice.

My latest project is PIC 24F, so I've got to go back and learn the new MPLABX. Haven't touched PIC since pre-MPLABX.
 

Offline Someone

  • Super Contributor
  • ***
  • Posts: 2442
  • Country: au
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #33 on: January 16, 2016, 06:31:38 am »
For Atmel ATxmega etc, I really like Codevision AVR, a brilliant IDE written by a brilliant engineer in Romania. The registers are SO EASY to set up within your code, to get you going quickly. So for the cost factor, Codevision AVR can save you heaps of labour by avoiding sifting through a ton of register settings. It would be nice if all chips had such a nice IDE.

Yes, Codevision is very nice.
And they have a free version that is very capable, a great step beyond Arduino.
 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 19369
  • Country: nl
    • NCT Developments
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #34 on: January 16, 2016, 06:52:12 am »
I'm not feeling anything since both companies make crappy, bottom of the barrel products and therefore I don't use their parts in any of my designs. Over the years (decades) Microchip and Atmel products have proven to me to have higher (hidden) costs in the long run than the low price of the devices make up for.

That's why they are so successful.

If you write your code properly and do a robust hardware design around the chips, what's you problem? The only thing I dislike is the price of Atmels debuggers, and Microchip's mediocre IDE. Their chips are fine.
Just compare the specs of their chips to A-brands like TI or Analog. You'll see the Microchip ones are always worse. The hidden costs are in stupid things like chips not behaving according to specifications, little immunity to external noise, logic levels which are choosen stupidly, etc, etc. For example: if you use an ADC from TI or Analog and power it from 5V then most ones will be suitable for use with 3.3V or lower logic levels. Now look at the ones from Microchip  :palm:. Atmel ditto in other places BTW.
« Last Edit: January 16, 2016, 06:55:20 am by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 12123
  • Country: gb
    • Mike's Electric Stuff
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #35 on: January 16, 2016, 11:49:05 am »
My latest project is PIC 24F, so I've got to go back and learn the new MPLABX. Haven't touched PIC since pre-MPLABX.

I'm still using the last  MPLAB 8.xx which supports a lot of PIC24 and PIC32 devices so see if it supports your device before jumping!
For new devices that only have minor differences like memory size (e.g. PIC32MX170) it's not hard to tweak device files to add support.
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline Stonent

  • Super Contributor
  • ***
  • Posts: 3824
  • Country: us
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #36 on: January 16, 2016, 12:30:32 pm »
Though the latest Atmel Studio is based on Visual Studio 2015 which I like.  I guess it's what you're used to, but I never really cared for Eclipse or any of the other IDEs.
The larger the government, the smaller the citizen.
 

Offline dannyf

  • Super Contributor
  • ***
  • Posts: 8229
  • Country: 00
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #37 on: January 16, 2016, 12:31:41 pm »
Quote
still selling like hotcakes.

For a business, what matters most isn't how it is selling now. Instead, how it is going to sell in the future.

Future looks quite bleak for Microchip.
================================
https://dannyelectronics.wordpress.com/
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4829
  • Country: gb
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #38 on: January 16, 2016, 01:18:13 pm »
My latest project is PIC 24F, so I've got to go back and learn the new MPLABX. Haven't touched PIC since pre-MPLABX.

I'm still using the last  MPLAB 8.xx which supports a lot of PIC24 and PIC32 devices so see if it supports your device before jumping!
For new devices that only have minor differences like memory size (e.g. PIC32MX170) it's not hard to tweak device files to add support.

I switch between both 8.92 and X on a daily basis, I still prefer 8.92 as it's far faster in the edit-compile-program-debug cycle despite its little character traits. X I use for the much newer chips that can't be hacked easily into 8 and occasionally to generate a set of #pragma config source code which I copy an paste into my own code: I can't be doing with automated code generators which all too easily overwrite your own careful handiwork and write excessive and largely useless commentary.
 

Offline MT

  • Super Contributor
  • ***
  • Posts: 1290
  • Country: cn
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #39 on: January 16, 2016, 01:42:56 pm »
Useless and no commentary's at all are one of ST's specialties.
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4829
  • Country: gb
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #40 on: January 17, 2016, 08:56:13 am »
Quote
still selling like hotcakes.

For a business, what matters most isn't how it is selling now. Instead, how it is going to sell in the future.

