Minutes from the 13 February ibis-atm meeting are attached.
IBIS Macromodel Task Group
Meeting date: 13 February 2018
Members (asterisk for those attending):
ANSYS: Dan Dvorscak
* Curtis Clark
Broadcom (Avago): Xingdong Dai
Bob Miller
Cadence Design Systems: Ambrish Varma
Brad Brim
Kumar Keshavan
Ken Willis
eASIC: David Banas
Marc Kowalski
Ericsson: Anders Ekholm
GlobalFoundries: Steve Parker
IBM Luis Armenta
Trevor Timpane
Intel: * Michael Mirmak
Keysight Technologies: Fangyi Rao
* Radek Biernacki
Ming Yan
Maxim Integrated Products: Hassan Rafat
Mentor, A Siemens Business: John Angulo
* Arpad Muranyi
Micron Technology: * Randy Wolff
* Justin Butterfield
SiSoft: * Walter Katz
Todd Westerhoff
* Mike LaBonte
SPISim: * Wei-hsing Huang
Synopsys: Rita Horner
Kevin Li
Teraspeed Consulting Group: Scott McMorrow
Teraspeed Labs: * Bob Ross
TI: Alfred Chong
The meeting was led by Arpad Muranyi. Curtis Clark took the minutes.
--------------------------------------------------------------------------------
Opens:
- None.
-------------
Review of ARs:
- Arpad to send BIRD189.5_draft16_v4 to the Interconnect and ATM lists.
- Done.
- Bob to propose language to address his concerns about fallback to legacy
package models with BIRD189.
- In progress.
--------------------------
Call for patent disclosure:
- None.
-------------------------
Review of Meeting Minutes:
- Arpad: Does anyone have any comments or corrections? [none]
- Walter: Motion to approve the minutes.
- Michael M.: Second.
- Arpad: Anyone opposed? [none]
-------------
New Discussion:
BIRD189.5_draft17_v1
- Discussion: Arpad introduced an example/question that had been discussed
during public and private Interconnect Group email exchanges. Arpad noted
that he thought a single Interconnect Model could not span two different
regions for different pins (e.g. pin-to-pad for some pins and pin-to-buffer
for others). He noted that Randy had raised some examples that suggested
that this restriction could be an inconvenience to him. He asked Randy if
he could elaborate. Randy described an example in which he had a full
package model for I/O and power pins, but did not have a detailed on-die
power delivery or I/O interconnect model. The on-die model consisted only of
decoupling caps as the PDN. His concern was whether the only way to handle
this example with the BIRD189 syntax was to create two models. One goes from
pin to pad and contains the full package model. The second model is necessary
to create the pad to buffer model for the PDN, and the inconvenience would be
the need to add a bunch of low-impedance paths to connect all the I/O pins
just so the PDN could be provided. Walter noted that two models would be
necessary, but the on-die model need only contain the PDN and did not have to
include low-impedance paths for all the I/O pins. Walter noted the following
line on page 10 of draft17_v1:
If an *_I/O pin_name appears only in a pin to die pad Interconnect Model
in the Interconnect Model Group, then the *_I/O pin_name electrical path
from the die pad to buffer shall be shorted.
Walter noted that this statement applies to any and all individual pins in the
model, so in Randy's example the I/O pins would be automatically shorted from
pad to buffer by the EDA tool. Randy agreed that this was an acceptable
solution for his example.
- Discussion: The group began reviewing and attempting to resolve the
outstanding comment fields in the current draft. Arpad started with this
sentence, which precedes his Figure XX2 illustrating the ambiguity of a pin
occurring multiple times as Aggressor_Only but never as a victim:
However, the rules defined in the Usage Rules section of the [Interconnect
Model Group] keyword above instruct the EDA vendor to select the first
model (marked as green) for the simulation of pin 4 in this case.
Arpad noted that the Rule to which the sentence referred had been deleted.
Walter noted that such a model should be considered incomplete, and we need
not worry about specifying how EDA tools deal with incomplete models. The
group agreed that we need only remove this sentence, as the previous sentence
already pointed out the ambiguity.
During the review, Mike L. and Randy noticed that we were reviewing topics
that had already been addressed. Mike L. then realized that Michael M. had
sent out a draft17_v2, but he had only sent it to Mike L. Arpad and Mike L.
took the AR to fold any changes discussed in the meeting into draft17_v2 to
create draft17_v3.
- Mike L.: Motion to adjourn.
- Curtis: Second.
- Arpad: Thank you all for joining.
AR: Arpad and Mike L. to fold changes discussed during the meeting into
draft17_v2 to create draft17_v3.
-------------
Next meeting: 20 February 2018 12:00pm PT
-------------
IBIS Interconnect SPICE Wish List:
1) Simulator directives