( ESNUG 459 Item 9 ) -------------------------------------------- [12/14/06]
Subject: ( ESNUG 458 #1 ) An unofficial PrimeTime-SI vs. CeltIC conspiracy
> Our Signal Integrity (SI) flow involves using Cadence CeltIC to perform
> cross-talk and noise analysis. CeltIC writes out the incremental SDF that
> is read into PrimeTime. When we view the timing reports, the incremental
> SDF values are lumped in with the delay values calculated from the SPEF.
> From these reports we cannot determine if a path is violating due to
> cross-talk. So we have some internal scripts that translate the SDF into
>
> set_annotated_delay -delta_only
>
> commands. We read these into plain vanilla PrimeTime and use
>
> report_timing -crosstalk_delta
>
> to be able to see the incremental SDF delays in a delta column. Then we
> can see if a violating path violates due to crosstalk, which is critical
> information when you get to the point of making semi-manual fixes to close
> timing. ... Taking away this very valuable capability and forcing us to
> buy a much more expensive PT-SI license is outrageous!
>
> - Dean Ranger
> Toshiba San Jose, CA
From: [ Bender of Futurama ]
Hi, John,
Hmm, that sounds like a pretty cool trick. Too bad it went away. We've
just been doing it the old-fashioned way, you know, reading in the design
with and without the CeltIC SDF and comparing the timing reports. But
since the SI in PT-SI is no good, I suppose Synopsys needs a new way to
sell those licenses. Anon please.
- [ Bender of Futurama ]
---- ---- ---- ---- ---- ---- ----
From: Dean.Ranger <dean.ranger=user domain=taec.toshiba spot mom>
To: John Cooley <jcooley=user domain=zeroskew spot mom>
Hi John,
I guess I should thank you for posting my 7/19/06 email about being outraged
they'd taken that feature out of PrimeTime. I burst out laughing when I saw
ESNUG 458 this morning because of the incredible irony: I had literrally
just finished this morning verifying that Synopsys had indeed "done the
right thing" and put that feature back into PrimeTime! (version 2006.06-SP3)
Are you sure you're not 'secretly' trying to give Synopsys some good press?
After a couple of months of hassling with Synopsys they agreed to put that
feature back. I would have told you sooner except that since I hadn't heard
any response to my 7/19/06 rant I figgured you were just ignoring it.
But seriously: I very much appreciate ESNUG. It's truely a uniquely useful
forum for our industry. Keep up the good work!
- Dean Ranger
Toshiba San Jose, CA
---- ---- ---- ---- ---- ---- ----
From: John Cooley <jcooley=user domain=zeroskew spot mom>
To: Dean.Ranger <dean.ranger=user domain=taec.toshiba spot mom>
Hi, Dean,
This is too funny! What had happened was that I found your letter in my
email "in" box just recently. You had sent it to me around DAC time and it
had gotten lost in the DAC noise. I was feeling guilty, so I put it
immediately on the top of ESNUG 458; hoping you'd get the attention you
rightfully deserved. I wanted to make up for my bad.
But it's at least good to know Synopsys, Inc. was listening to you and that
you got this fixed, Dean. Congrats!
- John Cooley
ESNUG/DeepChip.com Hollistion, MA
---- ---- ---- ---- ---- ---- ----
From: Dean.Ranger <dean.ranger=user domain=taec.toshiba spot mom>
To: John Cooley <jcooley=user domain=zeroskew spot mom>
Hi John,
Thanks for the explanation. That is funny. I'm not too surprised my email
got overlooked in the DAC mayhem. I'm impressed you finally got around to
it. And thanks for trying to "hold their feet to the fire."
- Dean Ranger
Toshiba San Jose, CA
Index
Next->Item
|
|