Future looks quite bleak for Microchip.

If you've ever had to deal with business cash flow and accounting, you'll know that's not quite true: cash is king.

While I agree it does seem counter intuitive and perhaps even absurd that we're still using these 8 bit devices, and that Microchip is still investing in them, to paraphrase Mark Twain, reports of the death of 8 bit have been greatly exaggerated. Repeatedly over many years now the death of 8 bit has been predicted but it's never quite happened.

A little frustrating is that the 16 bitters that Microchip have produced continually lag behind in the peripherals that the 8 bitters have. The 16 bit base architecture (ignoring the dsPIC for now) is way better than the 40 year old 8 bit, and is a joy to use in comparison.

The dsPIC on the other hand needs a lot of work on the part of the programmer to take advantage of the fixed point ALU, there is no means to automatically leverage its use with the compiler without explicit use of builtins or assembly language, or to use pre-baked library routines. With cheap M4F devices with ever improving peripherals it's hard to see where the dsPIC will fit.

In applications, the PIC32MX is just as good as any ARM M0 or M3. Unfortunately the powers that be have chosen to take a rather bizarre direction in the software support, using excessive abstraction and unnecessarily implementing a proprietary software framework more at home with a heavy weight OS than a microcontroller. Luckily on the MX series the old MLA library is still available but is deprecated, although it's an organically derived mess. The new MZ devices pretty much force you into using the new framework. To add to their woes, the first MZ C devices were of appalling quality, although more recently the MZ F devices are far better in that regard. As a competitor to M4F and M7, the PIC32MZ F is certainly a worthy contender.

 

Offline XFDDesign

  • Frequent Contributor
  • **
  • Posts: 435
  • Country: us
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #41 on: January 18, 2016, 05:30:05 am »
On the "would anyone pic a PIC for the CPU?" front, I have. I specifically got into the PIC32MZ line as it has a very low barrier to entry over anything arm. Every time I tried to get into ARM, I found it to be a completely rooted venture. Compilers are too expensive (GCC has been a nightmare trying to get to work), Ive not found any programmers for any of the litany of parts out there, and the software side of dev seems to have so much overhead to appease the part, I'd rather get audited by the IRS.

PIC32? Open the box and write my own code to twiddle a pin in less than 30 minutes. It's a capable processor, a reasonable amount of horsepower, and of course the great peripheral list (though I still use ADI converters over the "10bit" ADCs).
 

Offline benSTmax

  • Regular Contributor
  • *
  • Posts: 61
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #42 on: January 18, 2016, 10:30:05 am »
On the "would anyone pic a PIC for the CPU?" front ...
I think this was referring only to the 8-bit PIC cores, which are quite an awkward architecture, making the job of C a compiler quite difficult.
On the other hand, the 16-bit PIC cores (e.g. PIC24, dsPIC) are nice architectures and in my view they resemble more with the AVR rather than the 8-bit PIC cores.
The PIC32 and their MIPS cores are modern and they are an alternative to ARMs.
 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 19369
  • Country: nl
    • NCT Developments
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #43 on: January 18, 2016, 10:41:03 am »
@XFDDesign: You are not making a strong case for your abilities to figure something out.  >:D There are so many tutorials out there to get an ARM microcontroller going...
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Online AndyC_772

  • Super Contributor
  • ***
  • Posts: 3592
  • Country: gb
  • Professional design engineer
    • Cawte Engineering | Reliable Electronics
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #44 on: January 18, 2016, 11:09:51 am »
Every time I tried to get into ARM, I found it to be a completely rooted venture. Compilers are too expensive (GCC has been a nightmare trying to get to work), Ive not found any programmers for any of the litany of parts out there, and the software side of dev seems to have so much overhead to appease the part, I'd rather get audited by the IRS.

Giving up on ARM is a bad idea in the long run, I'm afraid. The microcontroller world is getting ever closer to standardising on Cortex-M, and having the ability to use this enormous range of parts is absolutely necessary for an embedded developer, IMHO.

Download yourself a trial version of CrossWorks and see how you get on. After days spent trying to get Eclipse up and running, searching out "helpful" tutorials that were inevitably outdated, misleading or just plain wrong, I found it an absolute breeze.

