Looks like all Arduino.org pages now redirect to Arduino.cc.
No sign of Star Otto or Primo anymore....
That's a shame! I really longing for something like STAR OTTO. A lot stronger then the MEGA.
Come on Arduino: give us the STAR OTTO 
FWIW, 2018 is likely to bring Cortex-M7 microcontrollers at 400 & 600 MHz...
M7 is pretty incredible. It has 2 integer execution units, so sometimes it can get 2 instructions per clock. The FPU has both 32 bit float and 64 bit double. M7 also has branch prediction, so many branches that currently take 3 cycles in M3 & M4 become single cycle if predicted correctly.
M4 at "only" 180 MHz on Star Otto and Teensy 3.6 is going to look kinda sluggish by comparison.
Do you mean arduino will produce a new board based on Cortex M7 and if yes, what manufacturer (Atmel, ST...) ?
bobcousins:
Technically it looks great, but what sort of corporate morons release a non-Open Source board into an Open Source ecosystem?ST clearly have no clue about Open Source, and I'm not sure about Arduino Srl either. Let's hope it gets some cheap clones.
What do you mean by 'Non-Open Source'?
Every one of STM development boards have documentation including the PCB design files.
Many months ago, when Bob wrote those words, Arduino.org's page had changed from "coming soon" to showing the Star Otto as released. The web pages didn't have the schematic back then. I don't know if I would agree with the "corporate morons" sentiment, but at least at the time Bob wrote those words, their web page did say Star Otto was released (it wasn't) and no design files were present.
But Arduino.org did publish the schematic and other design info a short while later. I remember, because I spent a few hours looking over the schematic and the ST chip's pinouts, trying to figure out if they had exposed the pins necessary to make a Ethernet PHY shield and a ULPI USB shield.
Whether Star Otto ever fully met the open source hardware definition is a good question. Phil Torrone of Adafruit claimed it didn't. He posted several messages on mail lists to call out Musto and Arduino.org. While he probably was correct, it really looked to me like a technicality. Arduino.org did publish the full schematic in PDF, and other files I didn't try to open.
Except for a handful of "lite version" at the San Mateo Maker Faire (lacking the LCD, audio and SDRAM), it doesn't look like they ever actually sold any Star Otto boards.
I too was looking at an Otto like board and also was disappointed they didn't bring out the Ethernet to an RJ45 header on the board.
Atmel has the ATSAME70Q21 chip (look at the specs... it's awesome) . It will be interesting to see if folks pursue dev boards with it. it runs at 300Mhz. Doesn't have an on-chip MIPI DSI interface like 32-bit STM32F469 MCU that the Otto was to use. If Atmel could add that type of functionality to the great stuff that the ATSAME70Q21 already has.... that would be awesome.
Adafruit seems to be enthusiastic about the new Atmel SAMD5x chips. Somewhat similar to the samd21 boards out now, but with CM4F CPU, faster clocks, and more memory. It sounds pretty sweet, and would permit both uno and mega sized boards...
Otoh, arduino.cc and Atmel have had a relationship for a while now; one of the nice things about the Otto and primo was bringing in some new vendors...
So Paul- your little boot chip that loads code onto Freescale arm chips- would it work with other vendors chips as well?
westfw:
So Paul- your little boot chip that loads code onto Freescale arm chips- would it work with other vendors chips as well?
Nope, definitely won't work on any non-Freescale chip. Even on the Freescale chips, it only works with a pretty narrow subset having the same flash memory configuration. Like most bootloaders, it's pretty tightly tied to specific hardware.
But the bootloader currently used for Arduino Zero probably can be used with minimal modification. The SAMD50 looks like it's mostly register compatible with SAMD21. Many of the peripherals seem to have extra registers, mostly for use with DMA. The timers appear to be a little different, but mostly the same stuff arranged differently to accommodate newer modes while still supporting the old ones.
This backwards compatibility with SAMD20 is almost certainly why Adafruit is so interested. They're probably feeling pretty cramped with Python. But they know how much engineering time is needed to switch to another chip with all different peripherals.
SAMD50 has several new peripherals that look pretty amazing. It'll be interesting to see if anyone really makes effective use of the 4 bit SDIO controller with the Arduino SD library. We've had this on Teensy 3.5 & 3.6 for about a year. So far, we're mostly using it to just speed up the SD library as-is, still fetching 1 sector at a time, still all done in the code from user function calls. To really leverage the speed of modern cards, you really have to implement substantial caching in RAM and use multi-sector transfers, ideally done asynchronously with DMA while the rest of the user's program runs. That's on my bucket list... So far, there's been little interest from anyone else in the Arduino world. Maybe with SAMD20 now having this hardware and enough RAM, Arduino might finally want to implement a higher performance SD library?
your little boot chip that loads code onto Freescale arm chips- would it work with other vendors chips as well?
