[ibis-macro] Minutes from the 18 May 2010 and 25 May 2010 ibis-atm meetings
- From: "Mike LaBonte (milabont)" <milabont@xxxxxxxxx>
- To: <ibis-macro@xxxxxxxxxxxxx>
- Date: Thu, 27 May 2010 11:22:34 -0500
Minutes from the 18 May 2010 and 25 May 2010 ibis-atm meetings are
attached.
Mike
IBIS Macromodel Task Group
Meeting date: 18 May 2010
Members (asterisk for those attending):
Adge Hawes, IBM
* Ambrish Varma, Cadence Design Systems
* Anders Ekholm, Ericsson
* Arpad Muranyi, Mentor Graphics Corp.
Barry Katz, SiSoft
* Bob Ross, Teraspeed Consulting Group
Brad Brim, Sigrity
Brad Griffin, Cadence Design Systems
Chris Herrick, Ansoft
Chris McGrath, Synopsys
Danil Kirsanov, Ansoft
David Banas, Xilinx
Deepak Ramaswany, Ansoft
Donald Telian, consultant
Doug White, Cisco Systems
* Eckhard Lenski, Nokia-Siemens Networks
Eckhard Miersch, Sigrity
Essaid Bensoudane, ST Microelectronics
* Fangyi Rao, Agilent
Ganesh Narayanaswamy, ST Micro
Gang Kang, Sigrity
Hemant Shah, Cadence Design Systems
Ian Dodd, consultant
Jerry Chuang, Xilinx
Joe Abler, IBM
* John Angulo, Mentor Graphics
John Shields, Mentor Graphics
* Ken Willis, Sigrity
Kellee Crisafulli, Celsionix
Kumar Keshavan, Sigrity
Lance Wang, Cadence Design Systems
Luis Boluna, Cisco Systems
Michael Mirmak, Intel Corp.
* Mike LaBonte, Cisco Systems
Mike Steinberger, SiSoft
Mustansir Fanaswalla, Xilinx
Patrick O'Halloran, Tiburon Design Automation
Paul Fernando, NCSU
Pavani Jella, TI
Radek Biernacki, Agilent (EESof)
* Randy Wolff, Micron Technology
Ray Komow, Cadence Design Systems
Richard Mellitz, Intel
Richard Ward, Texas Instruments
Samuel Mertens, Ansoft
Sam Chitwood, Sigrity
Sanjeev Gupta, Agilent
* Scott McMorrow, Teraspeed Consulting Group
Shangli Wu, Cadence Design Systems
Sid Singh, Extreme Networks
Stephen Scearce, Cisco Systems
Steve Kaufer, Mentor Graphics
Steve Pytel, Ansoft
Syed Huq, Cisco Systems
Syed Sadeghi, ST Micro
Ted Mido, Synopsys
Terry Jernberg, Cadence Design Systems
* Todd Westerhoff, SiSoft
Vladimir Dmitriev-Zdorov, Mentor Graphics
Vikas Gupta, Xilinx
Vuk Borich, Agilent
Walter Katz, SiSoft
Wenyi Jin, LSI Logic
Zhen Mu, Mentor Graphics
------------------------------------------------------------------------
Opens:
- Ambrish asked about LiveMeeting issues
--------------------------
Call for patent disclosure:
- No one declared a patent.
-------------
Review of ARs:
- Arpad: Update section 2 text based on updated AMI flow
- In progress
- Bob: Discuss ISS spec review process at Open Forum
- Michael Mirmak accepted the editing task
- Arpad: Write parameter passing syntax proposal (BIRD draft)
for *-AMS models in IBIS that is consistent with the
parameter passing syntax of the AMI models
- Would like to discuss in this meeting
- TBD: Propose a parameter passing syntax for the SPICE
- [External ...] also?
- TBD
- Arpad: Review the documentation (annotation) in the macro libraries.
- Deferred until a demand arises or we have nothing else to do
-------------
New Discussion:
Arpad showed the new AMI flow diagrams:
- Spelling errors were fixed
- "StatEye type analysis" changed to "Statistical analysis"
- There were no further comments on this
Arpad showed the updated IBIS-AMI Flow Correction BIRD:
- Statement of the Issue is updated
- All of section 2 is replaced
- Arpad had compared the files side by side with WinMerge
- Fangyi: On page 6, 2nd paragraph of Step 2:
- This seems to imply statistical analysis
- Arpad: There is some difference between the text and the flow drawing
- This is a fine point
- Ken: What if Tx uses a modified impulse response and Rx uses GetWave?
