.... I know the oscillator has been discussed, but the Flash/Eeprom/OTP question hasn't even been asked of OnSemi.
Continuing this line of thought, another reality check before final design starts, would be to ask OnSemi 'Is 180nm still the right place on the price curve'
- during the long gestation of P2, the whole price curve has moved, and I noticed this Feb 2017 news release from Microchip :
https://www.microchip.com/pressreleasepage/sst-announces-ESF3-qual-on-110-nm-cmos
"For the first time SST’s ESF3 technology is available to fabless and Integrated Device Manufacturers (IDMs) on a highly cost-effective 8-inch 110 nm CMOS platform. This ESF3 platform is qualified for 300,000 erase and program cycles, making it ideally suited for smart card and other high-endurance IC designs."
Maybe in 2017+, that 110nm process is now considered the lowest cost node ?
We have so much custom analog in 180 nanometers, that we really can't switch to anything else right now. We are going to start at 180 nanometers.
Good point.... is that Analog 3V3, or 1.8V operating ?
Perhaps they can mix 180nm Analog and 110nm (or whatever) logic - 'custom' is ultimately just artwork lines, but it would need to be a common Core Vcc & oxides.
We have so much custom analog in 180 nanometers, that we really can't switch to anything else right now. We are going to start at 180 nanometers.
Good point.... is that Analog 3V3, or 1.8V operating ?
Perhaps they can mix 180nm Analog and 110nm (or whatever) logic - 'custom' is ultimately just artwork lines, but it would need to be a common Core Vcc & oxides.
I/O and analog are 3.3V (340nm), while the core logic is 1.8V (180nm)
At no particular time, though it would be good to get them some feedback, certainly before we spend a lot of money. I am going to email each of them a link to this thread in a day or two.
I think P2 will find itself being used in newer inventions, including robotics, machine control, signal processing (I think we can make a great little low-part count o-scope), and as a favorite general-purpose microcontroller by people who value the quick development from flexible I/O, forum support and symmetrical design. The kinds of "products" it goes into would be between 10 and 1K units in the near-term, likely not high volumes.
The role of education? Absolutely! Education has fueled the effort and will also be a recipient of the benefits of P2. We will segment our customers appropriately - the educational ones may not know the part number or name, but of course the commercial users will. As we simplify the P2 with the passing of time it's more important than ever that the first-time educator not get confused by our story [we make "multicore microcontrollers" and robots - a messy story for them to sort out when all of these users live in the same space, or web site].
C compiler and Spin are important for obvious reasons, along with a Blockly port at some point when customers have exceeded P1 Blockly-level coding (I don't see this happening too quick).
My personal interests would be to sell it in a module format. Ideally Parallax only offers one board/module format to get people started and the variety of hardware is grown by the community.
I see that many have gone to the trouble of a steep learning curve with ESP chips when they were already familiar and comfortable with other chips. Why? Is it because they are such a great and cheap solution for IoT apps? Yep, and that nothing else did come close to it (but I can get chips from the manufacturer for around $1 too).
So too, many will go through the steep learning curve for the P2 (from PIC/ARM/Arduino etc) if there is a killer app for it which could be motion control but maybe something simpler which it does really well. Then you don't need huge evaluation boards, just simple cheap little modules to get started or build it into your own project. Motion control also covers robotics and drones as well as 3D printers. My area of interest though is completely different to this although I acknowledge its "gotoitness" in this area provided the type of tools that makers are familiar with are available.
Just saw your message Ken - Yes, a processor for the PEOPLE (and not one that will become obsolete within a couple of years either, a wise investment of one's time I think).
It's true that people make huge Investments of time in learning an architecture. Everyone likes to fantasize about building a million units of something, but most common is maybe tens or hundreds. For each $1-$8 microcontroller purchased and used, they invested maybe 10x to 100x that in development time. The crazy part is that their time is rarely considered, if not ignored, while the unit cost of the microcontroller was made out to be 95% of the matter, when it was more like 5%, or less. I think most of us routinely suffer this delusion. We might as well pay $20 for a microcontroller that really makes it easy to get the job done.
I like the idea of Parallax making a module and one Activity like board for the module as a reference. Then the community can make boards that the module fits in for various needs/uses.
I think that makes it a lot friendlier for hobbyists and the like.
My personal interests would be to sell it in a module format. Ideally Parallax only offers one board/module format to get people started and the variety of hardware is grown by the community.
Certainly, having an initial module ready to ramp volume as soon as the chips arrive, will be vital.
