Yikes. That's not very much room for any mistake at sub-200nm nodes... I know Chip and Beau have been working on the Propeller II rather hard for several years... Only to be dealt a blow due to a short on the PII die.
And, the short's around the CPU core or inside it? If inside, I guess it's Game Over.
Hopefully, Chip and Beau get the chance to fix the problems. Sorry that it happened. T-T
We learned today that TSMC's next C018LP shuttle run isn't until mid-July. That means four more months before we'll have another chip to try out. I figured there'd be another shuttle any day, but it's not so.
I'm not a backend guy, but isn't there a tool that should catch this sort of thing? e.g. LVS, or GDS to RTL/gate formal equivalence checking? I do know of some former coworkers that got burned when a mask was improperly generated for the process they were using. Something to do with wet etch versus dry etch. It supposedly took someone many weeks or months to figure out what went wrong.
I can ask our backend guys how they would catch this sort of thing if you would like me to. They just taped out a couple of chips, so hopefully they have some time.
I'm not a backend guy, but isn't there a tool that should catch this sort of thing? e.g. LVS, or GDS to RTL/gate formal equivalence checking? I do know of some former coworkers that got burned when a mask was improperly generated for the process they were using. Something to do with wet etch versus dry etch. It supposedly took someone many weeks or months to figure out what went wrong.
I can ask our backend guys how they would catch this sort of thing if you would like me to. They just taped out a couple of chips, so hopefully they have some time.
See posts #21 and #48 in this thread for an answer to that question.
Two months hopefully should be enough time to verify the fill layer alignment (or redoing the offending layer) before shuttle is up for the taking.
That's a rather educational process, however. Hopefully the Silicon gremlins go away this time.
ADDED: The pie chart style of CPU cores (COGs) arrangement seems to be a rather smart idea, concerning the IO delay / trace travel timing, as well as being auto-routed. That may have avoided the needs for true superscalar integer units (probably delayed until Propeller III). What's of that have the potential of turning Propeller II into an efficient number cruncher. I will be giving it a try once it's fixed and ready - just curious how fast it can go compared to Propeller 1 (140 MHz being the record - correct me if I am wrong).
See posts #21 and #48 in this thread for an answer to that question.
I figured that something like Encounter Conformal Equivalence Checker should be able to handle it. Maybe there's a better way for this type of error. Anyways just offering to ask some backend guys that do many chips a year in 40 nm and below processes.
Sorry to hear that the first testing batch didn't work.
To the creators and engineers of Prop2, please keep going and don't give up. Hopefully you will create this magnificient product which is the Prop2.
Since I'm the person who have a bit of a sixth sense, I know that the new Prop2 is going to rock the world, and I will give all my blessings to Parallax Inc.
We learned today that TSMC's next C018LP shuttle run isn't until mid-July. That means four more months before we'll have another chip to try out. I figured there'd be another shuttle any day, but it's not so.
zapping through the net about parallel processing I found this forum and started reading this thread.
I have no real clue what exactly went wrong, anyhow please allow some words:
After reading this thread I decided to buy a Propeller and start learning!
Unfortunately it was a bit of pain to find here in Bahrain, but it is now on my desk ;-)
To make it short:
The open and honest way you guys have disclosed the trouble you are in, no pointing "it was the fault of..." and the way of explaining simply impressed me.
Even though I am only a hobby enthusiast that, most probably, never will generate decent revenue to parallax (sorry for that) I want to thank you!
The open and honest way you guys have disclosed the trouble you are in, no pointing "it was the fault of..." and the way of explaining simply impressed me.
Even though I am only a hobby enthusiast that, most probably, never will generate decent revenue to parallax (sorry for that) I want to thank you!
Best Regards
Frank
Welcome to the family Frank! There''s a lot of people that feel the same way you do.
Even though I am only a hobby enthusiast that, most probably, never will generate decent revenue to parallax (sorry for that) I want to thank you!
