EmbeddedRelated.com
Forums

Atmel Bought by Microchip

Started by rickman June 28, 2016
Phil Hobbs wrote:
> On 06/29/2016 12:24 PM, Les Cargill wrote: >> rickman wrote: >>> Not sure how I missed this one for two months. I see there is already >>> some serious contention with Atmel employees. >>> >>> http://www.eetimes.com/document.asp?doc_id=1329412 >>> >>> I wonder what it will be like for the sales force to be selling PICs, >>> AVRs and ARMs all into much of the same market space? >> >> I think you're at the wrong end of the telescope. This is pretty clearly >> a consolidation in a falling market. Microchip doesn't appear to be >> able to conform to severance agreements. >> >>> I wonder if the >>> PIC32 will finally bite the dust with ARMs all around it crowding it out >>> of the market? >> >> I've never held that ARM was much in the same space as PIC32. At least >> the tooling is pretty radically different. And I can't say I'm too sad >> if AVR just goes away. > > It would be a pity for the Arduino folks. >
Perhaps I'm in error but I hold those as reasonably separate - I doubt they'd dump Arduino. The AVR stacks are pretty raunchy, and I've run into obsolescence problems with them.
>> >> The problem is that I don't have any direct exposure to the tiny ARM >> offerings, mainly because the peripheral mix chosen on projects I >> worked on fit PIC32 or PIC24 better. > > Some Cortex M0s are under half a buck at distributor prices, which is a > lot cheaper than an ATmega. >
This is also true. I haven't had the privilege yet.
>> >> Really, when we're saying ARM these days, it more means something >> big enough to build a cell phone on. > > Well, you don't know how the other half lives. ;)
I just meant as a function of population of chips.
>> >>> Or maybe the dsPIC will go away? >>> >> > > Cheers > > Phil Hobbs > >
-- Les Cargill
rickman wrote:
> On 6/29/2016 12:24 PM, Les Cargill wrote: >> rickman wrote: >>> Not sure how I missed this one for two months. I see there is already >>> some serious contention with Atmel employees. >>> >>> http://www.eetimes.com/document.asp?doc_id=1329412 >>> >>> I wonder what it will be like for the sales force to be selling PICs, >>> AVRs and ARMs all into much of the same market space? >> >> I think you're at the wrong end of the telescope. This is pretty clearly >> a consolidation in a falling market. Microchip doesn't appear to be >> able to conform to severance agreements. >> >>> I wonder if the >>> PIC32 will finally bite the dust with ARMs all around it crowding it out >>> of the market? >> >> I've never held that ARM was much in the same space as PIC32. At least >> the tooling is pretty radically different. And I can't say I'm too sad >> if AVR just goes away. > > Why would you say the ARMs aren't "in the same space as PIC32"? They > are both available at about the same performance levels, both come with > a range of memory and peripherals.
What we'd found is that the PIC32 fit better. The Cortex M had, in general, fewer peripherals. But this is gonna be an observer bias thing. I was also pushing for an Ethernet port, and it seems like PIC had a better story there.
> I think the main difference is there > is one company making PIC32s while there are dozens making ARM devices > for the same applications. >
This is also true. I just found the PIC series to be very easy to work with. -- Les Cargill
On 6/29/2016 1:59 PM, Les Cargill wrote:
> Phil Hobbs wrote: >> On 06/29/2016 12:24 PM, Les Cargill wrote: >>> rickman wrote: >>>> Not sure how I missed this one for two months. I see there is already >>>> some serious contention with Atmel employees. >>>> >>>> http://www.eetimes.com/document.asp?doc_id=1329412 >>>> >>>> I wonder what it will be like for the sales force to be selling PICs, >>>> AVRs and ARMs all into much of the same market space? >>> >>> I think you're at the wrong end of the telescope. This is pretty clearly >>> a consolidation in a falling market. Microchip doesn't appear to be >>> able to conform to severance agreements. >>> >>>> I wonder if the >>>> PIC32 will finally bite the dust with ARMs all around it crowding it >>>> out >>>> of the market? >>> >>> I've never held that ARM was much in the same space as PIC32. At least >>> the tooling is pretty radically different. And I can't say I'm too sad >>> if AVR just goes away. >> >> It would be a pity for the Arduino folks. >> > > Perhaps I'm in error but I hold those as reasonably separate - I doubt > they'd dump Arduino.
I don't think Arduino is much of a factor in the big picture. What matters is sales and Arduino is oneseys-twoseys. They care about sales of millions. I don't know what sort of profitability the AVR provides to Atmel, but I think it might be hard for Microchip to shut it down without doing damage to the bottom line.
> The AVR stacks are pretty raunchy, and I've run into obsolescence > problems with them.
Not sure what that means.
>>> The problem is that I don't have any direct exposure to the tiny ARM >>> offerings, mainly because the peripheral mix chosen on projects I >>> worked on fit PIC32 or PIC24 better. >> >> Some Cortex M0s are under half a buck at distributor prices, which is a >> lot cheaper than an ATmega. >> > > This is also true. I haven't had the privilege yet.
Not sure if this is a valid comparison. CM0s under a dollar are very limited devices. I can't imagine there aren't AVRs with similar capabilities at similar prices. -- Rick C
Tim Wescott wrote:
> On Wed, 29 Jun 2016 11:24:50 -0500, Les Cargill wrote: > >> >> The problem is that I don't have any direct exposure to the tiny ARM >> offerings, mainly because the peripheral mix chosen on projects I worked >> on fit PIC32 or PIC24 better. >> >> Really, when we're saying ARM these days, it more means something big >> enough to build a cell phone on. > > I disagree. The ARM A-core is for cell-phone sized stuff (and possibly > desktop). But the Cortex M-cores show up in some pretty small > processors. Anything that can be had for under a buck in quantity isn't > a "big processor". >
Yeah. So many processors, so little time. -- Les Cargill
On 6/29/2016 2:07 PM, Les Cargill wrote:
> rickman wrote: >> On 6/29/2016 12:24 PM, Les Cargill wrote: >>> rickman wrote: >>>> Not sure how I missed this one for two months. I see there is already >>>> some serious contention with Atmel employees. >>>> >>>> http://www.eetimes.com/document.asp?doc_id=1329412 >>>> >>>> I wonder what it will be like for the sales force to be selling PICs, >>>> AVRs and ARMs all into much of the same market space? >>> >>> I think you're at the wrong end of the telescope. This is pretty clearly >>> a consolidation in a falling market. Microchip doesn't appear to be >>> able to conform to severance agreements. >>> >>>> I wonder if the >>>> PIC32 will finally bite the dust with ARMs all around it crowding it >>>> out >>>> of the market? >>> >>> I've never held that ARM was much in the same space as PIC32. At least >>> the tooling is pretty radically different. And I can't say I'm too sad >>> if AVR just goes away. >> >> Why would you say the ARMs aren't "in the same space as PIC32"? They >> are both available at about the same performance levels, both come with >> a range of memory and peripherals. > > > What we'd found is that the PIC32 fit better. The Cortex M had, in > general, fewer peripherals. But this is gonna be an observer bias > thing. I was also pushing for an Ethernet port, and it seems like > PIC had a better story there.
I recall a discussion on this some years ago. I can't imagine this is still true in any meaningful way unless you have very unusual requirements. Let's face it, if the ARM CMx devices didn't cut the mustard in some way, they would not have taken off as they did. I also remember a conversation some years back where I forecast the rise of the ARM to dominate the MCU market. Some argued that the core didn't matter in a meaningful way, much more important were the peripherals which change between different makers even with the same CPU core. That is all true, but it has become irrelevant as users pick a CPU core and become comfortable with that. Even if they have to switch makers to get a peripheral they want, as long as the new devices is supported by the same tools they often are willing to do that. With the PIC32 you are stuck with one maker. I don't know how profitable the PIC32 is, but there are various costs associated with supporting a line and if the sales don't continue to justify it, the line can get dropped. Which do you think has the larger sales, the PIC32 or the ARM CMx devices? Microchip may not axe any of the MCU lines. But with this merger the first two to be considered I would think would be dsPIC and PIC32.
