Give me a chance ... to do the optimization. I've already "rediscovered" for example that bit 8 is not the best one to use for chained reset and will probably move it to bit 9 [noparse]:)[/noparse] Also, I think I can remove the I2C communication requirement at start up completely, but I2C connections are still necessary for booting from EEPROM. Once all Propellers are booted, I believe can use the I2C clock bit for asynchronous messaging and eliminate the fairly slow I2C "device code" completely which would be a big (wish I thought of this last week ... uuugh!).
Adding the I2C slave code was very complicated and took up another COG. In retrospect, it looks like it was a big waste of time and resources. ... not very minimum
> for a minimum config, I think a one wire interface would be the easiest.
Guess that depends on the one wire interface software [noparse]:)[/noparse]
That would be a horrible bottleneck in a 1280 MIPS system (8 Propellers). Not sure what the point of that would be.
> Not quite sure what you are asking? "clock-slaved"??
Aren't you using Beau's high-speed serial link that depends on all Propellers sharing the same clock phase?
I'm curious about the aggregate throughput of N links.
@Ray (Cluso99), or anyone else who has used Beau's high-speed serial object.
I'm also curious about the number of simultaneous multiple high speed serial links (14Mbps?) because the most optimistic data rate I can achieve with one COG on an 8 bit bus between Propellers at any one time would be 26.4Mbps. I have it working now at 16 to 18Mbps. Adding COGs for Parallel bus would complicate bus timing in a big way.
Having two 14Mbps links active simultaneously would be 28Mbps aggregate throughput at any one time which beats my most optimistic 8 bit parallel bus data rate. Of course the application data setup/extraction would play huge role in data transfer performance, but assuming they are equal, the fully implemented serial mesh would have superior data transfer performance, but would have fewer free pins (48? vs 136) per 8-Propeller system. Of course one could create a 2-host serial mesh and have free pins in a similar context.
Yes Phil, I guess it's time to for me to look at that 14Mbps code [noparse]:)[/noparse] Pas touche'
Comments
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve
Propeller Tools
even if it means sacrificing some speed. I call it
Ultra Minimum.
humanoido
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve
Propeller Tools
humanoido
Not quite sure what you are asking?·· "clock-slaved"??
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBlade,·RamBlade, RetroBlade,·TwinBlade,·SixBlade, website
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index) ZiCog (Z80) , MoCog (6809)
· Search the Propeller forums·(uses advanced Google search)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
Guess that depends on the one wire interface software [noparse]:)[/noparse]
That would be a horrible bottleneck in a 1280 MIPS system (8 Propellers). Not sure what the point of that would be.
> Not quite sure what you are asking? "clock-slaved"??
Aren't you using Beau's high-speed serial link that depends on all Propellers sharing the same clock phase?
I'm curious about the aggregate throughput of N links.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve
Propeller Tools
I'm also curious about the number of simultaneous multiple high speed serial links (14Mbps?) because the most optimistic data rate I can achieve with one COG on an 8 bit bus between Propellers at any one time would be 26.4Mbps. I have it working now at 16 to 18Mbps. Adding COGs for Parallel bus would complicate bus timing in a big way.
Having two 14Mbps links active simultaneously would be 28Mbps aggregate throughput at any one time which beats my most optimistic 8 bit parallel bus data rate. Of course the application data setup/extraction would play huge role in data transfer performance, but assuming they are equal, the fully implemented serial mesh would have superior data transfer performance, but would have fewer free pins (48? vs 136) per 8-Propeller system. Of course one could create a 2-host serial mesh and have free pins in a similar context.
Yes Phil, I guess it's time to for me to look at that 14Mbps code [noparse]:)[/noparse] Pas touche'
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve
Propeller Tools
Find it here: http://forums.parallax.com/showthread.php?p=839672
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve
Propeller Tools
- Propalyzer: Propeller PC Logic Analyzer
- BMA: An on-chip PASM Debugger
- SPUD: Spin Source Level Debugger
Post Edited (jazzed) : 9/10/2009 3:30:45 PM GMT