Best Regards
Frank
Hello Frank, and welcome to the crowd! You are most welcome and that's okay if you only end up needing a few chips. It takes a lot of peanuts to fill the bag, and having many customers ensures a broad base of support for Propeller 1 (and 2). I admit to knowing little about Bahrain so I had to research it myself http://en.wikipedia.org/wiki/Bahrain. You might be the first Propeller programmer in Bahrain - hopefully you have a spare Propeller, EEPROM and crystal in case something goes wrong (normal day at my desk - yesterday I hooked up a 12V 15Ahr SLA battery in reverse to my power supply and the cables burned apart). We are here to help even if you need another chip!
thanks a lot for the warm welcome.
In fact it seems that the PE Kit I bought I got the only propeller which is available in Bahrain.
Tells me I need to be extra careful :-D
I have started looking into spin (can't stop me comparing it with occam) and will search for the beginners section here in the Forum now.
Ok enough off topic for this thread ;-)
As I like the atmosphere here in the Forum you might see me more often.
Have fun and please excuse my english. This happens if germans try to write in a foreign language.
Frank
To make it short:
The open and honest way you guys have disclosed the trouble you are in, no pointing "it was the fault of..." and the way of explaining simply impressed me.
Even though I am only a hobby enthusiast that, most probably, never will generate decent revenue to parallax (sorry for that) I want to thank you!
Best Regards
Frank
Willkommen. Eigentlich habe ich ein paar Freunde, die Art durch alle Arten von Kommunikation und dergleichen auf die Spur zu Menschen, die Probleme verursachen. Sie haben mir versichert, dass dies einzig Ausfall war die Schuld dieses Mannes: http://forums.parallax.com/image.php?u=41324&dateline=1289257809
Unfortunately, I didn't save the original.... such a loss.
Google from German to English doesn't sound at all like me:)
Welcome. Actually, I have a few friends that way through all forms of communication, and the like to track down people who cause the problems. They assured me that this failure was only the guilt of this man: http://forums.parallax.com/image.php...ine=1289257809
Of course, my friends can not associate names and pictures ... but this man knows who he is and what he has done
If you have the most fun here and want to take more money to spend on Chinese products, you must follow the Erco posts.
Check out Mike Green posts. He's been around forever and is almost always right. And of course all the presenters get involved in great conversations, so you can select one and read his (or her) submissions.
Spin is fabulous.
You will never regret your interest in the propeller ... unless you end up with a warehouse full of Chinese products and have no idea what to do with all
I'm guessing your German is better than Google's English too:)
Welcome to the fold, Frank. Parallax has been doing business this way for 20+ years and a lot of us here have been customers for much of that period. They are a really wonderful company to do business with because they really do care about our needs, and have many times proven a willingness to go an extra kilometer when it doesn't profit them but can help one of us.
Comments
...go back to the vendor who reduced the core logic and require that they do the density fill for Poly and Diffusion layers?
...tweak the FILL cells you modified?
...deal individually with each instance?
And, the short's around the CPU core or inside it? If inside, I guess it's Game Over.
Hopefully, Chip and Beau get the chance to fix the problems. Sorry that it happened. T-T
I can ask our backend guys how they would catch this sort of thing if you would like me to. They just taped out a couple of chips, so hopefully they have some time.
See posts #21 and #48 in this thread for an answer to that question.
That's a rather educational process, however. Hopefully the Silicon gremlins go away this time.
ADDED: The pie chart style of CPU cores (COGs) arrangement seems to be a rather smart idea, concerning the IO delay / trace travel timing, as well as being auto-routed. That may have avoided the needs for true superscalar integer units (probably delayed until Propeller III). What's of that have the potential of turning Propeller II into an efficient number cruncher. I will be giving it a try once it's fixed and ready - just curious how fast it can go compared to Propeller 1 (140 MHz being the record - correct me if I am wrong).
I figured that something like Encounter Conformal Equivalence Checker should be able to handle it. Maybe there's a better way for this type of error. Anyways just offering to ask some backend guys that do many chips a year in 40 nm and below processes.