>> I think the main difference is there >> is one company making PIC32s while there are dozens making ARM devices >> for the same applications. >> > > This is also true. I just found the PIC series to be very easy to work > with.
Compared to what? -- Rick C
rickman wrote:
> On 6/29/2016 1:59 PM, Les Cargill wrote: >> Phil Hobbs wrote: >>> On 06/29/2016 12:24 PM, Les Cargill wrote: >>>> rickman wrote: >>>>> Not sure how I missed this one for two months. I see there is already >>>>> some serious contention with Atmel employees. >>>>> >>>>> http://www.eetimes.com/document.asp?doc_id=1329412 >>>>> >>>>> I wonder what it will be like for the sales force to be selling PICs, >>>>> AVRs and ARMs all into much of the same market space? >>>> >>>> I think you're at the wrong end of the telescope. This is pretty >>>> clearly >>>> a consolidation in a falling market. Microchip doesn't appear to be >>>> able to conform to severance agreements. >>>> >>>>> I wonder if the >>>>> PIC32 will finally bite the dust with ARMs all around it crowding it >>>>> out >>>>> of the market? >>>> >>>> I've never held that ARM was much in the same space as PIC32. At least >>>> the tooling is pretty radically different. And I can't say I'm too sad >>>> if AVR just goes away. >>> >>> It would be a pity for the Arduino folks. >>> >> >> Perhaps I'm in error but I hold those as reasonably separate - I doubt >> they'd dump Arduino. > > I don't think Arduino is much of a factor in the big picture. What > matters is sales and Arduino is oneseys-twoseys. They care about sales > of millions. I don't know what sort of profitability the AVR provides > to Atmel, but I think it might be hard for Microchip to shut it down > without doing damage to the bottom line. >
Entirely possible. The article tells a story that's very messy.
> >> The AVR stacks are pretty raunchy, and I've run into obsolescence >> problems with them. > > Not sure what that means. > >
I can't make it much clearer than that. The contributed software for things like USB are abysmal. USB is like that, but this was pretty bad. At a very slow-moving company, different revs of certain processors were quite incompatible.
>>>> The problem is that I don't have any direct exposure to the tiny ARM >>>> offerings, mainly because the peripheral mix chosen on projects I >>>> worked on fit PIC32 or PIC24 better. >>> >>> Some Cortex M0s are under half a buck at distributor prices, which is a >>> lot cheaper than an ATmega. >>> >> >> This is also true. I haven't had the privilege yet. > > Not sure if this is a valid comparison. CM0s under a dollar are very > limited devices. I can't imagine there aren't AVRs with similar > capabilities at similar prices. >
I would think there were. -- Les Cargill
rickman wrote:
> On 6/29/2016 2:07 PM, Les Cargill wrote: >> rickman wrote: >>> On 6/29/2016 12:24 PM, Les Cargill wrote: >>>> rickman wrote: >>>>> Not sure how I missed this one for two months. I see there is already >>>>> some serious contention with Atmel employees. >>>>> >>>>> http://www.eetimes.com/document.asp?doc_id=1329412 >>>>> >>>>> I wonder what it will be like for the sales force to be selling PICs, >>>>> AVRs and ARMs all into much of the same market space? >>>> >>>> I think you're at the wrong end of the telescope. This is pretty >>>> clearly >>>> a consolidation in a falling market. Microchip doesn't appear to be >>>> able to conform to severance agreements. >>>> >>>>> I wonder if the >>>>> PIC32 will finally bite the dust with ARMs all around it crowding it >>>>> out >>>>> of the market? >>>> >>>> I've never held that ARM was much in the same space as PIC32. At least >>>> the tooling is pretty radically different. And I can't say I'm too sad >>>> if AVR just goes away. >>> >>> Why would you say the ARMs aren't "in the same space as PIC32"? They >>> are both available at about the same performance levels, both come with >>> a range of memory and peripherals. >> >> >> What we'd found is that the PIC32 fit better. The Cortex M had, in >> general, fewer peripherals. But this is gonna be an observer bias >> thing. I was also pushing for an Ethernet port, and it seems like >> PIC had a better story there. > > I recall a discussion on this some years ago. I can't imagine this is > still true in any meaningful way unless you have very unusual > requirements.
