is we is... or is we aint?
kbash
Posts: 117
The silence here is more frightening than all the "loud" arguments that went back and forth when we were waiting for the megawatt P2 ...
so I guess this question isn't completely out of the question is ... is something resembling a P2 really gonna happen?
When... where... what? some of us out here have waited for years and years... ( at great sacrifice to our lives and business ) can anyone giving us a REAL idea about what is actually happening with a new processor right now?
Ken.... Chip... anyone in authority.... please, we've believed in you for a very long time. Please... what's happening?
so I guess this question isn't completely out of the question is ... is something resembling a P2 really gonna happen?
When... where... what? some of us out here have waited for years and years... ( at great sacrifice to our lives and business ) can anyone giving us a REAL idea about what is actually happening with a new processor right now?
Ken.... Chip... anyone in authority.... please, we've believed in you for a very long time. Please... what's happening?
Comments
See here
It's probably for the best that this section is quiet as it lets Chip finish this project without distraction.
Be patient. Getting anxious won't change anything or make it go faster.. trust me. I'm one of the people you mentioned..
+2
Remember that it's a small group of people developing the P2. I get the impression that Chip knows what he wants and is working very hard at getting the new chip designed and ready for testing.
Now we should remember that is is not http://www.professionaladviser.com/ifaonline/news/1314165/fis-private-equity-renamed-isis-capital, and even the simple form "is" has a new meaning that has our responsible adults scurrying to catalog all web occurrences.
So, the question should have been "r we or rnt we?" To which I would reply: You can't spell Parallax without an "r."
You can't spell "Parallax" without "lax".
Oh, never mind.
And you can't spell Parallax without a para....
Just let Chip be... and we will all have a nice Christmas. I am looking forward to Tackyon Forth of the Propeller2.
It'll have sixteen cogs a blazin' when it comes!
It'll be quite the sight to see as it bit bangs USB
It'll be coming 'round the mountain when it comes!
Sorry, only joking.
What was the point here again? Oh, yeah...Are we there yet?
so, there ya go!
en reference to rasperry pi, beaglebone, & Galileo.
However I don't think the Propellers are to be compared to Raspberry Pi, Beagle Bone, etc. They are completely different machines. Those are boards not chips. They are designed to be full up computers running Linux or other OS, not micro-controllers.
It's like comparing KIA cars to a London buses.
More valid comparisons are to be made with the small ARM Cortex micro-controllers, the STM32F4 and such like. Those are micro-controllers. of a similar size and price, with on chip RAM and FLASH. They don't have many processor cores but that is to a large extent offset by having a ton of hardware peripherals on chip, UART, SPI, PWM, I2C, USB, RTC, etc etc. Not to mention floating point hardware.
The Prop II has some very stiff competition in the 32 bit micro-controller space today. As a single example look at this: http://fi.mouser.com/ProductDetail/STMicroelectronics/STM32F4DISCOVERY/?qs=J2qbEwLrpCGdWLY96ibNeQ== That's a whole dev board for a 32 bit floating point MCU for only 12 euro. There are many other similar examples. The processors can be had for prices down to less than a dollar. It's crazy stuff.
The Propellers though will be a hell of a lot easier to program multi-tasking software for.
A couple of years back they started renaming "cores" to "tiles" and "threads" to "logical cores" in all their marketing material. This is a huge pile of marketing dooie by XMOS which really annoys myself and many others. Note that little word "logical" in there. It's an underhanded deception.
But yes, I agree, the XMOS devices are directly comparable to the Propeller II and will be stiff competition.
So by way of a couple quick comparison points:
XMOS cores only have 64K of RAM each. RAM in one core cannot be accessed by another.
XMOS pins are dedicated to cores. Pins on one core cannot be used by another. There is also complication about how the way XMOS pins are grouped into "ports"
Heater,
I'll omit "only", specially because the implied device against which you are comparing the XMOS have "only" 2K of RAM per core and that memory is also non accessible by others cores.
Is the 512K of shared ram that can be accessed by any core and until you don't have an official (hw) hubexec capable to "(semi)transparently" execute code from it at roughly the same core speed this, IMHO, is to be considered only data space, even if with some tricks (read LMM) you find a way to execute it at a cost of big speed penalties.
I completely agree with other points.
Yes, these things are not so simple to compare.
For example in one view the Propeller 2 will have 512K of shared RAM, accessible by all cores, and each core has 512 internal registers.
With that view in mind we have the comparison:
1) Prop II has 512K of RAM for code and data. Accessible by all cores.
2) XMOS has, in the 4 core ("Tile") models, only 256K. Which is divided up between cores. The low end single tile devices only have 64K total.
3) If we take the Propellers cog space into account we have a 32K more space. What a bonus!
The P2 is a winner in the code/data space comparison. Hence my "only".
You do have a point, if outright speed if an issue and your programs are bigger than COG sized, the the Spin byte codes or C's LMM may not cut it.
See here
Yesterday, I got the booter program in the Prop2 talking to the PC. There seems to be one hang-up, yet, inhibiting loader downloading. I've been flushing minor bugs out for the last few days and I think I've maybe one or two to fix before I can start downloading from PNut.exe.
Once I'm downloading, I'm going to get hubexec working. After that, the only think I'll need to add to the core is the hub-streaming I/O stuff. Then, there should just be some cog refinement work, the hub CORDIC system, and the pin smarts.
Things have been going slowly, but once I get the ability to download from PNut.exe, I know things are going to accelerate.
I'm glad to see that you guys are still here!
Ummm... I don't think we could even be chased off.
refresh,refresh,refresh....
Still here. Looking forward to Chipmas again this year. If if it is only the FPGA image to check out.
Waiting quietly and patiently
Glad you're on the home straight .