32-bit AVR

Do you have a question? Post it now! No Registration Necessary

Translate This Thread From English to

Threaded View
I have heard that Atmel has a 32-bit AVR-like architecture (not ARM) in
development, near release.  Is this true?



Re: 32-bit AVR
On Sun, 16 Nov 2003 19:58:13 GMT, "Ian McBride"

Quoted text here. Click to load it

Atmel doesn't talk...

bob


Re: 32-bit AVR
On Sun, 16 Nov 2003 19:58:13 GMT, "Ian McBride"

Quoted text here. Click to load it

I should say that Atmel doesn't  talk very much anyway...
They are very tight lipped about everything.   New and old products.
bob


Re: 32-bit AVR
Quoted text here. Click to load it

 Anything is possible, but this makes little commercial sense :

- They have an ARM product line already, this would canabalise that..
- Others (Philips, Sharp, et al ) have ARM Microcontrollers in release
now
  ARM is quickly becomming the next 80C51
- Tools/Training for 'yet another 32 bit core' would take the std
gestation time

 It _would_ make sense for Atmel to release a single-die FLASH ARM, in
something like Mega128 pinouts/peripherals, with ~512K Bytes FLASH.
 For a Flash-centric company, Atmel have surprisingly small-memory
single-chip microcontroller FLASH variants.
( this excludes their niche dual-die BGA-only ARM offerings )
 

-jg

Re: 32-bit AVR
Quoted text here. Click to load it

ARM own's arm, so they pay a fee for each chip
Customers can 'jump' ship to another ARM processor.
AVR customers may be thinking of using an ARM (I know I am) and ATMEL
doesn't have anything that looks good to me.  But something like a beefed up
AVR and they may hang on to those customers.

Quoted text here. Click to load it

See above; A good reason for them to do something of their own.

Quoted text here. Click to load it

Sure - but then they make 8051's and AVR's. You could argue why even bother
with AVR's.

Quoted text here. Click to load it

It would, but if everyone thought that way we would never get anything new.

Quoted text here. Click to load it

Couldn't agree more.

Ralph



Re: 32-bit AVR
Quoted text here. Click to load it

 Correct, but those same customers then have to choose between
'Green Silicon/Single Source' OR Proven family/Multisourced.

 Look at Motorola's MCORE phase-out as an example of just how hard
even the biggest player finds hitting 'critical mass' on
a new 32 bit core.

-jg

Re: 32-bit AVR
Quoted text here. Click to load it



Atmel has had single die flash ARMs for a long time, but
these are ASICs and not standard products...
Single die flash + ARM standard products are coming ...

--
Best Regards
Ulf at atmel dot com
We've slightly trimmed the long signature. Click to see the full one.
Re: 32-bit AVR
Quoted text here. Click to load it

Please contact me if you ever make a single die flash version of the
55800A.  That was a very nice chip for my application other than the
lack of flash.  I think the inclusion of the RTC was a *great* idea.  I
don't think I have seen that done by anyone else.  

--

Rick "rickman" Collins

snipped-for-privacy@XYarius.com
We've slightly trimmed the long signature. Click to see the full one.
Re: 32-bit AVR

<snip>
Quoted text here. Click to load it

Info on the LPC2114/2124 is sparse, but this document
http://www.semiconductors.philips.com/acrobat/literature/9397/75012211.pdf
shows a block diagram  with :
 RTC, 16K RAM, 128KB FLASH, 2 UARTS, SPI,i2c, ADC, PWM, PLL.....
so it's reasonably similar to (55800A+FLASH+i2c+MoreSRAM-1uart-SomePins)...

-jg




Re: 32-bit AVR
Quoted text here. Click to load it

but -external bus!!!! That Philips chip is really cool, concerning
peripherals, but the lack of an external bus is a pitty!!! The Atmel chips
are fine, but most of them have little peripherals like SPI, I2C, ADC, etc.
except for the 55800, but this one lacks flash.

Cheers,
ER!K


--
---
LINS Technologies
http://www.lins.de



We've slightly trimmed the long signature. Click to see the full one.
Re: 32-bit AVR
Quoted text here. Click to load it
http://www.semiconductors.philips.com/acrobat/literature/9397/75012211.pdf
Quoted text here. Click to load it
etc.