I'm not sure "only offers one board" will quite fit ? - it may need 2 PCBs to cover the bases needed.
The RaspPi Zero form factor gives one guideline, and one decision there, will be to include WiFi/Bluetooth, or do two builds.
This also gives useful code frameworks, as talking to the RaspPi on all 0.1" interfaces, at their top speeds, all at the same time will get peoples attention.
Another segment would need a more complete PCB, and a LCD piggyback would extend the potential a lot there.
Education and Development openings for the second module I see would be more along the direction of the RedPitaya.
ie able to be used as a Logic Analyser, and Scope, even if that means adding some good-spec ADC's.
Although I have zero use for a P2 (since I never approach the limits of the P1 in my projects), I do have a couple recommendations for its launch:
1) A well thought out and heavily documented "covers everything" module with a library of tutorials on day 1.
2) A barebones module for very easy DIY integration for those able to provide any and all support circuitry on their own.
3) At least one heavily documented and current technological example project. My first thought was a 3+ axis CNC control board. You could cover a lot of bells and whistles with a project like that and it hits CNC, 3D Printing, etc. However, since my knowledge of the P2 architecture is limited, maybe others know of a prime example to show off its unique performance to prove to potential users that they must have one to learn.
Personally, I am salivating over the prospect of a single chip handling eight axes, each with two feedback devices, one for the actuator and one for the load.
Now if it will be possible to handle BiSS in firmware...
Now if it will be possible to handle BiSS in firmware...
An extension to SSI I presume. Looks to be reasonably straight forward synchronous serial, not unlike SPI. Should be compatible with existing SmartPin sync-serial mode.
Now if it will be possible to handle BiSS in firmware...
An extension to SSI I presume. Looks to be reasonably straight forward synchronous serial, not unlike SPI. Should be compatible with existing SmartPin sync-serial mode.
Cool....Kinematics should be a breeze for the P2 so now we have a single chip that can handle an entire 6-axis robot fitted with absolute encoders (?)
I think P2 will find itself being used in newer inventions, including robotics, machine control, signal processing (I think we can make a great little low-part count o-scope), and as a favorite general-purpose microcontroller by people who value the quick development from flexible I/O, forum support and symmetrical design. The kinds of "products" it goes into would be between 10 and 1K units in the near-term, likely not high volumes.
The role of education? Absolutely! Education has fueled the effort and will also be a recipient of the benefits of P2. We will segment our customers appropriately - the educational ones may not know the part number or name, but of course the commercial users will. As we simplify the P2 with the passing of time it's more important than ever that the first-time educator not get confused by our story [we make "multicore microcontrollers" and robots - a messy story for them to sort out when all of these users live in the same space, or web site].
C compiler and Spin are important for obvious reasons, along with a Blockly port at some point when customers have exceeded P1 Blockly-level coding (I don't see this happening too quick).
My personal interests would be to sell it in a module format. Ideally Parallax only offers one board/module format to get people started and the variety of hardware is grown by the community.
P2 is a processor for PEOPLE.
That's how I see it.
Ken Gracey
I have already used my P1 project boards as an o’scope, logic analyzer, multimeter, serial data sniffer, data logger, and various other instruments as needed. The P2 would be an even better choice for combining all those and more into a bench top electronics workstation. Add a Pi or similar micro and a monitor to that and you could have a very inexpensive software and hardware development system.
Maybe an initial board same size and similar layout as the project board with a second connector for the "B" port on the opposite side of the pcb to current "A" port?
Above, I mentioned "one board from Parallax". Maybe that single board is really a module holding the necessary P2 parts and the second board is a place to put it. Something like that.
I think my point is that if we're making all the boards that's a sign of under-adoption (?) by the community.
I'll wish that a little more effort have been done with Ethernet.
It would greatly benefit to all those applications (robotic, motion control, test and measurement, signal processing) designed to feed data to a higher-end processor (or PC).
The benefit of Ethernet over USB is that Ethernet only requires one third (or 1/5) the R&D effort on software development (not to mention the benefits of not needing a new driver with every new Windows version).
Think that any USB application will need drivers for at least three platforms (Windows, Mac, Linux, all of them x86). Nowadays people is also asking for USB Raspberry Pi and Android drivers (both ARM).
With ethernet you don't have to worry about any of that (it doesn't matter if it is Win/Mac/Linux, x86 or ARM).
Just a simple hardware CRC32 instruction (or peripheral) combined with the amazing power of 16 cogs could open the door to something UNIQUE.