- Statistical analysis would shut off the GetWave part
- Todd: This gives channel output without EQ
- Ken: You want to allow that
- Todd: Right, we don't want to call things invalid if not necessary
- Fangyi: This is about how the tool interprets it, not the real value
- Use_Init_Output is not valid here
- Todd and Ambrish agreed
- Todd: The slide is titled "statistical"
- Arpad: Each file has it's own Use_Init_Output
- Ambrish: Each is valid for it's own model
- Arpad: How will a result be displayed if it is not used?
- The 2nd note is the problem here
- The tool has to disregard TRUE or FALSE settings
- Todd: That flow is really talking only about time domain
- Fangyi: There should be separate flows for statistical and time domain
- Ambrish: It would be best to disallow it, not disregard
- Todd: GetWave flags have nothing to do with statistical
- These are "don't cares" for that
- The spec does not elaborate that
- Fangyi: Agree
- Ken: Agree
- It should be valid if only one of Tx and Rx has GetWave
- Arpad: The flags on slide 2 are required to be meaningful
- Todd: Are people assuming flow #7 means you can't do statistical?
- Note 1 means if both are FALSE then the model does nothing
- In time domain the model does nothing
- Walter: Have to watch out for TX and RX usage
- Used on the input to TX GetWave, not output
- Todd: The effect of the impulse response must be felt
- Ambrish: Page 145 says EDA tool must use it
- Arpad: This has nothing to do with Init output
- Each AMI is associated with Tx or Rx
- It is not clear which one the flag applies to on this page
- Todd: The letter of the law here is a little broken in places
Arpad showed pages 144 & 145 of the IBIS spec:
- Arpad: Two sentences were combined into one
- Together these give us Note 1
- Ambrish: The logic is not correct
- Arpad: The policy is to avoid having conditions where there is no result
- Todd: Sometimes one model is Init only and the other is GetWave only
- We have neglected this case
- The Rx EQ effect will be missing
- Arpad showed slide 7:
- We enumerated all cases
- Todd: That is for time domain only
- John was right in calling for at least a warning
- Ken: Can we run statistical analysis when one is missing?
- Scott: There could be an approximation used?
- Mike: EDA tools should warn as they see fit
- But we should explain the issues
- Ambrish: We can't predict all possible problem scenarios
- John: We need to have these cases understood
- Arpad: We can't count on GetWave doing something in a statistical analysis
- Walter: An Rx might have only clock recovery
- That would be fine for statistical analysis
- There would simply be no clock ticks from GetWave
- Arpad: So we should not force Use_Init_Output TRUE
- Ken: Right
AR: Arpad update AMI flow diagram and text
Walter: Would like to get my 4 BIRDs published
AR: Mike post BIRDs from Walter
Next meeting: 1 June 2010 12:00pm PT
--------
IBIS Interconnect SPICE Wish List:
1) Simulator directives
IBIS Macromodel Task Group
Meeting date: 18 May 2010
Members (asterisk for those attending):
Adge Hawes, IBM
Ambrish Varma, Cadence Design Systems
* Anders Ekholm, Ericsson
* Arpad Muranyi, Mentor Graphics Corp.
Barry Katz, SiSoft
* Bob Ross, Teraspeed Consulting Group
Brad Brim, Sigrity
Brad Griffin, Cadence Design Systems
Chris Herrick, Ansoft
Chris McGrath, Synopsys
* Danil Kirsanov, Ansoft
David Banas, Xilinx
Deepak Ramaswany, Ansoft
Donald Telian, consultant
Doug White, Cisco Systems
* Eckhard Lenski, Nokia-Siemens Networks
Eckhard Miersch, Sigrity
Essaid Bensoudane, ST Microelectronics
* Fangyi Rao, Agilent
Ganesh Narayanaswamy, ST Micro
Gang Kang, Sigrity
Hemant Shah, Cadence Design Systems
Ian Dodd, consultant
Jerry Chuang, Xilinx
Joe Abler, IBM
* John Angulo, Mentor Graphics
John Shields, Mentor Graphics
Ken Willis, Sigrity
Kellee Crisafulli, Celsionix
Kumar Keshavan, Sigrity
Lance Wang, Cadence Design Systems
Luis Boluna, Cisco Systems
* Michael Mirmak, Intel Corp.
