While I really want one too [noparse][[/noparse]or maybe two or three], my annual income of $300,000NT is merely equal to about One Million Japanese Yen.
[noparse][[/noparse]You can let the marketing department do the math to USD].
Still I spend hundreds of USD each year on pursuit of this pleasure and technical odessey.
It looks to me like the bread n' butter of the Propeller is going after the 'Gamer market'.
If it goes inside an X-Box or a Playstation or ?, Parallax will be living on the volume, not the hobbyist and hackers. Nonetheless, new technologies depend on popularity to rapidly deploy. So I really don't feel that we will be left out in the pricing.
Alternatively, there is a lot of discussion of a 3rd world computer these days and it certainly looks an interesting alternative to Intel. Is Parallax seeking greatness in the computer annals of fame? Could be. Why else would it go into making its own chip?
Alternatively, the Propeller could possilbly manage all the I/O in a traditional Intel platform. Do you dare to imagine how many chips might be sold in that fashion?
Then there is the question of fitting one of these into a cellular phone. What would it do there? Video conference calls while following your Parallax IPO? China and India are driving the cellular phone market well over the billion user market.
Eric, you really know how to get a jump on things. I doubt if you will get any sellers though.
In sum, the more volume that this has --- the better the price to us little guys. I suspect Parallax has a very good employee profit sharing plan and everyone there is doing the utmost to realize a new dream.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)
······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
As a college student, I don't really have much money. I did have some set away for one of those Penguins, but this really caught my attention. I've started the initial design phase of a simple 2D RPG and looking at what was done with the XGS, this chip could do so much and very easily. So I'm definitely grabbing one of these when they come out. Not sure if I'll wait for the "Hydra" game development kit or whether I'll build my own, but either way, I think my little RPG will be done on one of these beauties. Now if only I knew how much I might be dropping on this. I think I can wait a little bit to find out because in my opinion, it's worth the wait.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Truth can be dangerous... Trust can be deadly.
Still I spend hundreds of USD each year on pursuit of this pleasure and technical odessey.
It looks to me like the bread n' butter of the Propeller is going after the 'Gamer market'.
No, it's not particularly suited for gamers. It just doesn't have the memory capabilities needed for that.
(To me it looks like the Hydra is right at the boundaries of what it can do)
It should be a cost-effective replacement for the PS3 dev-kit, though, to learn all about inter-process communication and how that influences game-design.
Kramer said...
If it goes inside an X-Box or a Playstation or ?, Parallax will be living on the volume, not the hobbyist and hackers. Nonetheless, new technologies depend on popularity to rapidly deploy. So I really don't feel that we will be left out in the pricing.
There's no way that M$ or Sony is going to adopt a chip from a relatively small-time company as Parallax.
Apple might, if it looks cool enough, but that's another case...
(On the other hand, they don't have much to brag about in the gaming industry. Just google for 'Atmark Pippin'... )
Kramer said...
Alternatively, there is a lot of discussion of a 3rd world computer these days and it certainly looks an interesting alternative to Intel. Is Parallax seeking greatness in the computer annals of fame?
No, this thing doesn't seem to have the processing-power of more than an old Pentium chip, and that's if we overlook the lack of memory-space. And Parallax probably won't be able to deliver the quantity of chips needed in any 3rd world project. (given that any 'PC' designed with this chip would need an entirely new OS designed specifically for it, whereas any old 486 or Pentium clone can reuse well.. OpenDOS, Linux, Minix, OS/2 and so on... )
Kramer said...
Alternatively, the Propeller could possilbly manage all the I/O in a traditional Intel platform.
Sorry, but no, I don't think so. How is an 80MHz chip going to keep up with the 2 - 4GHz processors of today?
Kramer said...
Then there is the question of fitting one of these into a cellular phone. What would it do there?
Nothing, probably.
Even my PDA has more processing-power than this chip.
(It's a Psion netBook with a 190MHz ARM chip)
And most smartphones now have similar or better chips...
This chip is definitely levered towards hobbyists and schools.
When is the WASC book coming out?
(What's A SuperComputer... )
Going over hints and discussions in the forum, and the mailing-list before that, and adding a bottle of Hindsight, the Propeller is the obvious next step...
Users have always wanted more speed, but there's a limit to how fast you can go without incurring speed-related problems. Adding more cores sidesteps that problem.
(New) Users always nag about interrupts, which Basic can't do. I haven't seen any mention of interrupts in the Propeller, either, but with the possibility of having cores dedicated to spacific tasks, well... It's another neat sidestep...
'I need more memory! My program can't be split up into 2KB segments!' no longer a problem, according to a post in this thread.
Also, if you look at what is happening in the computing-world today, you'll see more and more multi-processing/multi-core systems showing up, and guess what, there are no cheap trainers for engineers and programmers to 'mess about on' and learn the basics...
Anyway...
I'd like to see the makers behind the BasicAtom and other clones try to copy this one...
And if I can afford one, it'll go into my car...
Scrolling text on a long LED bar in the rear window, heads-up display, central-locking system, fuel-economy indicator(by reading off RPM, gear-selection, incline and so on)...
I was thinking of doing this with my BS2P, and using a PalmPilot as keyboard and display, but the enhanced IO-capabilities of the Propeller will make it a better choice.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Don't visit my new website...
Post Edited (Gadgetman) : 2/18/2006 10:51:19 AM GMT
Ken should have a contest to see who can guess the closest to the "actual selling price" of the P8X32A-DIP. Who ever gets the closest without going over gets a free Propeller developement kit.
To make sure there isn't 100 winners only the first correct guess wins.
Bean.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95
Of Course there are the commercial aspects of this chip but I really think Parallax is sticking to hobbyists and education.
I seems as if it may be a great chip for robotics.
also the hydra... Its main purpose seems to be to educate people on using multi-cell processors.
IMO, the chip will be around $10 give or take, and the basic dev kit will be around $50-$60.
I don't think Parallax is here to sell out.· The chip is not some Pentium whatever.· And all of the R&D is theres and not outsourced, which would probably help keep costs resonable.
I'm sure also that it was alot of fun for them the whole process (trips to china , ect)
First impressions: It is similar to your standard microcontroller (PIC or Atmel) but darn fast. It has eight parallel processors running at 80Mhz (20MIPS) each. That's 160 MIPS. Memory and IO access use a unique "cog" architecture under the hood, but it is mostly obfuscated unless you do direct assembly programming. For those of you that are experienced programmers, the main hurdle is going to be the new language they use. It looks a convergence of pascal, basic, and python. I sure hope someone comes up with a c compiler.
Fortunately, they have created a bunch of "objects" for doing all your standard stuff. They even have objects that can read a standard PS2 mouse and keyboard. One of their coolest objects allows you, without any extra hardware and a minimal amount of software, to create NTSC (or PAL) or VGA output directly. I'm not talking about boring B/W blocks and text, I'm talking about 64 color animated graphics and text. The best part is that fully animated video only takes two of the eight "cogs". So, you have 6 other cogs to do other stuff. You could do music, motion control, IO, anything. You could even create four separate video streams using all eight cogs. There is even an extra option to modulate the direct video signal and create a short range broadcast signal (picked up on channels 2-16). This works from about 20' away without any extra antenna (just radiating off of the chip, sorry FCC).
The final pricing hasn't been set yet, but raw chips (40 dip or 44 LQFP) should be around $20 each. Their whole dev. board kit should be around $200.
Posted by Phillip Torrone | 11:48 AM
Post Edited (Steve Joblin) : 2/18/2006 3:14:01 PM GMT
Why a C compiler?
C doesn't have any syntax that helps in developing multi-processor programs, so that means a less than ideal solution.
(I expect the Parallax IDE is tooled specifically for that)
As for Dev-kits, Al(bless him) already posted that what you NEED is a USB2SER adapter at $29, a EEPROM and a 3.3V voltage regulator.
What the 'pro' dev-kit seems to add is PS/2 connectors, an external Xtal, experimenter area and video output.
Knowing Parallax, they will most probably post the schematics for the board anyway, so that those who can't afford it can set up those circuits they need on an experimenter board or breadboard.
If the chip and board together is no more than $150, I'll buy the board.
(I assume that those boards will use the LQFP version as seen on the Demo-board pictures.)
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Don't visit my new website...
Why a C compiler?
C doesn't have any syntax that helps in developing multi-processor programs, so that means a less than ideal solution.
(I expect the Parallax IDE is tooled specifically for that)
I've worked with multi-processing·systems and one of the things I've found (that actually surprised me) is that·you really don't need any special syntax for it.·It works just like either string processing·function calls or multithreading in any modern (ie, pre-emptive multitasking) OS -·depending on whether you're using·MIMD-style or SIMD-style, respectively.·A thin little API·is really all you need.
I'm sure there's some clever multi-processing syntax out there that someone's come up with, but in actual practice,·C's lack of built-in multi-processing concepts is really not a hinderance at all.
Plus, if you look at the Propeller as an "introduction to real-industry multiprocessing" as many of the posters in this thread have stated, then take a look at the mainstream game developers working on XBox360 and PS3. If you know a lot about the mainstream games industry, it's a VERY safe bet that they are not abandoning C/C++ in favor of special multi-processing languages. So using C on a Propeller·would be the closest thing to what they do.
EDIT: BTW, Hello to everyone· . I'm relatively new to Parallax and hardware (but learning fast!), and I'm really thrilled with what I've seen of Parallax so far. Seems like a great company with great products and an excellent community .
Jon Williams said...
If another cog defines the pin as an input (default state) and reads the pin, it will see the current output value from that other cog.
This sounds like the formula for imminent disaster. I have heard and read numerous software-related horror stories about Read-Write-Modify situations where the reading of a pin's register after writing to it came back with the stored result rather than the external signal.
Is there a safeguard built into the compiler(s) for this situation? Is there ANY way to monitor for this beyond creating a running list of which COG did what to each pin as the programmer writes his code?
Or am I seeing the sky falling?
Looks like the world will beat a path to your door with this design!
kenjj
I am still wondering about the Propeller's High level language - seems that it will be some kind of Object Oriented language to allow quick integration of numerous devices in a multi-tasking environment. Basic is just too much simplicity. It could only be a sub-set of a greater whole.
Multi-Processing.· Multi-tasking can be done too, but multi-processing is much better.
Gadgetman said...
(New) Users always nag about interrupts, which Basic can't do. I haven't seen any mention of interrupts in the Propeller, either, but with the possibility of having cores dedicated to spacific tasks, well... It's another neat sidestep...
Jon mentioned it earlier, but I'll say it again, the Propeller doesn't need interrupts.· By design, it has something that is much, much better... not really a sidestep, more like a better solution.· Just dedicate a cog to an task where you'd use interrupts in other processors, and set it to wait for a particular pin change... super fast response, absolutely no interrupts of other cogs' tasks.
Why a C compiler?
C doesn't have any syntax that helps in developing multi-processor programs, so that means a less than ideal solution.
(I expect the Parallax IDE is tooled specifically for that)
I've worked with multi-processing·systems and one of the things I've found (that actually surprised me) is that·you really don't need any special syntax for it.·It works just like either string processing·function calls or multithreading in any modern (ie, pre-emptive multitasking) OS -·depending on whether you're using·MIMD-style or SIMD-style, respectively.·A thin little API·is really all you need.
I'm sure there's some clever multi-processing syntax out there that someone's come up with, but in actual practice,·C's lack of built-in multi-processing concepts is really not a hinderance at all.
Plus, if you look at the Propeller as an "introduction to real-industry multiprocessing" as many of the posters in this thread have stated, then take a look at the mainstream game developers working on XBox360 and PS3. If you know a lot about the mainstream games industry, it's a VERY safe bet that they are not abandoning C/C++ in favor of special multi-processing languages. So using C on a Propeller·would be the closest thing to what they do.
EDIT: BTW, Hello to everyone· . I'm relatively new to Parallax and hardware (but learning fast!), and I'm really thrilled with what I've seen of Parallax so far. Seems like a great company with great products and an excellent community .
Welcome Nick!
Everybody,
many of the things we've done with the Propeller design take a completely different approach to things like multi-processing, interrupts, objects, etc.· Of course, some will love it, some will not.· Nothing's perfect, and perfection is subjective, but I think many people who·actually get their·hands on one, and take the time to learn it, will feel refreshed and unburdened.
As far as language, I know there are wars started on topics like this; very silly.· Let's not go there, but I have to say that if you are one to want to squeeze all the power out of something that you can, you are doing yourself a great disservice by not learning the native language of the device itself, but instead are relegating yourself to get only as far as 1) the·cleverness of the cross-compiler writer, and 2)·the limitations of a general-purpose language applied to an embedded system device, will allow you to go.· Yes, learning a new language is hard to swallow for some people, but
those who don't bother may be missing out on some very exciting stuff.· The Propeller's language is not difficult to learn or use for most people.
I would urge anyone who would stay away from the Propeller simply because there are no tools for the language of their choice, to just take a chance to learn something new; you may find you like it.· If you're successful, you've gained a lot, if you're not you've lost only a little.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ --Jeff Martin
· Sr. Software Engineer ·Parallax, Inc.
Post Edited (Jeff Martin) : 2/18/2006 11:58:28 PM GMT
This will allow me to take some of my projects that are multi processor (8 bit's PIC's & ARV's) and possibly implement the same functionality into one chip.
Looking forward to SPIN!
Mike
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"OEM NMEA GPS Module" Now available on ebay for only $17.49
Jon Williams said...
If another cog defines the pin as an input (default state) and reads the pin, it will see the current output value from that other cog.
This sounds like the formula for imminent disaster. I have heard and read numerous software-related horror stories about Read-Write-Modify situations where the reading of a pin's register after writing to it came back with the stored result rather than the external signal.
Is there a safeguard built into the compiler(s) for this situation? Is there ANY way to monitor for this beyond creating a running list of which COG did what to each pin as the programmer writes his code?
Or am I seeing the sky falling?
Looks like the world will beat a path to your door with this design!
kenjj
I understand your concern, but we've worked hard to eliminate that problem.· The inputs read by any cog are directly from the sense-amps, not from the output latch.· A cog reading the state of pins immediately after they are written by another cog reports the correct state of the pins as they exist at the physical pin.· No read-modify-write issue there.· It's amazingly simple and straight-forward.
This will allow me to take some of my projects that are multi processor (8 bit's PIC's & ARV's) and possibly implement the same functionality into one chip.
Looking forward to SPIN!
Mike
Thank Mike.· Implementing your projects one one chip is very likely.· We expect to see a lot of that in the first few months after official release.· Glad you're looking forward to Spin;· I'm looking forward to teaching it.· We're working hard to get it all ready to go.· You're patience,·and that of others,·is very appreciated.
Can you say anything about RF? Can the propeller send and receive radio signals? If it can broadcast video, it can probably broadcast radio. What about receiving?
Can you say anything about RF? Can the propeller send and receive radio signals? If it can broadcast video, it can probably broadcast radio. What about receiving?
Thanks.
Yes, it can be made to broadcast AM and FM.· Receiving, in any case that I know of, is always more difficult than transmitting.· I belive it can be done with fast ADCs but can not provide more details since I have no direct experiance trying such a thing.
I feel lucky to have been one of the participants in the meeting at Parallax yesterday, and I can attest, this is different from any chip I have seen before. The hardware architecture creative, highlly symmetrical, artful. An amazing IDE and an amazing instruction set go with it, both at the asm level and at the Spin (higher level language) level. I'm not sure what I am allowed to say. We were not sworn to secrecy, but on the other hand, I'm can't say much because still getting my head around the details. It does require a different way of thinking. I'm barely past the the blink an LED stage.
The way I explained Propeller to my technophobic friends was this: Think of it as sliced bread. (The next best thing!) You have different spreads (rasberry, peanut butter, honey,...) and those you put on a slice of bread. The bread is the computer, and the spreads are the programs. Programs have different flavors, consider them tastes or tasks (eating). If you have one slice of bread and 8 spreads on that slice and 8 people wanting to taste, okay, well, it gets complicated. Only one person at time (without being awkward), and it becomes a real problem of how to arrange the spreads on the surface. Okay, well, compare that with 8 slices of bread, and each has its own spread, and 8 people can pass them around and taste in turn. That takes some coordination, but, remember, these are the technophobic friends and always ready to party!
While much of the discussion here has been about the high speed capabilities, I look to low power, long battery life stuff. That is fully taken into consideration in this chip with an amazing clocking scheme that can shift gears on the fly. Some commands can stop an individual cog's operation, while waiting for an event on a pin or from the system tick, thus reducing the current draw, while ready to continue execution in one tick. And propeller as a whole can run on its internal 20khz clock, so that it draw microamps of current but scan still monitors pins and ticks and can wake up quickly and do something and accelerate to high gear when necessary.
Can any cog start and stop any other cog? Or are the cogs 'prioritized' so that only a higher level cog can 'control' a lower level cog? How much interaction is there cog-to-cog?
How many max MIPS can an individual cog run? Is it 20 MIPS as 'unofficially' stated in a previous post?
How many clock ticks per cycle?
How many clock cycles does each instruction take?
Are there any future plans for a scaled back (and lower priced) propeller chip, say with only 4 or 2 cogs? I'd guess that the bulk of the use of the propeller would be with less than 8 cogs....
I think I read that each cog is identical so that would imply each cog can start/stop another.
If the clock is 80mhz and the mips is 20 then each instruction must take 4 clocks each.
I can see a day when the next generation of the chip comes out and it pipelines for one instruction per clock for 80 mips per cog.
So if each cog gets 2k bytes of ram/instruction space and the instructions are 32 bits then that gives you 512 instructions per cog. That is like having 8 sx28 chips in one package in a way.
Looking at the diagram, the hub can access 32k of ram and 32k of rom. The cogs must have a way of commanding the hub to transfer data to/from the global memory. I guess with a slight performance hit as compared to access the local 2k of ram allotted to each cog.
I say this because since the cog instruction space seems to be shared with the data space, this says that self modifying code is possible.
Thinking about it, you could partition the code in a way that it could overlay a part of the cog ram so you could swap in different routines from global ram, effectively expanding the local cog ram up to the limit of 32k with paging techniques.
I think that the sky is the limit here. I get really frustrated with chips that lock you into using the chip only one way. This chip really seems to let you be creative like no other before.
We all know of the programmable/reconfigurable hardware chips but rarely see them in use. Why? because sometimes it is hard to break out of the literal mind set. We want a chip to do one thing and one thing only.
You need DSP functions, You use a DSP, you need lots of IO you use an IO loaded chip.
Give us a chip that can be reconfigured on the fly and we fall a part trying to get our brains wrapped around the possibilities.
This chip seems to be what the industry needs. A cpu in the standard sense but with on the fly configuration capabilities. A bridge if you will. Use it a a· mega controller or morph it into something new.
I bet if you ask, 'can it do this?' or 'can it do that?', that the answer would boil down to, it can if you program it that way.
As far as size goes, 9mm/9mm would almost fit in my toothbrush so I guess it would fit in a cell phone. Besides, I'm sure a custom version could be created with less IO pins if the customer needed.
8 processors, not 8 threads but 8 processors. This chip is going to rock the controller world and set a new standard for future controllers.
Pricing, I'll toss my two cents into this.......
From the picture, I doubt the monitor will be included as part of the basic kit. Why? Because we all have TV/monitors around. Heck, that monitor is so small and cute it could be a product on it's own and they would make a mint selling it as an add on.
The board has the chip, two regulators, a crystal, a few caps, a breadboard, some extras like keyboard/mouse port (should be two ports, one for each I would think unless a splitter is included) and what seems to be a serial/usb programming key.
so...
Key = 30 bucks ( I think that has all ready been established).
Hardware (less chip) 20 bucks. Based on the sx48/52 with less parts at 10. Doubled it for fun.
Chip = 80 bucks, based on BS pricing. True, this beats the pants of BS chips but they must have priced BS chips at what they felt the market would tolerate. This would be introduced at the top end BS price and the BS chips would then get a price cut.
So my guess is going to be $130.00 to $150.00 range for a dev board (with programming key).
Post Edited (ToneDefJam) : 2/19/2006 2:58:01 AM GMT
I'll let Parallax staff answer this directly, but I think you'll be happily shocked at the chip cost. I also attended their 2 days training session with Tracy Allen - It was nice meeting you Tracy!.· PS:· It's an amazing chip.
Comments
Does anyone want to sell their kit from the class today?
If so, PM me
[noparse][[/noparse]You can let the marketing department do the math to USD].
Still I spend hundreds of USD each year on pursuit of this pleasure and technical odessey.
It looks to me like the bread n' butter of the Propeller is going after the 'Gamer market'.
If it goes inside an X-Box or a Playstation or ?, Parallax will be living on the volume, not the hobbyist and hackers. Nonetheless, new technologies depend on popularity to rapidly deploy. So I really don't feel that we will be left out in the pricing.
Alternatively, there is a lot of discussion of a 3rd world computer these days and it certainly looks an interesting alternative to Intel. Is Parallax seeking greatness in the computer annals of fame? Could be. Why else would it go into making its own chip?
Alternatively, the Propeller could possilbly manage all the I/O in a traditional Intel platform. Do you dare to imagine how many chips might be sold in that fashion?
Then there is the question of fitting one of these into a cellular phone. What would it do there? Video conference calls while following your Parallax IPO? China and India are driving the cellular phone market well over the billion user market.
Eric, you really know how to get a jump on things. I doubt if you will get any sellers though.
In sum, the more volume that this has --- the better the price to us little guys. I suspect Parallax has a very good employee profit sharing plan and everyone there is doing the utmost to realize a new dream.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)
······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Truth can be dangerous... Trust can be deadly.
No, it's not particularly suited for gamers. It just doesn't have the memory capabilities needed for that.
(To me it looks like the Hydra is right at the boundaries of what it can do)
It should be a cost-effective replacement for the PS3 dev-kit, though, to learn all about inter-process communication and how that influences game-design.
There's no way that M$ or Sony is going to adopt a chip from a relatively small-time company as Parallax.
Apple might, if it looks cool enough, but that's another case...
(On the other hand, they don't have much to brag about in the gaming industry. Just google for 'Atmark Pippin'... )
No, this thing doesn't seem to have the processing-power of more than an old Pentium chip, and that's if we overlook the lack of memory-space. And Parallax probably won't be able to deliver the quantity of chips needed in any 3rd world project. (given that any 'PC' designed with this chip would need an entirely new OS designed specifically for it, whereas any old 486 or Pentium clone can reuse well.. OpenDOS, Linux, Minix, OS/2 and so on... )
Sorry, but no, I don't think so. How is an 80MHz chip going to keep up with the 2 - 4GHz processors of today?
Nothing, probably.
Even my PDA has more processing-power than this chip.
(It's a Psion netBook with a 190MHz ARM chip)
And most smartphones now have similar or better chips...
This chip is definitely levered towards hobbyists and schools.
When is the WASC book coming out?
(What's A SuperComputer... )
Going over hints and discussions in the forum, and the mailing-list before that, and adding a bottle of Hindsight, the Propeller is the obvious next step...
Users have always wanted more speed, but there's a limit to how fast you can go without incurring speed-related problems. Adding more cores sidesteps that problem.
(New) Users always nag about interrupts, which Basic can't do. I haven't seen any mention of interrupts in the Propeller, either, but with the possibility of having cores dedicated to spacific tasks, well... It's another neat sidestep...
'I need more memory! My program can't be split up into 2KB segments!' no longer a problem, according to a post in this thread.
Also, if you look at what is happening in the computing-world today, you'll see more and more multi-processing/multi-core systems showing up, and guess what, there are no cheap trainers for engineers and programmers to 'mess about on' and learn the basics...
Anyway...
I'd like to see the makers behind the BasicAtom and other clones try to copy this one...
And if I can afford one, it'll go into my car...
Scrolling text on a long LED bar in the rear window, heads-up display, central-locking system, fuel-economy indicator(by reading off RPM, gear-selection, incline and so on)...
I was thinking of doing this with my BS2P, and using a PalmPilot as keyboard and display, but the enhanced IO-capabilities of the Propeller will make it a better choice.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Don't visit my new website...
Post Edited (Gadgetman) : 2/18/2006 10:51:19 AM GMT
To make sure there isn't 100 winners only the first correct guess wins.
Bean.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95
http://www.parallax.com/detail.asp?product_id=30012
"SX-Video OSD module" Now available from Parallax for only·$49.95
http://www.parallax.com/detail.asp?product_id=30015
Product web site: www.sxvm.com
"Ability may get you to the top, but it takes character to keep you there."
Post Edited (Bean (Hitt Consulting)) : 2/18/2006 1:44:08 PM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Don't visit my new website...
I seems as if it may be a great chip for robotics.
also the hydra... Its main purpose seems to be to educate people on using multi-cell processors.
IMO, the chip will be around $10 give or take, and the basic dev kit will be around $50-$60.
I don't think Parallax is here to sell out.· The chip is not some Pentium whatever.· And all of the R&D is theres and not outsourced, which would probably help keep costs resonable.
I'm sure also that it was alot of fun for them the whole process (trips to china
another great product from a great company.
-Dan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
DTQ
·
First impressions: It is similar to your standard microcontroller (PIC or Atmel) but darn fast. It has eight parallel processors running at 80Mhz (20MIPS) each. That's 160 MIPS. Memory and IO access use a unique "cog" architecture under the hood, but it is mostly obfuscated unless you do direct assembly programming. For those of you that are experienced programmers, the main hurdle is going to be the new language they use. It looks a convergence of pascal, basic, and python. I sure hope someone comes up with a c compiler.
Fortunately, they have created a bunch of "objects" for doing all your standard stuff. They even have objects that can read a standard PS2 mouse and keyboard. One of their coolest objects allows you, without any extra hardware and a minimal amount of software, to create NTSC (or PAL) or VGA output directly. I'm not talking about boring B/W blocks and text, I'm talking about 64 color animated graphics and text. The best part is that fully animated video only takes two of the eight "cogs". So, you have 6 other cogs to do other stuff. You could do music, motion control, IO, anything. You could even create four separate video streams using all eight cogs. There is even an extra option to modulate the direct video signal and create a short range broadcast signal (picked up on channels 2-16). This works from about 20' away without any extra antenna (just radiating off of the chip, sorry FCC).
The final pricing hasn't been set yet, but raw chips (40 dip or 44 LQFP) should be around $20 each. Their whole dev. board kit should be around $200.
Posted by Phillip Torrone | 11:48 AM
Post Edited (Steve Joblin) : 2/18/2006 3:14:01 PM GMT
Why a C compiler?
C doesn't have any syntax that helps in developing multi-processor programs, so that means a less than ideal solution.
(I expect the Parallax IDE is tooled specifically for that)
As for Dev-kits, Al(bless him) already posted that what you NEED is a USB2SER adapter at $29, a EEPROM and a 3.3V voltage regulator.
What the 'pro' dev-kit seems to add is PS/2 connectors, an external Xtal, experimenter area and video output.
Knowing Parallax, they will most probably post the schematics for the board anyway, so that those who can't afford it can set up those circuits they need on an experimenter board or breadboard.
If the chip and board together is no more than $150, I'll buy the board.
(I assume that those boards will use the LQFP version as seen on the Demo-board pictures.)
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Don't visit my new website...
I'm sure there's some clever multi-processing syntax out there that someone's come up with, but in actual practice,·C's lack of built-in multi-processing concepts is really not a hinderance at all.
Plus, if you look at the Propeller as an "introduction to real-industry multiprocessing" as many of the posters in this thread have stated, then take a look at the mainstream game developers working on XBox360 and PS3. If you know a lot about the mainstream games industry, it's a VERY safe bet that they are not abandoning C/C++ in favor of special multi-processing languages. So using C on a Propeller·would be the closest thing to what they do.
EDIT: BTW, Hello to everyone·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Nick Sabalausky
Post Edited (Abscissa) : 2/18/2006 6:06:09 PM GMT
Sid
This sounds like the formula for imminent disaster. I have heard and read numerous software-related horror stories about Read-Write-Modify situations where the reading of a pin's register after writing to it came back with the stored result rather than the external signal.
Is there a safeguard built into the compiler(s) for this situation? Is there ANY way to monitor for this beyond creating a running list of which COG did what to each pin as the programmer writes his code?
Or am I seeing the sky falling?
Looks like the world will beat a path to your door with this design!
kenjj
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
Everybody,
many of the things we've done with the Propeller design take a completely different approach to things like multi-processing, interrupts, objects, etc.· Of course, some will love it, some will not.· Nothing's perfect, and perfection is subjective, but I think many people who·actually get their·hands on one, and take the time to learn it, will feel refreshed and unburdened.
As far as language, I know there are wars started on topics like this; very silly.· Let's not go there, but I have to say that if you are one to want to squeeze all the power out of something that you can, you are doing yourself a great disservice by not learning the native language of the device itself, but instead are relegating yourself to get only as far as 1) the·cleverness of the cross-compiler writer, and 2)·the limitations of a general-purpose language applied to an embedded system device, will allow you to go.· Yes, learning a new language is hard to swallow for some people, but
those who don't bother may be missing out on some very exciting stuff.· The Propeller's language is not difficult to learn or use for most people.
I would urge anyone who would stay away from the Propeller simply because there are no tools for the language of their choice, to just take a chance to learn something new; you may find you like it.· If you're successful, you've gained a lot, if you're not you've lost only a little.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
Post Edited (Jeff Martin) : 2/18/2006 11:58:28 PM GMT
This will allow me to take some of my projects that are multi processor (8 bit's PIC's & ARV's) and possibly implement the same functionality into one chip.
Looking forward to SPIN!
Mike
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"OEM NMEA GPS Module" Now available on ebay for only $17.49
Product web site: http://www.allsurplus.net/Axiom/
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
Can you say anything about RF? Can the propeller send and receive radio signals? If it can broadcast video, it can probably broadcast radio. What about receiving?
Thanks.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.speechchips.com
Speech & Video IC's for BasicStamps
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
The way I explained Propeller to my technophobic friends was this: Think of it as sliced bread. (The next best thing!) You have different spreads (rasberry, peanut butter, honey,...) and those you put on a slice of bread. The bread is the computer, and the spreads are the programs. Programs have different flavors, consider them tastes or tasks (eating). If you have one slice of bread and 8 spreads on that slice and 8 people wanting to taste, okay, well, it gets complicated. Only one person at time (without being awkward), and it becomes a real problem of how to arrange the spreads on the surface. Okay, well, compare that with 8 slices of bread, and each has its own spread, and 8 people can pass them around and taste in turn. That takes some coordination, but, remember, these are the technophobic friends and always ready to party!
While much of the discussion here has been about the high speed capabilities, I look to low power, long battery life stuff. That is fully taken into consideration in this chip with an amazing clocking scheme that can shift gears on the fly. Some commands can stop an individual cog's operation, while waiting for an event on a pin or from the system tick, thus reducing the current draw, while ready to continue execution in one tick. And propeller as a whole can run on its internal 20khz clock, so that it draw microamps of current but scan still monitors pins and ticks and can wake up quickly and do something and accelerate to high gear when necessary.
I toast Parallax!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
How many max MIPS can an individual cog run? Is it 20 MIPS as 'unofficially' stated in a previous post?
How many clock ticks per cycle?
How many clock cycles does each instruction take?
Are there any future plans for a scaled back (and lower priced) propeller chip, say with only 4 or 2 cogs? I'd guess that the bulk of the use of the propeller would be with less than 8 cogs....
Rick
If the clock is 80mhz and the mips is 20 then each instruction must take 4 clocks each.
I can see a day when the next generation of the chip comes out and it pipelines for one instruction per clock for 80 mips per cog.
So if each cog gets 2k bytes of ram/instruction space and the instructions are 32 bits then that gives you 512 instructions per cog. That is like having 8 sx28 chips in one package in a way.
Looking at the diagram, the hub can access 32k of ram and 32k of rom. The cogs must have a way of commanding the hub to transfer data to/from the global memory. I guess with a slight performance hit as compared to access the local 2k of ram allotted to each cog.
I say this because since the cog instruction space seems to be shared with the data space, this says that self modifying code is possible.
Thinking about it, you could partition the code in a way that it could overlay a part of the cog ram so you could swap in different routines from global ram, effectively expanding the local cog ram up to the limit of 32k with paging techniques.
I think that the sky is the limit here. I get really frustrated with chips that lock you into using the chip only one way. This chip really seems to let you be creative like no other before.
We all know of the programmable/reconfigurable hardware chips but rarely see them in use. Why? because sometimes it is hard to break out of the literal mind set. We want a chip to do one thing and one thing only.
You need DSP functions, You use a DSP, you need lots of IO you use an IO loaded chip.
Give us a chip that can be reconfigured on the fly and we fall a part trying to get our brains wrapped around the possibilities.
This chip seems to be what the industry needs. A cpu in the standard sense but with on the fly configuration capabilities. A bridge if you will. Use it a a· mega controller or morph it into something new.
I bet if you ask, 'can it do this?' or 'can it do that?', that the answer would boil down to, it can if you program it that way.
As far as size goes, 9mm/9mm would almost fit in my toothbrush so I guess it would fit in a cell phone. Besides, I'm sure a custom version could be created with less IO pins if the customer needed.
8 processors, not 8 threads but 8 processors. This chip is going to rock the controller world and set a new standard for future controllers.
Pricing, I'll toss my two cents into this.......
From the picture, I doubt the monitor will be included as part of the basic kit. Why? Because we all have TV/monitors around. Heck, that monitor is so small and cute it could be a product on it's own and they would make a mint selling it as an add on.
The board has the chip, two regulators, a crystal, a few caps, a breadboard, some extras like keyboard/mouse port (should be two ports, one for each I would think unless a splitter is included) and what seems to be a serial/usb programming key.
so...
Key = 30 bucks ( I think that has all ready been established).
Hardware (less chip) 20 bucks. Based on the sx48/52 with less parts at 10. Doubled it for fun.
Chip = 80 bucks, based on BS pricing. True, this beats the pants of BS chips but they must have priced BS chips at what they felt the market would tolerate. This would be introduced at the top end BS price and the BS chips would then get a price cut.
So my guess is going to be $130.00 to $150.00 range for a dev board (with programming key).
Post Edited (ToneDefJam) : 2/19/2006 2:58:01 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Martin Hebel
Southern Illinois University Carbondale -Electronic Systems Technologies
and
SelmaWare Solutions - StampPlot - Graphical Data Acquisition and Control
Personal Links with plenty of BASIC Stamp info
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.speechchips.com
Speech & Video IC's for BasicStamps
RoboGeek
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
There are·3 kinds of people in the world,
the dreamers, the do-ers, and the "Oh, what's this button do"-ers.
Formerly bugg.
www.parallax.com
www.goldmine-elec.com
www.expresspcb.com
www.startrek.com
·