IBIS Macromodel Task Group Meeting date: 29 November 2016 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 Cisco: Seungyong (Brian) Baek 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 Graphics: John Angulo * Arpad Muranyi Micron Technology: * Randy Wolff Justin Butterfield QLogic Corp.: James Zhou Andy Joy SiSoft: * Walter Katz Todd Westerhoff * Mike LaBonte Synopsys: Rita Horner Kevin Li Teraspeed Consulting Group: Scott McMorrow Teraspeed Labs: * Bob Ross TI: Alfred Chong The meeting was led by Arpad Muranyi. -------------------------------------------------------------------------------- Opens: - Arpad noted that regularly scheduled meetings in December would occur on the 6th, 13th, 20th, and 27th. Arpad noted that we typically cancel the IBIS meeting that occurs in the last week of December. The meeting on December 27th is cancelled (official motion appears in the minutes below). ------------- Review of ARs: - Walter to create draft 5 of the file name relaxation proposal and send it to the ATM list. - Done. Posted to the work archives. - Bob M. to update BIRD 147.4 draft to include discussed changes to BCI_ID. - Done. New version sent out prior to this meeting. - Michael M. to incorporate discussed changes into the Format and Usage Out Clarifications BIRD draft. - Done. Posted to the work archives. - Michael M. to incorporate discussed changes into the Deterministic Noise Support BIRD draft. - Done. Posted to the work archives. -------------------------- Call for patent disclosure: - None. ------------------------- Review of Meeting Minutes: - Arpad: Does anyone have any comments or corrections? [none] - Radek: Motion to approve the minutes. - Bob M.: Second. - Arpad: Anyone opposed? [none] ------------- New Discussion: BIRD 147.4 draft 4, update to include changes to BCI_ID. - Bob M.: [sharing draft4 from his recent email] - Discussion: Bob M. reviewed changes to the language of the BCI_ID section. The section now refers to Section 3, item 3 of the spec. so it can inherit any changes from the file name rule relaxation proposal. The language explains how the BCI_ID value set by the EDA tool differs from a file name as defined in Section 3. The language clarifies that the complete filenames created by the model from the BCI_ID shall conform to Section 3, item 3. The language introduces the concept of a "namespace" as what the BCI_ID value set by the EDA tool defines. It introduces the terminology "namespace set" to illustrate the point that two models on the same channel (say Tx and Rx) may have different values set in their BCI_ID by the EDA tool. If the models' working directories are different, different BCI_ID values would be required to point each of them to a common location. The group reviewed the language regarding path-like quantities and file names, and editorial suggestions were discussed. Bob M. modified the proposal as the group reviewed it. The group discussed the paragraph that described what an end user might do if the EDA tool did not modify BCI_ID. It was viewed as unnecessarily describing a workaround for a tool that did not support BCI_ID. Radek also pointed out that it suggested the end user modify the .ami file, which we usually don't recommend. The paragraph was removed entirely. Bob R. asked that Bob M. issue the modified version as draft 5 of BIRD 147.4. Bob M. agreed to do so. - Bob M.: Motion to submit BIRD 147.4 to the Open Forum (draft 5). - Bob R.: Second. [no one opposed] Relaxation of IBIS filename restrictions: - Walter: [sharing draft 5 of the proposal] - I modified some language based on last meeting: - "../" is not permitted, except that it is permitted if generated by the EDA tool. - "/" is not permitted at the beginning of a file name, except that it is permitted if generated by the EDA tool. - We want to make sure model makers understand the format of what they make and put in IBIS files, and that the EDA tool can insert something different if necessary. - File names in IBIS files have one set of rules. - Exceptions are noted for the things the EDA tool may generate. - I changed the language in item 3 to say that a file name may NOT be a directory. - I added a sentence to say, "Directory (path) names follow the same rules as filenames, except that they shall be the name of a directory." - Previously we only had defined rules for file names. - Now we say paths follow the same rules as file names. - Bob M.: If you have a directory name, should we require it to end in a directory delimeter? - Walter: I thought that would make sense, too. - However, the AMI Supporting_Files parameter's description explicitly prohibits the delimeter from appearing at the end of the string. - Discussion: Bob R. expressed some concern over the language "except that it is permitted if generated by the EDA tool." He felt it might have hidden implications and might not be clear enough. Radek noted that the description was still somewhat incomplete, as there was no provision to provide a drive letter (e.g. C:) as might be seen in Windows. He suggested we might just refer to absolute or relative paths for the operating system, and that the only thing we might insist upon is that "/" is the directory separator and not "\". Walter modified the proposal as the group discussed and came to agreement on revised language (creating draft 6). - Bob R.: Motion to submit this to the IBIS Open forum as modified (draft 6). - Bob M.: Second. [no one opposed] Rx Deterministic Noise Support in AMI: - Michael M.: [sharing draft 2] - The only change made in this revision was to rename the parameters from Rx_Dn and Rx_Rn (Rx_Noise) to Rx_BoundedUniform and Rx_Gaussian (Rx_Noise). - Walter: My only problem with the new names is that they don't make it clear whether it's noise or jitter. - Perhaps we could add "Noise" to the end? - For jitter you often see BUJ (bounded uniform jitter). - Discussion: Michael M. proposed simply adding "Noise" at the ends of the new parameter names (Rx_BoundedUniformNoise, Rx_GaussianNoise). Bob R. noted his preference for using "_" between words, unless a convention already existed. Michael M. noted that he had reviewed other AMI parameter names, including the PAM4 parameters, and had only seen that "_" was always used to separate Tx and Rx from the following words. To be entirely consistent for the new parameter names, Michael asked if we should add "Unbounded" to the Gaussian noise parameter name (Rx_UnboundedGaussianNoise). Radek, Curtis and others noted that unbounded was implicit in Gaussian, and bounded was implicit in uniform, but no one objected to the proposed explicit names. Michael agreed to send out a draft 3 containing the name changes. Meeting on December 27th: - Arpad: Motion to cancel the meeting on December 27th. - Curtis: Second. [no one opposed] - Arpad: Thank you all for joining. AR: Bob M. to submit BIRD 147.4 draft 5 to the IBIS Open Forum as BIRD 147.4. AR: Walter to put draft 6 of the file name relaxation proposal into BIRD format and submit it to the IBIS Open Forum. AR: Michael M. to incorporate discussed changes into the Deterministic Noise Support BIRD draft and send out draft 3. ------------- Next meeting: 06 December 2016 12:00pm PT ------------- IBIS Interconnect SPICE Wish List: 1) Simulator directives