I was dealing with a very slow organization, so the "many years ago" component dominates my observation.
> Let's face it, if the ARM CMx devices didn't cut the > mustard in some way, they would not have taken off as they did. >
It's hard to say, really. This is not a market in which "the market has spoken", and you even say why down the page a bit.
> I also remember a conversation some years back where I forecast the rise > of the ARM to dominate the MCU market. Some argued that the core didn't > matter in a meaningful way, much more important were the peripherals > which change between different makers even with the same CPU core.
Right. That mentality dominated what I'd seen.
> That > is all true, but it has become irrelevant as users pick a CPU core and > become comfortable with that. Even if they have to switch makers to get > a peripheral they want, as long as the new devices is supported by the > same tools they often are willing to do that. >
So there ya go.
> With the PIC32 you are stuck with one maker. I don't know how > profitable the PIC32 is, but there are various costs associated with > supporting a line and if the sales don't continue to justify it, the > line can get dropped. Which do you think has the larger sales, the > PIC32 or the ARM CMx devices? Microchip may not axe any of the MCU > lines. But with this merger the first two to be considered I would > think would be dsPIC and PIC32. >
Could be.
>>> I think the main difference is there >>> is one company making PIC32s while there are dozens making ARM devices >>> for the same applications. >>> >> >> This is also true. I just found the PIC series to be very easy to work >> with. > > Compared to what? >
AVR. Again, I will eventually see a Cortex Mx but haven't had the privilege yet. -- Les Cargill
On 06/29/2016 02:17 PM, rickman wrote:
> On 6/29/2016 2:07 PM, Les Cargill wrote: >> rickman wrote: >>> On 6/29/2016 12:24 PM, Les Cargill wrote: >>>> rickman wrote: >>>>> Not sure how I missed this one for two months. I see there is already >>>>> some serious contention with Atmel employees. >>>>> >>>>> http://www.eetimes.com/document.asp?doc_id=1329412 >>>>> >>>>> I wonder what it will be like for the sales force to be selling PICs, >>>>> AVRs and ARMs all into much of the same market space? >>>> >>>> I think you're at the wrong end of the telescope. This is pretty >>>> clearly >>>> a consolidation in a falling market. Microchip doesn't appear to be >>>> able to conform to severance agreements. >>>> >>>>> I wonder if the >>>>> PIC32 will finally bite the dust with ARMs all around it crowding it >>>>> out >>>>> of the market? >>>> >>>> I've never held that ARM was much in the same space as PIC32. At least >>>> the tooling is pretty radically different. And I can't say I'm too sad >>>> if AVR just goes away. >>> >>> Why would you say the ARMs aren't "in the same space as PIC32"? They >>> are both available at about the same performance levels, both come with >>> a range of memory and peripherals. >> >> >> What we'd found is that the PIC32 fit better. The Cortex M had, in >> general, fewer peripherals. But this is gonna be an observer bias >> thing. I was also pushing for an Ethernet port, and it seems like >> PIC had a better story there. > > I recall a discussion on this some years ago. I can't imagine this is > still true in any meaningful way unless you have very unusual > requirements. Let's face it, if the ARM CMx devices didn't cut the > mustard in some way, they would not have taken off as they did. > > I also remember a conversation some years back where I forecast the rise > of the ARM to dominate the MCU market. Some argued that the core didn't > matter in a meaningful way, much more important were the peripherals > which change between different makers even with the same CPU core. That > is all true, but it has become irrelevant as users pick a CPU core and > become comfortable with that. Even if they have to switch makers to get > a peripheral they want, as long as the new devices is supported by the > same tools they often are willing to do that. > > With the PIC32 you are stuck with one maker. I don't know how > profitable the PIC32 is, but there are various costs associated with > supporting a line and if the sales don't continue to justify it, the > line can get dropped.