Offline dannyf

  • Super Contributor
  • ***
  • Posts: 8229
  • Country: 00
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #45 on: January 18, 2016, 12:09:52 pm »
Quote
Every time I tried to get into ARM, I found it to be a completely rooted venture. Compilers are too expensive (GCC has been a nightmare trying to get to work), Ive not found any programmers for any of the litany of parts out there, and the software side of dev seems to have so much overhead to appease the part, I'd rather get audited by the IRS.

Two points:

1) you don't quite program the ARM chip, in that you are more than likely programming in C so the stuff that's the ARM core is fully transparent to you, just as if you were programming a PIC/AVR or a PC for that matter, with some minor differences of course.

2) the initial difficulties are with the software, both software necessary to get a project set up properly and software necessary to get it to do what you need. The first part is quite ARM- and IDE-specific: you are likely to use CMSIS, and many times vendor-supplied, chip specific start-up / system files. Some IDEs help a lot in this regard, like the new middleware in Keil, or CoIDE. The 2nd part relates to your approach to programming any powerful chip: do you use a library or not. Vendor supplied libraries can be a pain to learn for a beginner and requires a lot of reading and practicing and tracing the code vs. the datasheet, and a level of C-programming skills. This part is fully programmer dependent. If you cannot crack it, you cannot crack it. Not much anybody else can do to help you.

Overall, I have found the use of a good IDE like CoIDE quite helpful in getting it going. Otherwise, I have chip-specific templates that I can copy-and-paste to a new project. Coupled with my own middleware, I can get a project going with seconds. The downside with this approach is the large number of files that need to be compiled initially: a minimalist project for me on a STM32F3 is ~25MB and over 1000 source / header files, and the initial compilation takes a couple minutes on a fairly fast workstation, using Keil.

The advantage is obvious that it is platform dependent: I can copy my user code to different chips and fitted with the right library and middleware, the project is ready to go and just works.

If you are starting out, I would suggest that you take a look at CoIDE - I am a big fan of their products (pre 2.0). Select a chip, select functionalities you want on that chip, and hit compile and you are ready to go. Absolutely a delight to use. the 2.0 beta, however, is a total disappointment.

================================
https://dannyelectronics.wordpress.com/
 

Online andersm

  • Super Contributor
  • ***
  • Posts: 1139
  • Country: fi
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #46 on: January 18, 2016, 12:30:09 pm »
On the other hand, the 16-bit PIC cores (e.g. PIC24, dsPIC) are nice architectures and in my view they resemble more with the AVR rather than the 8-bit PIC cores.
IMO they have the same problem as all 16-bitters, namely that you run out of address space even quicker than on an 8-bit CPU, and then have to start adding ugly kludges that don't play well with C and overhead.

Quote
The PIC32 and their MIPS cores are modern and they are an alternative to ARMs.
PIC24 is far more modern (or at least newer) than MIPS! Personally I don't think MIPS make a great microcontroller core. There's too much historical baggage for compatibility, and the large register file consumes more stack and increases latencies. Cortex-M is much more elegant, and since it's a completely new architecture, the designers have been free to innovate. It's possible the architecture doesn't scale as well, but for high-end applications there's Cortex-A and Cortex-R.

Offline dannyf

  • Super Contributor
  • ***
  • Posts: 8229
  • Country: 00
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #47 on: January 18, 2016, 12:51:35 pm »
pic24 is the part that microchip should have marketed earlier and more.

What a shame.
================================
https://dannyelectronics.wordpress.com/
 

Offline dannyf

  • Super Contributor
  • ***
  • Posts: 8229
  • Country: 00
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #48 on: January 18, 2016, 12:52:52 pm »
I suspect that their reluctance to market pic24 is to protect their 8-bit pics - a golden goose then.

Unfortunately, as the market moved on, it is that very golden goose that killed them / their desire to innovate.
================================
https://dannyelectronics.wordpress.com/
 

Offline coppice

  • Super Contributor
  • ***
  • Posts: 5352
  • Country: gb
Re: Microchip to acquire Atmel - how do you feel about this?
« Reply #49 on: January 18, 2016, 12:54:37 pm »
pic24 is the part that microchip should have marketed earlier and more.

What a shame.
They tried pretty hard, but Microchip have never been that successful with larger MCUs. Their big successes have always been in the low pin count market.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf