This would allow packing 6 truncated ASCII characters per long, which is much more efficient than putting 4 ASCII characters in a 32-bit long. We don't really need lower case characters. Upper case should be sufficient. After all, the Spin language is case insensitive.
Excellent idea, never did like all those funky character encoding and don't start me on internationalization and unicode.
More seriously, who says a LONG can't be 40 bits. An extra 4 bits per address field or 4086 instructions per COG.
Or perhaps a LONG could be 48 bits for 128K instructions per COG.
This would gives us the bigger COG space we need whilst not going crazy with the transistor count and chip size.
As an extra benefit it would boost the precision of fixed point maths on the Prop. by a useful amount.
Heater, I understand what you are saying. However are you sure that is directly comparable for the legions of engineers who are out using uC's as uC', and not as small SOC's like it appears you need/ed?
I think the term 'embedded is starting to become a bit of a melange, with its roots in the mcu era slowly giving way to this SOC ARM/Mips/Power-PC Operating System on everything world.
I do think the Prop is a very cool, unique and daring even product. However its hard to see it as a real uC in the historical sense. If I as a layman, actually someone interested in the area since the mid-80's also when the 68K first came out, then I wonder about the bulk of the current engineers working in the field. For all of the devices out there that do not use or require an OS, I would expect they are not as familiar/comfortable with non-hard hardware peripherals.
I can certainly be wrong though.
We are in a growth economy. Well, arguably in the US it's something else, but let's put that in some politics corner, and just take growth in general as a given for this little discussion.
People are expensive, and they need stuff! The cost of that stuff generally goes up over time, as do materials, goods, services, etc...
It's not necessary for Parallax to grow to compete in share with the larger vendors. However, it is necessary for them to see growth in revenue, and margin (most importantly margin) over time, so that they can pay their people, deal with their costs of goods, and remain well capitalized, (money in the bank) so they can innovate as needed to keep the enterprise viable.
Two basic definitions:
Value is the product of labor. Labor costs time. Value comes from labor products that save time.
Wealth is innovation applied to labor over time. Innovation is that effort, or time investment, that improves the output of labor, increasing value over time.
Because others are innovating, the simple product of labor, or value, is a decreasing thing over time. Sometimes this is a slow ramp, and other times it's a rather fast ramp. In the tech sector, it's fast. For something like textiles, it's much slower.
So the cost of the prop includes all the labor and materials required to produce it, plus the cost of the next prop, and those investments required to maintain the business. Anything over and above that is real profit, and is the return on the effort for those that own it, and take the risk on it.
That's really why they must sell more. How much more is something they know, and it's linked to their costs and needs directly, and linked to the state of the industry niche they play in, less directly as stated above.
BigFoot,
Kye's FAT32 SD card drivers are very well written and fast, much faster than the one I wrote. All of the I2C drivers I've written are independent of the system clock frequency. They should run fine at 6.25MHz (100MHz). Pretty much all of the serial drivers will also work fine at that speed. Sphinx is an operating system for the Prop 1 that includes a Spin compiler / assembler. Its SD card I/O only supports FAT16, but that's good for up to 2GB cards.
Sphinx has the advantage that all of its I/O drivers are loaded into cogs so there's very little hub memory required for I/O freeing up most of the 32K for the Spin program.
It's planning ahead... Parallax is very aware of their sales of the BASIC STAMP and the Propeller. If more of us are starting to use the Propeller over the BASIC STAMP, then they simply need to increase volume because of the difference in price. Parallax has a strong desire to stay in the hobby, education markets, but the Propeller needs sales volume to keep the gears moving. If Parallax sells a million chips to a commercial product, it's win/win for everyone.
OBC, very astute though I don't think many people really understand the implications.
My understanding of Parallax' history is limited. I thought they originally had a mixed revenue stream based upon Education and Hobbiest. At some point I'm guessing that evolved to Education, Hobbiest, Prop Hobbiest, DIY/Robotics.
I thought I'd read somewhere that Parallax saw the majority of its revenue from the Educational/Hobbiest Basic Stamp, though I could be mistaken.
I assume this has been sufficient to pay the bills, not too mention expanding with some nice equipment, and R&D expansion which resulted in the Prop itself. But in the current age of uber-uC/Arduino/ARM/PSOC's dev kits currently available and often sold as a loss leader to the same institutions for future mind/market-share, I would expect this has got to have dropped Basic Stamp revenue significantly.
With this in mind, one has to use some common sense and figure that the BS revenue is at some point going to decline to irrelevency.
To survive, much less have the resources and capitol to invest in R&D of any significance requires a heavy hitter that can cover the revenue shortfalls of the old BS. Unless Parallax has already seen Prop/Robotics revenue grow enough to cover the losses from BS waning, that doesn't bode well mid-term to long-term. * I hope they have seen such growth in the Prop, or other areas, and that I am way off track.
OBC's 10 Prop to 1 BS, if its remotely true, is a bit mind numbing. I just wince every time I see a post by someone who says the Prop/Prop II is perfect for their needs, as a hobbiest...
Outside of the Parallax forum, and Nuts and Volts, I don't see Prop exposure that even approaches the old Basic Stamp's. Where then is the 10 to 1, 8 or 5 to 1 sales of Props either happening now, or to come from in the future? Outside of some of the forum posters with modest runs of dozens or maybe hundred/hundreds of units, the Hobbiest niche as a primary revenue source versus that being lost by a declining Basic Stamp would seem to be suspect. A move to really Commercialize the Prop is thus seen as rather a necessity to grow/maintain revenue.
Prop can not, and doesn't have to really fight ARM/PIC/AVR for market share as much as find a niche wherein it is a better solution by some demonstrable metric such as price (doubtful), RAD ( not sure, but doubtful for the average engineer), feature set (not sure, you can get some fast cheap chips), or some other metric.
I hate to be a downer, and realize I am not one of the long time regular posters and as such might be seen as trolling. I really am interested in the Prop enough to continue to hang around, even on Superbowl Sunday (blah).....
Without knowing the financials of Parallax, its impossible to say whether they are currently coasting along with adequate revenue forecast for the next several years enough to continue current operations/expenditures, or if the Basic Stamp revenue is shrinking without enough positive offset from Prop to make this commercial venture more necessity versus simple market expansion experiment.
I am guess it is the later out of ignorance of how successful their DIY/Robotics line is.
If I am right, all I am hoping is that some of my comments regarding commercial adoption might strike a chord in someone higher up.
In that vein, Parallax should really set up a Design Challenge in the forums, Nuts and Volts, Circuit Cellar, etc for some sort of industrial/medical/etc type of device prior to bringing their Commercial site online. That way they could have some premium examples of what the Prop can do here and now, today, versus what I consider negative examples such as the Hydra which lead down a certain assumption path we probably don't want to explore in the Commercial venue.
Either way, this should be interesting to see what being cooked up.
Cheers,
Fred
EDIT
I just wanted to add, I hope I am not seen to be trying to air dirty laundry or some such with my musings about Parallax' revenue etc. If so, please let me know.
But what I really wanted to mention was that I'd just had a thought about how Parallax might go about addressing the features of its primary chip that is both its most interesting and most confusing.
The programmable peripheral blocks on the Prop seem to me to be still new and unknown in the field, even though there is another PSOC that has something similar with configurable(?) periphs.
It struck me that the name really needs to both correct and well known if it is too pass the smell test of an engineer.
One could call them virtual peripherals, however virtual is just so over used.
Would not something like Shader Peripherals be both correct as software configurable AND a terminology now well entrenched amongst the technically literate?
Let me help you... It's all about profit margins. (These are not evil things, they make it possible to write paychecks and stay in business)
The Propeller is far more inexpensive than the BASIC STAMP, which has been Parallax's core product. The Propeller and it's successors are on the horizon. It takes a lot more Propellers sold to equate the sale of a STAMP. If all of us were willing to purchase 10 Propellers every time we needed to purchase a single unit, then there would be no problem. Truth is, Parallax simply needs to sell more. A commercial arm of the company will permit us to "have our cake, and eat it too."
If Parallax sells a million chips to a commercial product, it's win/win for everyone.
I would have to agree, however, if they don't sell a million chips to a commercial product, I would suggest that they create their own commercial product to show off the capabilities of the Propeller. And I would recommend a product within the process control and automation markets. Just my viewpoint of course.
They could even narrow down the market further by supplying ready to use solutions for the CNC industry. Instead of Parallax Semiconductor, how about CNC Solutions, Inc.(if the name is not already taken)
Bruce
P.S. With the available objects, it would be easy for Parallax and the Propeller to enter such a market.
What I remember of the Basic Stamp was that it was one of the very few 'kits' available. I'm not sure even how advanced the PIC was, but I don't remember it being sold in a similar kit form. I'm sure Leon will correct me.
I think the Basic Stamp series just happened at the right time, right place, Perfect Storm types of things. How well did the Javelin take off, comparably speaking?
The market is now beyond full, and most importantly, all the players in the market are following the same strategy of very low cost development kits, educational outreach programs, etc, etc ad naseum.
With so many factors now being relatively even across the playing field, its suicidal to try and live on past laurels. For many here, Parallax has a history undoubtedly. For the rest of the society of engineers and hobbiests, high-schoolers and undergrads, Parallax is only a quick google away, and it comes up with Basic Stamp, Educational Product, and the Prop. The first 2 are going to have autonomic reactions of ancient history and yeuchhh respectively, and the 3rd is going to be cool, and then confusion, if the common non-Parallax forum responses are taken as normative.
I would wager that most of the loyal clientele in education are now late in career, and probably continually pushed to get up to date with Arduino/ARM, etc.
Not trying to be argumentative, just seriously doubtful that past glories or even focusing on the hobbiest market with the current Prop is or will be successful long term.
My concerns may quite possibly be causing Chip and Ken to be ROFLTAO right now....
1. What's the average microcontroller product lifetime in the commercial/industrial arena?
2. How long does it take to go from cutting-edge to mature to "venerable" to obsolete?
3. Can Parallax Semiconductor's chip development process keep up with this schedule?
By focusing on a hobbyist and education market, Parallax has tapped a loyal clientele for whom products mature to obsolescence very slowly. (Witness the long-term success of the BASIC Stamp line.) This is possible because Parallax is very good at continuing to add value to mature products via accessories and a growing array of educational materials. As a consequence, new core products can be introduced at a fairly leisurely pace. (The Prop I, for example, is now five years old.) Now Parallax is about to enter the fast lane. Does this necessarily entail that a Prop III be already in gestation with a Prop IV close behind in the visualization stage? In short, how much does this new reality shorten the required design cycle, and can Parallax Semiconductor keep up?
On a side note, this of course entirely anecdotal evidence, but all my work collegues know about the Arduino, two of them even bought one. One the other hand the Prop is very obscure and unknown, in fact I didn't know it until yesterday when I was searching for a powerful uC on farnell in a DIL package. Then I started reading about the prop and decided to use it for my project.
I wasn't suggesting "resting on one's laurels" as a strategy going forward, but merely pointing out that the education/hobbyist market provides more breathing room for innovations to mature than the commercial/industrial sector. If Parallax's culture is geared to the former, they may find the latter to be rather more frenetic.
This post is in reference to my two previous posts.
Take for example the Propeller Proto Board, it has provisions for:
Keyboard
Mouse
Display
Of course the keyboard and mouse are for user input, which is very essential to CNC. And the display is the output for the user interface, which is also very essential to CNC.
The Propeller can multi-task which is essential to CNC.
The Propeller can accept input and provide output which is essential to CNC.
Add some relays, perhaps some contactors, sensors, and parsing code and BAM you have a new niche.
I am sure the DIP40 will remain for the Prop I but in the case of the Prop II you will have to use one of the many available DIP pcbs with the smt prop II mounted. Everyone will produce a version!
I totally agree with OBC on
A commercial arm of the company will permit us to "have our cake, and eat it too."
Peter: So we can buy cheaper chips and also so Chip can play to get a Prop III sometime around 2018
Dave Hein: Forget 6bit ASCII - I worked on the mini Singer/ICL that had 6bit ASCII (60bit instruction) and they went to 8 bit (80bit instruction). The 8bit was a huge success! There are other solutions, but not 6bit ASCII.
Why Prop II
Because there are a few limitations of the Prop I that could be solved with a new rev. But once the genie was opened, a lot of extras flowed in. Deficiencies: Counters (to be able to serial input) and lots of other things; Hub RAM, Hub windows (1:8 not 1:16); better DAC; more I/Os.
New extras: hub access (up to 4x32bit per cycle); way more powerful counters; DAC; fifo's (512B or 2KB???); more I/O (92), improved ADC & DAC etc on the I/O; faster 160MHz & 1 cycle instructions; I2C/SPI/SD booting.
And there will be some interesting use of the extra fifo per cog.
Cog RAM >2K
There are a number of ways and yet still stay mostly compatible.
The S & D addresses could be pointers to larger cog space. Relative jump/calls.
However, more cog ram is likely to be only required for 1 or maybe 2 cogs. Also, remember that spin will get an 8x boost in performance (single cycle, double clock). Add to that the fifo will be usable as a stack(s) (IIRC stated by Chip).
I know banked cog ram was discarded, but I still would like it on at least 1 cog (maybe Prop III, although I would like a 40/48/64bit lw). And I would like a preemtive or priority cog able to get more frequent hub access, or get access (bit settable) to use "free" hub cycles.
General
The prop will probably not compete with many other avr and pic chips, and certainly not with ARM with big OSes. However, there are niche markets that would use plenty of props. I see the prop as an extremely simple and powerful micro with which I can do a lot of things. I only have an inventory of 1 part. With base boards, it is easy to add other things. If I wanted to interface a widget I could bit-bang the code to do what I want simply, without fear of interfering with the main code, whatever that is doing. This is the beauty of this design.
I see a great future for the Prop I and II. Just need to get the world to realise it's available and what it's capable of.
idbruce: CNC - I couldn't agree more. I have two projects that have been on the drawing board for 8 months - a RepRap (micro-mendel) 3D plastic extruder and a QuadCopter. Both are extremely suited to the prop, yet they are based on arduino and derivations. Really, anything that requires multiple sources and outputs in parallel are suitable. Just a miniature OS and you are away. Of course, robotics is another prime example. And of course, as has been suggested, PLC replacements (~CNC).
From a user perspective, I really can't see the rush for Prop2. There is so much that we haven't done with the current Propeller. If I had a Prop2 sitting on my desk, (and I was honest about finishing everything I should learn on the Prop1) I would have to set it aside for a couple years and finish learning the Prop1. (That's IF I could finish learning/doing everything can be done with what I already have on my desk.)
All of this clamoring for Prop2... There is so much we can do with what we have already and haven't scratched the surface!
From a business perspective, the Prop2 makes sense, because so much has been learned with Prop1 and I expect that these lessons will push Prop2 so far ahead of the game it won't be funny. It's good stuff for Prop2.
Now, with nearly 4K views on this thread, it's flaming hot.. Quit reading. Go dig out your prop1 and get busy!
OBC
No.... Don't do it! Don't you hit reply! I said go grab your Propeller board..
No.... Don't do it! Don't you hit reply! I said go grab your Propeller board..
'Been doing that all day with the Spinneret! (What was that? There's some sort of game on TV today?) But it's still fascinating to return to this thread from time to time!
What game... Not here! I am trying to finish some pcbs and send them off for prodn. And yes, there is plenty of scope left to tweek the prop. Still got to find out what those counters/video can be used for other than video. And how we can kludge something to help input using them Pitty Chip is so busy or we could ask for more detail.
After reading the very interesting thread on the Propeller II, I had a few additional questions for the Parallax team:
- What is the (expected) transistor count (10M?), fabrication process (180nm?), power consumption at max. freq (160 MHz)
- For comparison, what is the transistor count of the Prop I and fabrication process (300nm?)
- What takes the most room on the Prop II? We know I/O requires 7000 transistors, I guess RAM is part that requires the most transistors? Does ROM require a lot of transistors too? (I guess not)
- What are the sales proportion of each package for the Prop I? Is it like 40% DIP, 50% SMD, 10% BGA?
- Could we get a regular DIL package with a reduced pin count? I know there will be modules available, but a DIP is important for people who like DIP.
- Are there any plans to increase for the Prop III the long size in bits, to like 40 bits? This seems like the most painless way to increase the Cog RAM size with minimal changes to the ISA.
- Are C3 board going to be restocked soon?
Still got to find out what those counters/video can be used for other than video.
Shamefully I have yet to make a more than cursory study of the Props video capabilities.
It seems to me that what you can do with the video logic is shift out parallel bit patterns at high speed. What could those bit patterns be doing apart from lighting up pixels?
Well what about driving external devices like RAM? Generally to drive such things one needs to mess with toggling chips selects, write enables, output enables etc. As well as putting out the address and data at the correct point in the sequence.
Could this be done by having the appropriate read and write sequences set up in memory permanently and then just blatting them out on waitvid, placing address and data on the I/O pins at the appropriate points?
Whew! Just got through this thread, which is quite a culmination of things discussed at one point or another in the past.
I've personally found the Prop to be quite easy to program for. IMO, PASM is easier to learn than C - though a bit time consuming, and the most efficient instructions for a given task aren't always the most obvious, though still significantly simpler than the SX, and especially something like the x86 architecture (granted, I've only invested a couple hours before I wanted to smash my head through the table). To be fair, as someone who's never written any substantial programs, I'm not really qualified to make any comments on SPIN's/PASM's shortcomings.
Ken has mentioned Parallax's long-term commitment to the Propeller, which is particularly important in the industrial sector where product lifetime is substantially longer than the cutting-edge throw away nature of the consumer market. I've worked for a company that seemingly constantly struggled with component EoL issues, requiring considerable financial investments just to keep a product rolling. This assurance alone can prove to be invaluable to many potential customers.
As several people mentioned here, and as I've said in other threads, the marketing strategy of Parallax's needs to be addressed. For instance, I was working for a company that was phasing out the use of ASICs for their main product. To me, the Propeller seemed like the ideal solution providing the multitasking capability that was needed (instead, they went with an ARM -based uC and had to use a simple RTOS that used up almost all available memory). Unsurprisingly, none of the engineers heard of the Propeller, yet as soon as I mentioned Parallax, a couple of them responded with something along the lines of "Oh, that company that makes robotic toys", and dismissed my recommendation without the chance of rebuttal. I have a feeling that I'm not the only one that this happened to either. As much as I hate to say it, I think Parallax should somewhat disassociate their name from their IC offerings.
Though it is unlikely that Parallax will (ever) implement a Prop chip with an ARM based primary processor, I do think that at least something like this should be pursued. Since the P2 cogs are shaping up to be pretty killer, I don't think there will be an issue with a lack of processing power, but there's certainly going to be plenty of moaning about the 512 Long limit. Without going to paged memory, or using Longs larger than 32-bits, one of the COGs can be given more frequent hub access (perhaps even given it's own hub space). It could be implemented efficiently enough, the RDxx instruction would not be that much different from moving data to the ALU as many processors seem to require. I imagine the only major change to the current architecture would be in the hub logic. The only way I think this is possible without affecting the hub access frequency of all the other COGs, or using some fancy ram for the hub memory, would be to give at least one of the COGs it's own hub space not accessible by the other COGs. And in order for the COG to access the shared hub space, an offset of the hub address can be used which is then automatically resolved by the hub logic. Now, that's not to say that it couldn't be implemented in every COG, but that could just lead to wasted hub ram.
After reading the very interesting thread on the Propeller II, I had a few additional questions for the Parallax team:
- What is the (expected) transistor count (10M?), fabrication process (180nm?), power consumption at max. freq (160 MHz)
- For comparison, what is the transistor count of the Prop I and fabrication process (300nm?)
- What takes the most room on the Prop II? We know I/O requires 7000 transistors, I guess RAM is part that requires the most transistors? Does ROM require a lot of transistors too? (I guess not)
- What are the sales proportion of each package for the Prop I? Is it like 40% DIP, 50% SMD, 10% BGA?
- Could we get a regular DIL package with a reduced pin count? I know there will be modules available, but a DIP is important for people who like DIP.
- Are there any plans to increase for the Prop III the long size in bits, to like 40 bits? This seems like the most painless way to increase the Cog RAM size with minimal changes to the ISA.
- Are C3 board going to be restocked soon?
Thanks,
I could be way off, but I thought Chip, or Beau said the Prop2 would be using the 45, or 65nm process.
Looking at images of the original Propeller die, it looks like the ram (hub + cog) used up most of the silicon real estate.
A Prop2 DIP even at reduced pin count seems unlikely. Parallax has mentioned this before, but hey, you never know.
PropIII? If Chip got any funny new ideas, they would probably wind up in the Prop2.
I could be way off, but I thought Chip, or Beau said the Prop2 would be using the 45, or 65nm process.
Looking at images of the original Propeller die, it looks like the ram (hub + cog) used up most of the silicon real estate.
A Prop2 DIP even at reduced pin count seems unlikely. Parallax has mentioned this before, but hey, you never know.
PropIII? If Chip got any funny new ideas, they would probably wind up in the Prop2.
Thanks. Do you have recommendations for a good book on ASIC design? Not that I'm going to design an ASIC anytime soon (I'm sticking to my Spartan 3), but I find the whole subject fascinating.
Hmmmmm..... Yea, I can understand that mindset, however, I remember reading about toy manufacturers during world war II converting their operations towards military defense. Parallax will now have to get the Propeller battle ready to defend it's rightful place in the history of semiconductors. The Propeller is far more lethal than the BasicStamp, and adversaries should not underestimate it's power and potential, otherwise they might possibly lose a fair or substantial market share.
But that was not your point. Your point was that your colleagues just dismissed it. And that is truly the MAIN obstacle.
In order for Parallax to become a serious contender, they will have to fight several battles. And I think they should pick their battles very strategically and only fight the battles they know they can win. This will help build a reputation for the Propeller. Once the reputation has been achieved and has gathered an active audience, then and only then should they fight the title fight. At which point they should find the biggest and baddest chip on the block and attack it's weaknesses.
I think we need to maintain some perspective and not come across as "fan boys".
One of the things Ken mentioned it another thread, I believe it was referenced earlier in this one when someone pointed out that Leon mentioned that chip which shall not be named, that Parallax was not about attacking competitors or doing comparisons. I think for some applications the prop stands on it's own merit, for others not so much, that isn't a slam on the prop, no component can be the right fit for every problem.
Comments
Excellent idea, never did like all those funky character encoding and don't start me on internationalization and unicode.
More seriously, who says a LONG can't be 40 bits. An extra 4 bits per address field or 4086 instructions per COG.
Or perhaps a LONG could be 48 bits for 128K instructions per COG.
This would gives us the bigger COG space we need whilst not going crazy with the transistor count and chip size.
As an extra benefit it would boost the precision of fixed point maths on the Prop. by a useful amount.
I think the term 'embedded is starting to become a bit of a melange, with its roots in the mcu era slowly giving way to this SOC ARM/Mips/Power-PC Operating System on everything world.
I do think the Prop is a very cool, unique and daring even product. However its hard to see it as a real uC in the historical sense. If I as a layman, actually someone interested in the area since the mid-80's also when the 68K first came out, then I wonder about the bulk of the current engineers working in the field. For all of the devices out there that do not use or require an OS, I would expect they are not as familiar/comfortable with non-hard hardware peripherals.
I can certainly be wrong though.
People are expensive, and they need stuff! The cost of that stuff generally goes up over time, as do materials, goods, services, etc...
It's not necessary for Parallax to grow to compete in share with the larger vendors. However, it is necessary for them to see growth in revenue, and margin (most importantly margin) over time, so that they can pay their people, deal with their costs of goods, and remain well capitalized, (money in the bank) so they can innovate as needed to keep the enterprise viable.
Two basic definitions:
Value is the product of labor. Labor costs time. Value comes from labor products that save time.
Wealth is innovation applied to labor over time. Innovation is that effort, or time investment, that improves the output of labor, increasing value over time.
Because others are innovating, the simple product of labor, or value, is a decreasing thing over time. Sometimes this is a slow ramp, and other times it's a rather fast ramp. In the tech sector, it's fast. For something like textiles, it's much slower.
So the cost of the prop includes all the labor and materials required to produce it, plus the cost of the next prop, and those investments required to maintain the business. Anything over and above that is real profit, and is the return on the effort for those that own it, and take the risk on it.
That's really why they must sell more. How much more is something they know, and it's linked to their costs and needs directly, and linked to the state of the industry niche they play in, less directly as stated above.
Thanks Mike,
I will pass this on to our firmware guys.
It's planning ahead... Parallax is very aware of their sales of the BASIC STAMP and the Propeller. If more of us are starting to use the Propeller over the BASIC STAMP, then they simply need to increase volume because of the difference in price. Parallax has a strong desire to stay in the hobby, education markets, but the Propeller needs sales volume to keep the gears moving. If Parallax sells a million chips to a commercial product, it's win/win for everyone.
OBC
My understanding of Parallax' history is limited. I thought they originally had a mixed revenue stream based upon Education and Hobbiest. At some point I'm guessing that evolved to Education, Hobbiest, Prop Hobbiest, DIY/Robotics.
I thought I'd read somewhere that Parallax saw the majority of its revenue from the Educational/Hobbiest Basic Stamp, though I could be mistaken.
I assume this has been sufficient to pay the bills, not too mention expanding with some nice equipment, and R&D expansion which resulted in the Prop itself. But in the current age of uber-uC/Arduino/ARM/PSOC's dev kits currently available and often sold as a loss leader to the same institutions for future mind/market-share, I would expect this has got to have dropped Basic Stamp revenue significantly.
With this in mind, one has to use some common sense and figure that the BS revenue is at some point going to decline to irrelevency.
To survive, much less have the resources and capitol to invest in R&D of any significance requires a heavy hitter that can cover the revenue shortfalls of the old BS. Unless Parallax has already seen Prop/Robotics revenue grow enough to cover the losses from BS waning, that doesn't bode well mid-term to long-term. * I hope they have seen such growth in the Prop, or other areas, and that I am way off track.
OBC's 10 Prop to 1 BS, if its remotely true, is a bit mind numbing. I just wince every time I see a post by someone who says the Prop/Prop II is perfect for their needs, as a hobbiest...
Outside of the Parallax forum, and Nuts and Volts, I don't see Prop exposure that even approaches the old Basic Stamp's. Where then is the 10 to 1, 8 or 5 to 1 sales of Props either happening now, or to come from in the future? Outside of some of the forum posters with modest runs of dozens or maybe hundred/hundreds of units, the Hobbiest niche as a primary revenue source versus that being lost by a declining Basic Stamp would seem to be suspect. A move to really Commercialize the Prop is thus seen as rather a necessity to grow/maintain revenue.
Prop can not, and doesn't have to really fight ARM/PIC/AVR for market share as much as find a niche wherein it is a better solution by some demonstrable metric such as price (doubtful), RAD ( not sure, but doubtful for the average engineer), feature set (not sure, you can get some fast cheap chips), or some other metric.
I hate to be a downer, and realize I am not one of the long time regular posters and as such might be seen as trolling. I really am interested in the Prop enough to continue to hang around, even on Superbowl Sunday (blah).....
Without knowing the financials of Parallax, its impossible to say whether they are currently coasting along with adequate revenue forecast for the next several years enough to continue current operations/expenditures, or if the Basic Stamp revenue is shrinking without enough positive offset from Prop to make this commercial venture more necessity versus simple market expansion experiment.
I am guess it is the later out of ignorance of how successful their DIY/Robotics line is.
If I am right, all I am hoping is that some of my comments regarding commercial adoption might strike a chord in someone higher up.
In that vein, Parallax should really set up a Design Challenge in the forums, Nuts and Volts, Circuit Cellar, etc for some sort of industrial/medical/etc type of device prior to bringing their Commercial site online. That way they could have some premium examples of what the Prop can do here and now, today, versus what I consider negative examples such as the Hydra which lead down a certain assumption path we probably don't want to explore in the Commercial venue.
Either way, this should be interesting to see what being cooked up.
Cheers,
Fred
EDIT
I just wanted to add, I hope I am not seen to be trying to air dirty laundry or some such with my musings about Parallax' revenue etc. If so, please let me know.
But what I really wanted to mention was that I'd just had a thought about how Parallax might go about addressing the features of its primary chip that is both its most interesting and most confusing.
The programmable peripheral blocks on the Prop seem to me to be still new and unknown in the field, even though there is another PSOC that has something similar with configurable(?) periphs.
It struck me that the name really needs to both correct and well known if it is too pass the smell test of an engineer.
One could call them virtual peripherals, however virtual is just so over used.
Would not something like Shader Peripherals be both correct as software configurable AND a terminology now well entrenched amongst the technically literate?
I would have to agree, however, if they don't sell a million chips to a commercial product, I would suggest that they create their own commercial product to show off the capabilities of the Propeller. And I would recommend a product within the process control and automation markets. Just my viewpoint of course.
Bruce
They could even narrow down the market further by supplying ready to use solutions for the CNC industry. Instead of Parallax Semiconductor, how about CNC Solutions, Inc.(if the name is not already taken)
Bruce
P.S. With the available objects, it would be easy for Parallax and the Propeller to enter such a market.
What I remember of the Basic Stamp was that it was one of the very few 'kits' available. I'm not sure even how advanced the PIC was, but I don't remember it being sold in a similar kit form. I'm sure Leon will correct me.
I think the Basic Stamp series just happened at the right time, right place, Perfect Storm types of things. How well did the Javelin take off, comparably speaking?
The market is now beyond full, and most importantly, all the players in the market are following the same strategy of very low cost development kits, educational outreach programs, etc, etc ad naseum.
With so many factors now being relatively even across the playing field, its suicidal to try and live on past laurels. For many here, Parallax has a history undoubtedly. For the rest of the society of engineers and hobbiests, high-schoolers and undergrads, Parallax is only a quick google away, and it comes up with Basic Stamp, Educational Product, and the Prop. The first 2 are going to have autonomic reactions of ancient history and yeuchhh respectively, and the 3rd is going to be cool, and then confusion, if the common non-Parallax forum responses are taken as normative.
I would wager that most of the loyal clientele in education are now late in career, and probably continually pushed to get up to date with Arduino/ARM, etc.
Not trying to be argumentative, just seriously doubtful that past glories or even focusing on the hobbiest market with the current Prop is or will be successful long term.
My concerns may quite possibly be causing Chip and Ken to be ROFLTAO right now....
I wasn't suggesting "resting on one's laurels" as a strategy going forward, but merely pointing out that the education/hobbyist market provides more breathing room for innovations to mature than the commercial/industrial sector. If Parallax's culture is geared to the former, they may find the latter to be rather more frenetic.
-Phil
This post is in reference to my two previous posts.
Take for example the Propeller Proto Board, it has provisions for:
- Keyboard
- Mouse
- Display
Of course the keyboard and mouse are for user input, which is very essential to CNC. And the display is the output for the user interface, which is also very essential to CNC.The Propeller can multi-task which is essential to CNC.
The Propeller can accept input and provide output which is essential to CNC.
Add some relays, perhaps some contactors, sensors, and parsing code and BAM you have a new niche.
Bruce
I totally agree with OBC on Peter: So we can buy cheaper chips and also so Chip can play to get a Prop III sometime around 2018
Dave Hein: Forget 6bit ASCII - I worked on the mini Singer/ICL that had 6bit ASCII (60bit instruction) and they went to 8 bit (80bit instruction). The 8bit was a huge success! There are other solutions, but not 6bit ASCII.
Why Prop II
Because there are a few limitations of the Prop I that could be solved with a new rev. But once the genie was opened, a lot of extras flowed in. Deficiencies: Counters (to be able to serial input) and lots of other things; Hub RAM, Hub windows (1:8 not 1:16); better DAC; more I/Os.
New extras: hub access (up to 4x32bit per cycle); way more powerful counters; DAC; fifo's (512B or 2KB???); more I/O (92), improved ADC & DAC etc on the I/O; faster 160MHz & 1 cycle instructions; I2C/SPI/SD booting.
And there will be some interesting use of the extra fifo per cog.
Cog RAM >2K
There are a number of ways and yet still stay mostly compatible.
The S & D addresses could be pointers to larger cog space. Relative jump/calls.
However, more cog ram is likely to be only required for 1 or maybe 2 cogs. Also, remember that spin will get an 8x boost in performance (single cycle, double clock). Add to that the fifo will be usable as a stack(s) (IIRC stated by Chip).
I know banked cog ram was discarded, but I still would like it on at least 1 cog (maybe Prop III, although I would like a 40/48/64bit lw). And I would like a preemtive or priority cog able to get more frequent hub access, or get access (bit settable) to use "free" hub cycles.
General
The prop will probably not compete with many other avr and pic chips, and certainly not with ARM with big OSes. However, there are niche markets that would use plenty of props. I see the prop as an extremely simple and powerful micro with which I can do a lot of things. I only have an inventory of 1 part. With base boards, it is easy to add other things. If I wanted to interface a widget I could bit-bang the code to do what I want simply, without fear of interfering with the main code, whatever that is doing. This is the beauty of this design.
I see a great future for the Prop I and II. Just need to get the world to realise it's available and what it's capable of.
All of this clamoring for Prop2... There is so much we can do with what we have already and haven't scratched the surface!
From a business perspective, the Prop2 makes sense, because so much has been learned with Prop1 and I expect that these lessons will push Prop2 so far ahead of the game it won't be funny. It's good stuff for Prop2.
Now, with nearly 4K views on this thread, it's flaming hot.. Quit reading. Go dig out your prop1 and get busy!
OBC
No.... Don't do it! Don't you hit reply! I said go grab your Propeller board..
'Been doing that all day with the Spinneret! (What was that? There's some sort of game on TV today?) But it's still fascinating to return to this thread from time to time!
-Phil
Been coding SIDPLAYER and my Noname OS tonight.. TV only useful for connecting my Prop.
OBC
After reading the very interesting thread on the Propeller II, I had a few additional questions for the Parallax team:
- What is the (expected) transistor count (10M?), fabrication process (180nm?), power consumption at max. freq (160 MHz)
- For comparison, what is the transistor count of the Prop I and fabrication process (300nm?)
- What takes the most room on the Prop II? We know I/O requires 7000 transistors, I guess RAM is part that requires the most transistors? Does ROM require a lot of transistors too? (I guess not)
- What are the sales proportion of each package for the Prop I? Is it like 40% DIP, 50% SMD, 10% BGA?
- Could we get a regular DIL package with a reduced pin count? I know there will be modules available, but a DIP is important for people who like DIP.
- Are there any plans to increase for the Prop III the long size in bits, to like 40 bits? This seems like the most painless way to increase the Cog RAM size with minimal changes to the ISA.
- Are C3 board going to be restocked soon?
Thanks,
Underwater Cricket, Australia v Pakistan:)
Ouch..bad joke I know, could not resist.
Shamefully I have yet to make a more than cursory study of the Props video capabilities.
It seems to me that what you can do with the video logic is shift out parallel bit patterns at high speed. What could those bit patterns be doing apart from lighting up pixels?
Well what about driving external devices like RAM? Generally to drive such things one needs to mess with toggling chips selects, write enables, output enables etc. As well as putting out the address and data at the correct point in the sequence.
Could this be done by having the appropriate read and write sequences set up in memory permanently and then just blatting them out on waitvid, placing address and data on the I/O pins at the appropriate points?
Is there any gain in such a scheme?
I've personally found the Prop to be quite easy to program for. IMO, PASM is easier to learn than C - though a bit time consuming, and the most efficient instructions for a given task aren't always the most obvious, though still significantly simpler than the SX, and especially something like the x86 architecture (granted, I've only invested a couple hours before I wanted to smash my head through the table). To be fair, as someone who's never written any substantial programs, I'm not really qualified to make any comments on SPIN's/PASM's shortcomings.
Ken has mentioned Parallax's long-term commitment to the Propeller, which is particularly important in the industrial sector where product lifetime is substantially longer than the cutting-edge throw away nature of the consumer market. I've worked for a company that seemingly constantly struggled with component EoL issues, requiring considerable financial investments just to keep a product rolling. This assurance alone can prove to be invaluable to many potential customers.
As several people mentioned here, and as I've said in other threads, the marketing strategy of Parallax's needs to be addressed. For instance, I was working for a company that was phasing out the use of ASICs for their main product. To me, the Propeller seemed like the ideal solution providing the multitasking capability that was needed (instead, they went with an ARM -based uC and had to use a simple RTOS that used up almost all available memory). Unsurprisingly, none of the engineers heard of the Propeller, yet as soon as I mentioned Parallax, a couple of them responded with something along the lines of "Oh, that company that makes robotic toys", and dismissed my recommendation without the chance of rebuttal. I have a feeling that I'm not the only one that this happened to either. As much as I hate to say it, I think Parallax should somewhat disassociate their name from their IC offerings.
Though it is unlikely that Parallax will (ever) implement a Prop chip with an ARM based primary processor, I do think that at least something like this should be pursued. Since the P2 cogs are shaping up to be pretty killer, I don't think there will be an issue with a lack of processing power, but there's certainly going to be plenty of moaning about the 512 Long limit. Without going to paged memory, or using Longs larger than 32-bits, one of the COGs can be given more frequent hub access (perhaps even given it's own hub space). It could be implemented efficiently enough, the RDxx instruction would not be that much different from moving data to the ALU as many processors seem to require. I imagine the only major change to the current architecture would be in the hub logic. The only way I think this is possible without affecting the hub access frequency of all the other COGs, or using some fancy ram for the hub memory, would be to give at least one of the COGs it's own hub space not accessible by the other COGs. And in order for the COG to access the shared hub space, an offset of the hub address can be used which is then automatically resolved by the hub logic. Now, that's not to say that it couldn't be implemented in every COG, but that could just lead to wasted hub ram.
I could be way off, but I thought Chip, or Beau said the Prop2 would be using the 45, or 65nm process.
Looking at images of the original Propeller die, it looks like the ram (hub + cog) used up most of the silicon real estate.
A Prop2 DIP even at reduced pin count seems unlikely. Parallax has mentioned this before, but hey, you never know.
PropIII? If Chip got any funny new ideas, they would probably wind up in the Prop2.
Thanks. Do you have recommendations for a good book on ASIC design? Not that I'm going to design an ASIC anytime soon (I'm sticking to my Spartan 3), but I find the whole subject fascinating.
I thought it would be 90 nm.
Hmmmmm..... Yea, I can understand that mindset, however, I remember reading about toy manufacturers during world war II converting their operations towards military defense. Parallax will now have to get the Propeller battle ready to defend it's rightful place in the history of semiconductors. The Propeller is far more lethal than the BasicStamp, and adversaries should not underestimate it's power and potential, otherwise they might possibly lose a fair or substantial market share.
But that was not your point. Your point was that your colleagues just dismissed it. And that is truly the MAIN obstacle.
In order for Parallax to become a serious contender, they will have to fight several battles. And I think they should pick their battles very strategically and only fight the battles they know they can win. This will help build a reputation for the Propeller. Once the reputation has been achieved and has gathered an active audience, then and only then should they fight the title fight. At which point they should find the biggest and baddest chip on the block and attack it's weaknesses.
Bruce
What is that, currently?
I will leave that question for the experts, I am just trying to be a creative journalist
Bruce
P.S. I would answer "The Propeller", but that would be counterproductive for the Propeller to fight itself.
One of the things Ken mentioned it another thread, I believe it was referenced earlier in this one when someone pointed out that Leon mentioned that chip which shall not be named, that Parallax was not about attacking competitors or doing comparisons. I think for some applications the prop stands on it's own merit, for others not so much, that isn't a slam on the prop, no component can be the right fit for every problem.
C.W.