To the creators and engineers of Prop2, please keep going and don't give up. Hopefully you will create this magnificient product which is the Prop2.
Since I'm the person who have a bit of a sixth sense, I know that the new Prop2 is going to rock the world, and I will give all my blessings to Parallax Inc.
Glass half full: More time to get ready for it, more drivers
It will be worth the wait - please don't sweat it Chip!
Can you do ESD testing on the Prop2 you have ?
Good point! Yes, we can, but we'll only be able to check for power/ground-to-pin shorts on the I/O's after ESD hits.
.. and also measure changes to I/O leakage ? - which can be an earlier indication of getting close to the limits ?
The 99% perspiration comes after the 1% inspiration. And so, I can wait patiently.
I don't mind the wait, as long as this chip gets completed!
Keep up the good work!
zapping through the net about parallel processing I found this forum and started reading this thread.
I have no real clue what exactly went wrong, anyhow please allow some words:
After reading this thread I decided to buy a Propeller and start learning!
Unfortunately it was a bit of pain to find here in Bahrain, but it is now on my desk ;-)
To make it short:
The open and honest way you guys have disclosed the trouble you are in, no pointing "it was the fault of..." and the way of explaining simply impressed me.
Even though I am only a hobby enthusiast that, most probably, never will generate decent revenue to parallax (sorry for that) I want to thank you!
Best Regards
Frank
Welcome to the family Frank! There''s a lot of people that feel the same way you do.
Doc
Hello Frank, and welcome to the crowd! You are most welcome and that's okay if you only end up needing a few chips. It takes a lot of peanuts to fill the bag, and having many customers ensures a broad base of support for Propeller 1 (and 2). I admit to knowing little about Bahrain so I had to research it myself http://en.wikipedia.org/wiki/Bahrain. You might be the first Propeller programmer in Bahrain - hopefully you have a spare Propeller, EEPROM and crystal in case something goes wrong (normal day at my desk - yesterday I hooked up a 12V 15Ahr SLA battery in reverse to my power supply and the cables burned apart). We are here to help even if you need another chip!
Sincerely,
Ken Gracey
thanks a lot for the warm welcome.
In fact it seems that the PE Kit I bought I got the only propeller which is available in Bahrain.
Tells me I need to be extra careful :-D
I have started looking into spin (can't stop me comparing it with occam) and will search for the beginners section here in the Forum now.
Ok enough off topic for this thread ;-)
As I like the atmosphere here in the Forum you might see me more often.
Have fun and please excuse my english. This happens if germans try to write in a foreign language.
Frank
Willkommen. Eigentlich habe ich ein paar Freunde, die Art durch alle Arten von Kommunikation und dergleichen auf die Spur zu Menschen, die Probleme verursachen. Sie haben mir versichert, dass dies einzig Ausfall war die Schuld dieses Mannes:
http://forums.parallax.com/image.php?u=41324&dateline=1289257809
Nat
Hey Rich thanks for the welcome.
To answer your guessing:
Can you please post the original (english) text as well?
See you around
Frank (is reading about VGA projects and communication between props at the moment)
Unfortunately, I didn't save the original.... such a loss.
Google from German to English doesn't sound at all like me:)
Welcome. Actually, I have a few friends that way through all forms of communication, and the like to track down people who cause the problems. They assured me that this failure was only the guilt of this man:
http://forums.parallax.com/image.php...ine=1289257809
Of course, my friends can not associate names and pictures ... but this man knows who he is and what he has done
If you have the most fun here and want to take more money to spend on Chinese products, you must follow the Erco posts.
Check out Mike Green posts. He's been around forever and is almost always right. And of course all the presenters get involved in great conversations, so you can select one and read his (or her) submissions.
Spin is fabulous.
You will never regret your interest in the propeller ... unless you end up with a warehouse full of Chinese products and have no idea what to do with all
I'm guessing your German is better than Google's English too:)
Regards,
Rich