Nope, definitely won't work on any non-Freescale chip. Even on the Freescale chips, it only works with a pretty narrow subset having the same flash memory configuration. Like most bootloaders, it's pretty tightly tied to specific hardware.
I was under the impression that the teensy3 bootload chip used jtag (or maybe something proprietary to freescale) to load bootloader code into the Main ARM chips RAM, and then runs that to do the actual loading over usb, and was wondering if that first part (load from BLchip to RAM and execute) was relatively not-freescale specific. (I could be entirely wrong, of course, since I'm just analyzing the schematic...)
Each chip has plenty of vendor specific stuff, even if using similar protocols. For example, the Freescale chip has a module called MDM-AP, documented on pages 197-201:
Atmel (now Microchip) has something functionally similar (but entirely different) which they call DSU. It's documented on page 50-55:
This is only the beginning. Each chip is loaded with vastly different details of how to actually program its flash memory. Sure, they're both "flash", but just because they're called the same and use basically the same physical semiconductor process to store non-volatile data doesn't mean these different vendors did anything else similarly.
At least for access to RAM, they both do have the AHB-AP component. At least that part comes from ARM, so it's likely to be the same.
But even then, many thorny details can get in the way. For example, in the Freescale chip you can access the MDM-AP and even the AHB-AP while the chip is in reset mode. But Freescale's bus matrix obeys the reset and won't work at all. So if you want to get to anything that can't be reached directly inside the ARM core or through the core's private peripheral bus, you have to go through quite a lot of work to configure the debug unit to catch the reset vector and put the chip into halt mode before it can start, just to be able to release reset and gain access to everything else (like the RAM) which lies on the other side of that bus matrix.
I honestly haven't fiddled with Atmel's ARM chips much. But years ago I did play with the NXP LPC2000 ones (long before NXP acquired Freescale, both of whom are about to be acquired by Qualcomm). The LPC chips were riddled with other gotchas. One particularly difficult problem was they put the GPIO onto the AHB to avoid the high latency in whatever bus bridge they were using back then, but they added some (undocumented) bug where it wouldn't properly ignore bus cycles during debug mode. At the time I found an ugly workaround, but didn't continue farther with those chips since Freescale was soon to release the first ones with DSP extension and a good DMA controller.
Probably very few people outside of IAR, Segger and maybe the OpenOCD devs ever know of these internal chip issues....
This thread is on STAR Otto and started in 5/2016. It seems to be sure that this will never be available.
Other stuff on different controllers would better be discussed in new thread(s).
Given that there are so many different and affordable microcontrollers I don't understand this STAR Otto thread:
https://forum.arduino.cc/index.php?topic=495555.msg3413818#msg3413818
Sorry, I don't have Teensies, and was not able to get the STMs I have to work, so they are not part of that table. Other can provide comparison data for those if they want.
Hermann.
HermannSW:
This thread is on STAR Otto and started in 5/2016. It seems to be sure that this will never be available.
That might be the case... Just dropping in for a 3 monthly update. If indeed Otto was "about to be on sale", I think that is kicked into the long grass, by the recent hostile re-splitting of the amicable merger of the warring split Arduinos. If arduino.cc are not committing to Due, it seems unlikely they will commit to Otto. The Otto was a "pet project" of one of the devs at arduino.org, so that is probably the end of that. RIP, next to Tre.
Anyway, I agree with Paul, if the Otto did get launched now, Cortex M4 is looking a bit tame. I expect there to be some low end Cortex M7s around, if not already (market moves too fast for me to keep up!).
Looks like Federico Musto is relaunching as "What's Next".
The Warranties page pretty clearly says it's his CC Logisitics business, and a whois lookup shows he registered this domain name.
So far every appearance is 6 clones of existing Arduino boards, and 2 of his Dog Hunter variants of Yun. No sign of the Star Otto or Primo designs.
bobcousins:
Yes, that was the whole point of my post, historical precedent. Thank you for summarising.
HermannSW:
This thread is on STAR Otto and started in 5/2016. It seems to be sure that this will never be available.
as another historical precedence perhaps the never ending story (actually, finally ending though) of the TRE might be referenced, too... :-/
Now that otto is dead in the egg, perhaps someone will work to make a third party board with similar feature to the arduino IDE...
I'm saw, for example the ATSAMA5D3.
except for wifi, it seem to have even more capability... but at an bigger price also.