To the question of who will buy the P2, my opinion is that it will be mostly people that buy the P1.
Starting off with a 180nm process is a disadvantage, as most very low end ARMs use 150nm that produce the 50 cent chips, these are mostly pad limited designs that would not benefit from shrinking. The midrange ARMs are using 90 and moving to 65 nm. These are the parts in the sub $10 range in small quantities (the P2 price point). The high end ARMs used in Android devices are built with 22 nm processes. FPGAs use some of the smallest processes, and while FPGA small quantity prices are high, that is by design as the FPGA vendors are more interested in large volume customers.
In the last 11 years, the processor portion of an IC has become more of a commodity and the market has shaken out most other CPU designs. Custom hardware has been designed and is necessary as it is not possible to bit-bang USB 3.0, 100 Mb Ethernet, 4K HDMI. And yes these are complex in both hardware and software, the chip vendors are supplying drivers and sample code.
You can teach programming on most any platform, but I would think that using a platform (hardware or software) that a student is unlikely to see elsewhere is probably a more difficult task. I am not suggesting Parallax build an ARM processor, I think it makes more sense to buy one or to buy a Risc-V. Let some other company make that investment, spend your time developing course material.
Speaking of familiar "hooks" that would catch the eye of millions hackers around the world I think one of the first things to do with a P2 is create a Raspberry Pi Hat board.
Comments
- during the long gestation of P2, the whole price curve has moved, and I noticed this Feb 2017 news release from Microchip :
https://www.microchip.com/pressreleasepage/sst-announces-ESF3-qual-on-110-nm-cmos
"For the first time SST’s ESF3 technology is available to fabless and Integrated Device Manufacturers (IDMs) on a highly cost-effective 8-inch 110 nm CMOS platform. This ESF3 platform is qualified for 300,000 erase and program cycles, making it ideally suited for smart card and other high-endurance IC designs."
Maybe in 2017+, that 110nm process is now considered the lowest cost node ?
Addit: I also find this
http://www.onsemi.com/PowerSolutions/content.do?id=18579
http://www.onsemi.com/site/pdf/ON-Semiconductor-0.18um-BCD-IP-v5.pdf
Of course they avoid comparing with ROM-areas, but it does seem to avoid needing more layers, so it could be a 'late-ROM' candidate ?
Perhaps they can mix 180nm Analog and 110nm (or whatever) logic - 'custom' is ultimately just artwork lines, but it would need to be a common Core Vcc & oxides.
I/O and analog are 3.3V (340nm), while the core logic is 1.8V (180nm)
At no particular time, though it would be good to get them some feedback, certainly before we spend a lot of money. I am going to email each of them a link to this thread in a day or two.
Thanks!
The role of education? Absolutely! Education has fueled the effort and will also be a recipient of the benefits of P2. We will segment our customers appropriately - the educational ones may not know the part number or name, but of course the commercial users will. As we simplify the P2 with the passing of time it's more important than ever that the first-time educator not get confused by our story [we make "multicore microcontrollers" and robots - a messy story for them to sort out when all of these users live in the same space, or web site].
C compiler and Spin are important for obvious reasons, along with a Blockly port at some point when customers have exceeded P1 Blockly-level coding (I don't see this happening too quick).
My personal interests would be to sell it in a module format. Ideally Parallax only offers one board/module format to get people started and the variety of hardware is grown by the community.
P2 is a processor for PEOPLE.
That's how I see it.
Ken Gracey
So too, many will go through the steep learning curve for the P2 (from PIC/ARM/Arduino etc) if there is a killer app for it which could be motion control but maybe something simpler which it does really well. Then you don't need huge evaluation boards, just simple cheap little modules to get started or build it into your own project. Motion control also covers robotics and drones as well as 3D printers. My area of interest though is completely different to this although I acknowledge its "gotoitness" in this area provided the type of tools that makers are familiar with are available.
Just saw your message Ken - Yes, a processor for the PEOPLE (and not one that will become obsolete within a couple of years either, a wise investment of one's time I think).
I like the idea of Parallax making a module and one Activity like board for the module as a reference. Then the community can make boards that the module fits in for various needs/uses.
I think that makes it a lot friendlier for hobbyists and the like.
I'm not sure "only offers one board" will quite fit ? - it may need 2 PCBs to cover the bases needed.