 Do you mean external DATA bus, for (eg) Ethernet Chip, or external CODE
memory bus ?
External CODE turns a microcontroller into a microprocessor, and I guess
Philips reasoning
is if the FLASH is on chip, you do not NEED the pin/EMC costs of external
bus.

 I have not seen enough data to know how/when they support connection of
Octarts/Ethernet style
peripherals - but they indicate 48/64/128 pin devices so far, and I'm sure
if the demand is
there, they will make it.

-jg



Re: 32-bit AVR
Quoted text here. Click to load it

Yes, the Atmel parts fall short by not combining the best of all their
chips in one and the Philips chip is designed for similar apps as the
MSP430 where space is at a real premium with no need for an external
bus.  

The OKI ML67Q400x/500x chips are the ones that give you all of the above
along with an external bus (and has a simple serial port boot loader).  

I have thought that I might be able to use the Philips chip even without
the external bus by connecting to the FPGA via SPI or similar.  But the
Philips part requires a special programmer to boot load the code.  This
is just not what I want in my product.  

--

Rick "rickman" Collins

snipped-for-privacy@XYarius.com
We've slightly trimmed the long signature. Click to see the full one.
Re: 32-bit AVR
Quoted text here. Click to load it
Umm, the LPC210x just uses the serial port for loading.  

Robert Adsett

Re: 32-bit AVR
Quoted text here. Click to load it

I have seen several posts here that indicate that you need a
programmer.  Several people have indicated that they can't find info on
programming the flash.  Are the details in the data sheet?  

--

Rick "rickman" Collins

snipped-for-privacy@XYarius.com
We've slightly trimmed the long signature. Click to see the full one.
Re: 32-bit AVR
Quoted text here. Click to load it
The details are there (pg. 177 of the users manual. I seem to remember
they wern't all there in the previous rev of the users manual).  There is
also a windows utility available that downloads over the serial port
which is what I'm using now.  I haven't tried to write one myself but it
looks fairly straightforward, all the timing is taken care of on chip.  
Hardware wise you just need power, Rx, Tx and pull one additional pin
down on reset to trigger boot load mode.

Robert Adsett

Re: 32-bit AVR

Quoted text here. Click to load it
http://www.semiconductors.philips.com/acrobat/literature/9397/75012211.pdf
Quoted text here. Click to load it
etc.

look at Philips ARM roadmap, the LPC2200 144pin devices have external bus.
http://www.zlgmcu.com/philips/arm/ARM_Family.pdf


r.



Quoted text here. Click to load it



Re: Philips ARM's (was 32-bit AVR)
Quoted text here. Click to load it
 > > > <snip>
Quoted text here. Click to load it
chips

..and today Philips announce SJA2020 : ARM, 144 pins, 256KF, 6 CAN, 4 LIN,
60MHz
http://www.philipssemiconductors.com/news/content/file_1015.html

reads similar to the LPC2296 mention in the roadmap above.

_Six_ CAN ports !

-jg



Re: Philips ARM's (was 32-bit AVR)
Quoted text here. Click to load it

"First samples of the SJA2020 are now available in a 144-pin package and
mass production is scheduled to begin Q4 2004"

Like I said, call me *when*...


--

Rick "rickman" Collins

snipped-for-privacy@XYarius.com
We've slightly trimmed the long signature. Click to see the full one.
Re: Philips ARM's (was 32-bit AVR)
Quoted text here. Click to load it

Yes.... I did wonder if that was a typo - that's a L O N G from samples to
production, but the SJA2020 IS an automotive part, so maybe that's right ?

 I see they have also got data now up on the 64 pin models, with/without CAN
LPC2114 /2124  LPC2119/2129 - status claims "Samples available", and
prices are showing on the WEB, which are better signs.
-jg



Re: Philips ARM's (was 32-bit AVR)
Quoted text here. Click to load it

I think I saw a similar news release from Philips on another version of
this CPU that also has a long time from samples to production.  I'm not
sure this is really that long of a proto time frame.  Xilinx has been
sampling their new Spartan 3 FPGAs for over six months and production is
not likely until next year.  

--

Rick "rickman" Collins

snipped-for-privacy@XYarius.com
We've slightly trimmed the long signature. Click to see the full one.

Site Timeline