That's one of the great things about Microchip historically though--they virtually never leave you in the lurch. You can still buy PIC17C756es from 20 years ago, that never sold well. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
On 6/28/2016 3:55 PM, rickman wrote:
> Not sure how I missed this one for two months. I see there is already > some serious contention with Atmel employees. > > http://www.eetimes.com/document.asp?doc_id=1329412 > > I wonder what it will be like for the sales force to be selling PICs, > AVRs and ARMs all into much of the same market space? I wonder if the > PIC32 will finally bite the dust with ARMs all around it crowding it out > of the market? Or maybe the dsPIC will go away?
After Cypress dropped their bid for Atmel, and Atmel spurned Dialog, I thought that Intel might swoop in. Intel has been making noise about getting back into the embedded market, where they once had a huge presence with the 8051 product line (and before that the 804x line. They finally gave up on the idea of the x86 architecture in embedded. All these M&As of companies doing embedded stuff and Intel is left out. I think I'll go do a COP8 design.
On 06/29/2016 01:32 PM, sms wrote:
> On 6/29/2016 9:59 AM, bitrex wrote: >> On 06/29/2016 12:46 PM, Phil Hobbs wrote: >>> On 06/29/2016 12:24 PM, Les Cargill wrote: >>>> rickman wrote: >>>>> Not sure how I missed this one for two months. I see there is already >>>>> some serious contention with Atmel employees. >>>>> >>>>> http://www.eetimes.com/document.asp?doc_id=1329412 >>>>> >>>>> I wonder what it will be like for the sales force to be selling PICs, >>>>> AVRs and ARMs all into much of the same market space? >>>> >>>> I think you're at the wrong end of the telescope. This is pretty >>>> clearly >>>> a consolidation in a falling market. Microchip doesn't appear to be >>>> able to conform to severance agreements. >>>> >>>>> I wonder if the >>>>> PIC32 will finally bite the dust with ARMs all around it crowding it >>>>> out >>>>> of the market? >>>> >>>> I've never held that ARM was much in the same space as PIC32. At least >>>> the tooling is pretty radically different. And I can't say I'm too sad >>>> if AVR just goes away. >>> >>> It would be a pity for the Arduino folks. >> >> It definitely would. Maybe I'm biased, but after working with AVRs for >> several years IMO at the 8 bit low end the AVR optimized-for-C >> architecture, plus a free high-quality compiler with full optimizations >> for every device, is the clear winner over the old-ass PICs. >> >> Only turbonerds hand write ASM for any real shit these days. > > There are plenty of Arduino-compatible boards with non-AVR chips. > Microchip makes one. Plus there is one with an Atmel Cortex M0, SAMD21, > the Arduino Zero <https://www.arduino.cc/en/Main/ArduinoBoardZero>. > > I think that Arduino will gradually move away from the AVR line, but it > is a very profitable product line and Atmel would not discontinue it. > After all, you can buy a Cortex M? from a bunch of companies, but an AVR > user might not want to convert to ARM. > > My former colleague is a "Wizard of Make" and senior manager at Atmel > and runs around the world promoting Arduino. I hope he's safe.
Yeah, it's getting to the point that for many projects, the cost difference between an AVR Mega 8 bit part and a low end ARM that you shrug and say "Eh, why not." I think 8 bit parts will likely always have some market share in ultra-low cost and/or low power applications. The Vuse electronic cigarette uses an ATiny in both the battery pack and cartomizer, cheap enough to be disposable.