This link from another thread :
RISV-V as an Arduino format board...
http://hackaday.com/2017/05/20/arduino-cinque-the-risc-v-esp32-wifi-bluetooth-arduino/
The RaspPi Zero form factor gives one guideline, and one decision there, will be to include WiFi/Bluetooth, or do two builds.
This also gives useful code frameworks, as talking to the RaspPi on all 0.1" interfaces, at their top speeds, all at the same time will get peoples attention.
Another segment would need a more complete PCB, and a LCD piggyback would extend the potential a lot there.
Education and Development openings for the second module I see would be more along the direction of the RedPitaya.
ie able to be used as a Logic Analyser, and Scope, even if that means adding some good-spec ADC's.
1) A well thought out and heavily documented "covers everything" module with a library of tutorials on day 1.
2) A barebones module for very easy DIY integration for those able to provide any and all support circuitry on their own.
3) At least one heavily documented and current technological example project. My first thought was a 3+ axis CNC control board. You could cover a lot of bells and whistles with a project like that and it hits CNC, 3D Printing, etc. However, since my knowledge of the P2 architecture is limited, maybe others know of a prime example to show off its unique performance to prove to potential users that they must have one to learn.
Now if it will be possible to handle BiSS in firmware...
http://www.biss-interface.com/
Cool....Kinematics should be a breeze for the P2 so now we have a single chip that can handle an entire 6-axis robot fitted with absolute encoders (?)
I have already used my P1 project boards as an o’scope, logic analyzer, multimeter, serial data sniffer, data logger, and various other instruments as needed. The P2 would be an even better choice for combining all those and more into a bench top electronics workstation. Add a Pi or similar micro and a monitor to that and you could have a very inexpensive software and hardware development system.
I think my point is that if we're making all the boards that's a sign of under-adoption (?) by the community.
Ken Gracey
It would greatly benefit to all those applications (robotic, motion control, test and measurement, signal processing) designed to feed data to a higher-end processor (or PC).
The benefit of Ethernet over USB is that Ethernet only requires one third (or 1/5) the R&D effort on software development (not to mention the benefits of not needing a new driver with every new Windows version).
Think that any USB application will need drivers for at least three platforms (Windows, Mac, Linux, all of them x86). Nowadays people is also asking for USB Raspberry Pi and Android drivers (both ARM).
With ethernet you don't have to worry about any of that (it doesn't matter if it is Win/Mac/Linux, x86 or ARM).
Just a simple hardware CRC32 instruction (or peripheral) combined with the amazing power of 16 cogs could open the door to something UNIQUE.
Starting off with a 180nm process is a disadvantage, as most very low end ARMs use 150nm that produce the 50 cent chips, these are mostly pad limited designs that would not benefit from shrinking. The midrange ARMs are using 90 and moving to 65 nm. These are the parts in the sub $10 range in small quantities (the P2 price point). The high end ARMs used in Android devices are built with 22 nm processes. FPGAs use some of the smallest processes, and while FPGA small quantity prices are high, that is by design as the FPGA vendors are more interested in large volume customers.
In the last 11 years, the processor portion of an IC has become more of a commodity and the market has shaken out most other CPU designs. Custom hardware has been designed and is necessary as it is not possible to bit-bang USB 3.0, 100 Mb Ethernet, 4K HDMI. And yes these are complex in both hardware and software, the chip vendors are supplying drivers and sample code.
You can teach programming on most any platform, but I would think that using a platform (hardware or software) that a student is unlikely to see elsewhere is probably a more difficult task. I am not suggesting Parallax build an ARM processor, I think it makes more sense to buy one or to buy a Risc-V. Let some other company make that investment, spend your time developing course material.
The module idea is great.
Also seen some other MCU boards hijacking the Arduino IDE for their own MCUs.
Wonder if it'd make any sense (and be possible) for Parallax to make a P2 board that can use the Arduino IDE...
With, P2 you can now have the speed and memory and I/O to interface with low cost camera modules and full color LCD displays.
Plus, the horsepower to process sensor inputs, like 9DOF mems chips.
Also, the direct USB and many analog I/O could allow new applications.
It would be nice though if there were an easy way to add wireless to P2 for low cost...
That gets you WIFI and Bluetooth.
Network connectivity for less than the price of an Ethernet jack.
Not to mention a device that is increasingly familiar to millions of hackers that they can feel comfortable with.
It just seems that all the modules I can find are wireless+MCU... Somehow, seems would lower cost to get wireless w/o MCU...
Was just looking on Digikey... Like the look of CC2500. Needs just a couple external items and has SPI interface...