• AVR Freaks

Helpful ReplyHot!Is MCHP still developing the PIC32?

Page: < 12345.. > >> Showing page 4 of 10
Author
jdeguire
Super Member
  • Total Posts : 599
  • Reward points : 0
  • Joined: 2012/01/13 07:48:44
  • Location: United States
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/04 11:22:38 (permalink)
0
Jack_M
 
Would also be interested to know :)
Only parts i know of -besides the release date, obviously- are the updates to PIC32MK (which alongside the new flash and can controllers promise to solve all present major issues from I2C not working to just a single SRS hence interrupts that are stupid slow)
 
I'm also in the faction of don't care about MIPS core, i just want MCHP peripherals in a 32bit part that is not laughably slower than a dsPIC

Do you happen to know if the updated PIC32MK would include Ethernet?  None of the current ones do from what I recall, but a PIC32MK with Ethernet and CAN-FD could be a great part for our needs.
 
I don't directly care about MIPS32 vs. Cortex-M, but I do generally care about CPU performance (hence the benchmarks I posted earlier).  The SAME54 is a nice bump over the PIC32MX795 we currently use and the SAME70 is a huge bump over the PIC32MZEF we currently use.  The nicer interrupt handling on the Cortex-M cores is a nice bonus as well.  If my previous tests are actually correct, then the PIC32MK should be faster than the SAME54 even though they both run at 120MHz and the PIC32MK has a full 64-bit FPU vs. the 32-bit FPU in the Cortex-M4.  Truthfully, though, either part would probably work well enough for a large portion of our needs.
 
I personally don't care about whose peripherals are in the chip as long as they cover what we need (Ethernet, CAN-FD, ADC, the serial protocols) and they actually work.  That said, I totally get why one would want to stick with what they know and have since trying to deal with a variety of peripherals would make code maintenance more difficult.  We wrap peripheral code in C++ classes and that has been effective, but that is still extra code to deal with.
 
friesen
I just wish that 0.1% of the cost to take 1 chip to fab could be redirected to the complete ide to debugger ecosystem.  I have even considered building something from that ground up using mips toolchain + netbeans or some other ide that could handle makefiles.

One advantage to the Arm parts is they they use a more modern version of GCC as part of XC32.  The MIPS version is a CodeSourcery variant of GCC 4.8.3.  The compiler itself supports C++11, but the Dinkumware libraries included with it do not (std::atomic is not suported at all, for example).  Personally, I don't get why Microchip paid for those libraries, but I'm no expert in that area.  The Arm compiler uses GCC 6.2.1 with Newlib and libstdc++.  They all support C++14 and C11 (I think, may be C99 still), which would be nice for our needs.
 
My uses of the debugger on the Arm devices has been limited to the Xplained board onboard debuggers (EDBG), but I can't say I noticed a huge difference between the Arm and MIPS devices in the experience.  I've generally suspected that many issues I've seen are because of MPLAB X, though, and not the debug tool itself.  It does seem like newer versions of MPLAB X do better, so I think they are working on it.
 
If you did want to use your own toolchain, you can download the MPLAB X SDK on www.opensource4pic.org and follow the documentation to make a plugin for MPLAB X.  The docs are pretty out of date, but they also include the sources for several of their toolchain plugins you can use as a guide.
post edited by jdeguire - 2020/06/04 11:25:24
#61
JPortici
Super Member
  • Total Posts : 1175
  • Reward points : 0
  • Joined: 2012/11/17 06:27:45
  • Location: Grappaland
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/04 15:00:54 (permalink)
0
No ethernet that i know of.
In the current version of the GP/MC datasheet there is a migration guide that shows the differences between the current and future parts... try to open the device include files in the compiler for the new parts (they are already there) and see if there is a mention of ethernet..
 
Sticking with peripherals: Part of it is keeping using what you know (and that's really important in complex stuff like CANbus, you know the gotchas.) and part of it is keeping using what you know you will work for you.
For example, i usually need A LOT of truly independent IC and OC. Not many parts outside of dsPIC33EV,dsPIC33EP(MU),dsPIC33CK and PIC32MK that will fit my needs.
 
or some combinations of peripherals: i mentioned it before, PIC32MX is still among the cheapest chips with both CAN and USB (STM32 F1 doesn't count. You have both but can only use one at a given time)
 
or peripheral interconnection that you are used to, DMA, whatever else..
post edited by JPortici - 2020/06/04 15:06:25
#62
Sal Ammoniac
Super Member
  • Total Posts : 234
  • Reward points : 0
  • Joined: 2011/11/27 14:04:14
  • Location: 0
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/06 16:37:54 (permalink)
5 (1)
NorthGuyTechnically, the difference between MIPS, ARM, and RISC-V is very small. RISC-V is somewhat better because it was designed recently, while MIPS, and especially ARM have lots of legacy stuff. From the marketing viewpoint though RISC-V might be huge - ARM has been popular for far too long and is now losing its marketing luster.


Since Microchip already has ARM-based products (via the Atmel acquisition), perhaps what they should do is de-emphasize further development of the MIPS-based PIC32 line and concentrate on the ARM line while simultaneously developing a RISC-V based PIC32 line (perhaps calling it PIC32V) to replace the MIPS parts.
#63
jdeguire
Super Member
  • Total Posts : 599
  • Reward points : 0
  • Joined: 2012/01/13 07:48:44
  • Location: United States
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/08 11:19:28 (permalink)
0
I'd expect that Microchip would do the first part of your idea--sunset development on the MIPS devices--but I can't see them developing a RISC-V device anytime soon.  I say this only because I can see the company not wanting to support another architecture.  The current Arm devices already cover the PIC32's performance and peripheral range and extend beyond at the high-end with the Cortex-M7 and Cortex-A5 parts, so Microchip may not see value in introducing another CPU architecture if the current one they have already covers the market that want to cover.  The Cortex-A5 devices have display controllers and the ARM9 devices have 2D GPUs as well.
 
I'm also not sure how important 32-bit is to Microchip in general.  I've gotten the impression over time that 8-bit is still their bread and butter, but that's only conjecture on my end.
 
That said, I am interested to learn more about RISC-V just for my own curiosity.  This excerpt from Wikipedia really caught my eye:  
 
Unlike ARM's Thumb or the MIPS compressed set, space was reserved from the beginning so there is no separate operating mode. Standard and compressed instructions may be intermixed freely.
#64
Sal Ammoniac
Super Member
  • Total Posts : 234
  • Reward points : 0
  • Joined: 2011/11/27 14:04:14
  • Location: 0
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 08:46:57 (permalink)
5 (1)
IMO, the biggest reason for MCHP to develop a RISC-V based MCU is to avoid being caught out if RISC-V takes off and starts to displace ARM in popularity. It's better to be one of the early adopters than to play catch-up with the likes of ST, NXP, and anyone else who develops RISC-V MCUs early in the adoption cycle.
 
A secondary reason is to reduce licensing costs.
#65
NKurzman
A Guy on the Net
  • Total Posts : 18976
  • Reward points : 0
  • Joined: 2008/01/16 19:33:48
  • Location: 0
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 09:12:06 (permalink)
0
Because the. RISC-V is the hot CPU in the Embedded Space?
You are assuming the RISC-V will push everyone else out of the Market. Being an early adopter is great is the Product wins, not so much if it loses.  Right now customer want the ARM.  the RISC-V will need to convince people to switch to it.
#66
marcov
Super Member
  • Total Posts : 285
  • Reward points : 0
  • Joined: 2006/10/08 01:59:40
  • Location: Eindhoven, NL.
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 09:23:06 (permalink)
4 (1)
Sal Ammoniac
A secondary reason is to reduce licensing costs.



Generally, companies don't just buy a architecture license, but also a specific base design. Usually the design costs are higher than the architecture costs.
 
In the ARM space there used to be many companies that created the design themselves, but they have dwindled in recent years, since the CPUs are getting more and more complicated, and thus having an own design (and developing the next, and the next and ...) is getting prohibitorily expensive.
 
And that's the rub, the only Risc V design seller seems to be SiFive, so if their designs are more expensive (since they have less consumers to divide the costs), for all practical purposes Risc V is more expensive than eg.. MIPS.
#67
Sal Ammoniac
Super Member
  • Total Posts : 234
  • Reward points : 0
  • Joined: 2011/11/27 14:04:14
  • Location: 0
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 10:25:52 (permalink)
0
NKurzmanYou are assuming the RISC-V will push everyone else out of the Market.

 
I'm assuming no such thing.
 
Being an early adopter is great is the Product wins, not so much if it loses.  Right now customer want the ARM.  the RISC-V will need to convince people to switch to it.

 
RISC-V is building up momentum. Most of the big ARM MCU vendors are members of the RISC-V Foundation (Microchip is a Founding Member). It's better to be ready if the market goes in that direction than to be unprepared and play a catch up game.
 
#68
NorthGuy
Super Member
  • Total Posts : 6350
  • Reward points : 0
  • Joined: 2014/02/23 14:23:23
  • Location: Northern Canada
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 11:23:52 (permalink)
5 (1)
NKurzman
Being an early adopter is great is the Product wins, not so much if it loses.  Right now customer want the ARM.  the RISC-V will need to convince people to switch to it.



They have already missed the boat by going with MIPS for their PIC32, while they should've gone with emerging ARM. It'll be silly if they make the same mistake again now.
#69
NKurzman
A Guy on the Net
  • Total Posts : 18976
  • Reward points : 0
  • Joined: 2008/01/16 19:33:48
  • Location: 0
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 13:51:22 (permalink)
0
At the time, I had hoped they would go ARM, But it was not obvious at the time ARM would be the winner and MIPS the Loser.  Look at the PIC16-18.  How many Competitors have they defeated.  And Not because the PIC architecture was better.  They just bought in to ARM with Atmel.  Should they just grab a third 32 bit core (Harmony 4?) What will a Chip Buyer choose/ What makes me want a  RISC-V over an ARM.  will I be able to get the support and tools that ARM Provides?



#70
Sal Ammoniac
Super Member
  • Total Posts : 234
  • Reward points : 0
  • Joined: 2011/11/27 14:04:14
  • Location: 0
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 14:25:44 (permalink)
0
NKurzman
will I be able to get the support and tools that ARM Provides?



 
At this point in time, no. But expect that to change rapidly when one of the big players (NXP, ST, TI) introduces an MCU based on a RISC-V core.
#71
NorthGuy
Super Member
  • Total Posts : 6350
  • Reward points : 0
  • Joined: 2014/02/23 14:23:23
  • Location: Northern Canada
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 14:30:36 (permalink)
5 (1)
NKurzman
At the time, I had hoped they would go ARM, But it was not obvious at the time ARM would be the winner and MIPS the Loser.  Look at the PIC16-18.  How many Competitors have they defeated.  And Not because the PIC architecture was better.  They just bought in to ARM with Atmel.  Should they just grab a third 32 bit core (Harmony 4?) What will a Chip Buyer choose/ What makes me want a  RISC-V over an ARM.  will I be able to get the support and tools that ARM Provides?



They all have tools and support. There's nothing rational about this.
 
For some reason, recently, all the competing companies in certain space want to do the same thing. Meanwhile, it is possible to do something differently, say develop a new 32-bit core based on PIC24 and push it to the market as something unique, or develop a multi-core PIC with 16 or 32 crazy-fast PIC16 cores which can do practically anything - look what we've got - nobody can offer anything similar.
#72
Sal Ammoniac
Super Member
  • Total Posts : 234
  • Reward points : 0
  • Joined: 2011/11/27 14:04:14
  • Location: 0
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 15:36:19 (permalink)
0
NorthGuy
 
They all have tools and support.

 
True, but not at the same level as ARM tools support. It's getting there, though. IAR supports RISC-V in their Embedded Workbench suite. I doubt if Keil will ever support RISC-V as Keil is owned by ARM and they're not likely to want to support a competing ISA.
 
#73
MisterHemi
Super Member
  • Total Posts : 302
  • Reward points : 0
  • Joined: 2017/11/02 12:24:21
  • Location: Commerce, CA USA
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 15:40:30 (permalink)
0
NorthGuy
NKurzman
At the time, I had hoped they would go ARM, But it was not obvious at the time ARM would be the winner and MIPS the Loser.  Look at the PIC16-18.  How many Competitors have they defeated.  And Not because the PIC architecture was better.  They just bought in to ARM with Atmel.  Should they just grab a third 32 bit core (Harmony 4?) What will a Chip Buyer choose/ What makes me want a  RISC-V over an ARM.  will I be able to get the support and tools that ARM Provides?



They all have tools and support. There's nothing rational about this.
 
For some reason, recently, all the competing companies in certain space want to do the same thing. Meanwhile, it is possible to do something differently, say develop a new 32-bit core based on PIC24 and push it to the market as something unique, or develop a multi-core PIC with 16 or 32 crazy-fast PIC16 cores which can do practically anything - look what we've got - nobody can offer anything similar.




Yes, i'd like to see those things happen especially multi-core PICs and even a 64 bit architecture. 

My configuration:
MacBook Pro (Retina, 15-inch, Mid 2015) with MacOS Mojave (10.14.6) and MPLAB X IDE v5.30
 
Curiosity PIC32MZ EF 1 & 2, PIC24F Curiosity, XPRESS EVAL BOARD (PIC16F18855), SAMA5D3 Xplained and various custom boards.
#74
NorthGuy
Super Member
  • Total Posts : 6350
  • Reward points : 0
  • Joined: 2014/02/23 14:23:23
  • Location: Northern Canada
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 15:46:44 (permalink)
4 (1)
Sal Ammoniac
True, but not at the same level as ARM tools support. It's getting there, though. IAR supports RISC-V in their Embedded Workbench suite. I doubt if Keil will ever support RISC-V as Keil is owned by ARM and they're not likely to want to support a competing ISA.



Also, if Microchip decides to go to RISC-V, they will provide their own tools. RISC-V is very similar to MIPS, and Microchip's compilers are based on GCC, so this won't be difficult to do.
#75
NKurzman
A Guy on the Net
  • Total Posts : 18976
  • Reward points : 0
  • Joined: 2008/01/16 19:33:48
  • Location: 0
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 16:00:00 (permalink)
0
The question is what quirk of history makes a CPU popular.  And why does another fade away.
The Z80 , 68000, 8052 ( and yes I know they are not truly dead.
Look at the AVR would it be so popular without Ardunio?
The ARM was probably helpped by phones.  The X86 was helped by IBM, but it could have been a 68000 they say.  It was certainly a better CPU.  And what made the PIC16 so great? there were many better 8bit CPUs. Maybe the RISC-V is the next big thing. who knows.
 
why do all companies push in the same space?  It is where the money is going.  A 32Bit ARM is not a Budget buster of a CPU.  Plus these days Companies want LCD, IoT, Ethernet, USB.  SO a 32 bit CPU Fits.  And A lot of that looks like Linux.  Or Android.  The X86 manufacturers left the 32 bit space behind. ARM is still there.
#76
Stefiff
Senior Member
  • Total Posts : 99
  • Reward points : 0
  • Joined: 2012/07/15 15:26:29
  • Location: 0
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/09 21:09:51 (permalink)
2.5 (2)
In my opinion, it all comes down to money, as always. At the moment, everyone wants a lot of money, and right away. That's where the key is. No one is developing a product anymore that they will produce after a few years. Everyone wants to take something ready, and immediately start winning. Currently, all waiting for someone to release a ready-made core with RISC-V, and others will start using it. The maximum term for development in which investments are made is about half a year.
Samsung's new 5G modem is based on RISC-V core. They are unlikely to develop an arm modem anymore.
The new WD controller too.
Qualcomm, their last few cores are now completely standard arm. They gave up investing in new cores. This was not the case 2 years ago. Now they have moved their money at RISC-V.
The moment a mobile processor appears, the process will speed up.
 
The latest information is that Apple is abandoning Intel and switching to its own processors. They will announce it in 2 weeks. I wonder what the core will be. For now it will be some kind of ARM.
#77
NorthGuy
Super Member
  • Total Posts : 6350
  • Reward points : 0
  • Joined: 2014/02/23 14:23:23
  • Location: Northern Canada
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/10 06:00:43 (permalink)
5 (2)
NKurzman
And what made the PIC16 so great?



Simplicity. Everything great is simple.
#78
NKurzman
A Guy on the Net
  • Total Posts : 18976
  • Reward points : 0
  • Joined: 2008/01/16 19:33:48
  • Location: 0
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/10 09:48:21 (permalink)
0
Apple Bought Intel's Modem Division recently.
I think they Bought a CPU design company some years back.
Some of the RISC-V movement you are seeing is due US sanctions related to ARM.
#79
Keaton
Junior Member
  • Total Posts : 80
  • Reward points : 0
  • Joined: 2013/04/01 14:34:05
  • Location: Chandler AZ
  • Status: offline
Re: Is MCHP still developing the PIC32? 2020/06/11 00:25:09 (permalink)
0
NorthGuy
Keaton
The PIC32 brand is not dead, it is under current active development with new parts coming very soon



Would you share any information on the new parts?


Jack_M
Keaton
The PIC32 brand is not dead, it is under current active development with new parts coming very soon


Would also be interested to know :)
 

Sorry I cannot, as much as I wish I could. PIC32 is coming out with some new and exciting stuff very very soon, you just have to hang in there, it is coming. I promise!
 
I can set you up with the right marketing and field engineer to get any qualified Early adopter access, I would love to do that and help you.


 
#80
Page: < 12345.. > >> Showing page 4 of 10
Jump to:
© 2020 APG vNext Commercial Version 4.5