* Mike LaBonte, Cisco Systems
Mike Steinberger, SiSoft
Mustansir Fanaswalla, Xilinx
Patrick O'Halloran, Tiburon Design Automation
Paul Fernando, NCSU
Pavani Jella, TI
Radek Biernacki, Agilent (EESof)
* Randy Wolff, Micron Technology
Ray Komow, Cadence Design Systems
Richard Mellitz, Intel
Richard Ward, Texas Instruments
Samuel Mertens, Ansoft
Sam Chitwood, Sigrity
Sanjeev Gupta, Agilent
Scott McMorrow, Teraspeed Consulting Group
Shangli Wu, Cadence Design Systems
Sid Singh, Extreme Networks
Stephen Scearce, Cisco Systems
Steve Kaufer, Mentor Graphics
Steve Pytel, Ansoft
Syed Huq, Cisco Systems
Syed Sadeghi, ST Micro
Ted Mido, Synopsys
Terry Jernberg, Cadence Design Systems
* Todd Westerhoff, SiSoft
Vladimir Dmitriev-Zdorov, Mentor Graphics
Vikas Gupta, Xilinx
Vuk Borich, Agilent
* Walter Katz, SiSoft
Wenyi Jin, LSI Logic
* Zhen Mu, Mentor Graphics
------------------------------------------------------------------------
Opens:
- Mike L: The IQ group has an interest in the ISS spec
- It might allow more flexible circuits for a future [External Test Load]
keyword
--------------------------
Call for patent disclosure:
- No one declared a patent.
-------------
Review of ARs:
- Arpad: Rework AMI flow diagram
- Done
- Mike: Compile comments on ISS from minutes
- Done
- Arpad: Write parameter passing syntax proposal (BIRD draft)
for *-AMS models in IBIS that is consistent with the
parameter passing syntax of the AMI models
- TBD
- TBD: Propose a parameter passing syntax for the SPICE
- [External ...] also?
- TBD
- Arpad: Review the documentation (annotation) in the macro libraries.
- Deferred until a demand arises or we have nothing else to do
-------------
New Discussion:
Arpad showed the updated AMI flow diagrams:
- Arpad: Happy with this arrangement
- Slide 1:
- This incorporates most if not all comments
- The first 2 pages are about statistical analysis
- the rest are time domain
- The reason boolean inputs are retained:
- So boxes don't move around from page to page
- Shows the expected inputs for statistical analysis
- Note 2 rule added: GetWave_Exists always treated as false
- Mike L: Is there a better way to say "StatEye type analysis"?
- The Ts for GetWave_Exists might be Xs
- Todd: "Statistical analysis" will do
- StatEye does something more specific, but there should be no confusion
- Arpad: The change to "statistical" is OK
- Would rather leave the Ts (True) to clarify how we treat it
- Todd: It has to be clear that the model has to be configured a certain
way to support statistical
- Arpad: We said we would ignore Use_Init_Output for statistical analysis
- That is a detail
- Slide 2:
- Arpad: This is indicating convolution of waveform
- Fangyi: Why is the box repeated?
- Arpad: It is less cluttered that way
- Slide 3:
- Arpad: The statistical box is gone from bottom here
- Rx Init output goes to EDA tool convolution
- Have to run Rx convolution twice
- Todd: This assumes digital input to Tx GetWave?
- Arpad: Probably
- Fangyi: It says that
- Slide 4:
- Arpad: Mostly the booleans change from the previous slide
- Slide 5:
- Arpad: This is the most complete - both GetWaves
- Again no statistical box
- This is time domain
- Slide 6:
- Arpad: This has just the truth table
- Arpad: It needs the correct text explanation
- Walter: After thorough review I think this flow is correct
- Todd: This shows analytical flows
- We must not infer that statistical is valid only in that one case
- Arpad: The grayed out boxes are the indication
- Todd: Maybe "StatEye" is better for slide 1
- Fangyi: Statistical has some ambiguity
- Maybe "probability" is OK
- Arpad: "Statistical" might be ambiguous
- There may be a new form of analysis at some point
- Todd: "Statistical" is OK
Arpad: Fangyi had commented about a flow issue:
- Fangyi: It comes to the same conclusion
- When Use_Init_Output=true it emphasized return impulse is not passed
into the next step
- Todd: Change "it is" to "they are" in note 2
Arpad: Looked at ISS meeting minutes
- Arpad made a short list of significant items
- Mike L: Looking at the ISS document, it has obvious format problems
- Arpad: I will send out my comments
- Bob: There are artifacts and open questions embedded
- Walter: That all needs to be done
- We should be able to ask for a content review by all of IBIS
- Mike L: This is not like the formal BIRD process
- We are not keeping it secret
- Arpad: By finishing in subcommittee we would avoid a can of worms
- Bob: Not concerned about too much review
- Walter: We might formulate a plan for this at the upcoming summit
- Arpad: We should find a technical writer to handle this
- Then we could focus on AMI
- Walter: Wonderful idea
- Bob: We should be able to just tidy it up a bit
- We are just capturing a portion of HSPICE
- Would like someone to volunteer to clean this up
- Walter: We have 3 more meetings before the summit
- SiSoft would like to preview our summit presentation here
- We should focus on getting the flow approved
- Also decide on what our committee will present
- The summit can help us decide how to proceed
- Arpad: I will continue working on the flows
Next meeting: 25 May 2010 12:00pm PT
--------
IBIS Interconnect SPICE Wish List:
1) Simulator directives
Other related posts:
