Attendees:
Mark Hahn, Cadence (Chair)
Tom Dewey, Mentor Graphics
Ibna Faruque, Synopsys
Vassilios Gerousis, Motorola
Dan Moritz, LSI Logic
Greg Schulte, Ambit
Jim Swift, IBM
Andres Teene, LSI Logic (Symbios)
New action items:
Who When What
---------- ------ --------
1. Jim 10/13 Add operating conditions to the taxonomy
2. Mark 9/29 Schedule a teleconference with Steve Schulz,
Vassilios to discuss SLDL relationship
3. Mark 9/29 Prepare a more detailed breakdown of tasks
and schedule
Open action items:
Who When What
---------- ------ --------
1. Jin, Jim 8/25 Discuss PVT-dependent constraints and relationship
to conceptual model
-> Need to reassign
2. Jim 9/15 Investigate IBM strawman possibility
-> 9/29
3. Greg 9/15 Add description of min/max delay constraints,
-> 9/29 boundary conditions specified on internal pins
to taxonomy
Next Meeting:
The next meeting will be a teleconference on
Tuesday, 9/29/98, from 9-11 am PDT.
Details:
1. Discussion on Ambit acquisition
Cadence and Ambit are reviewing whether transferring the
copyright to OVI for the strawman still makes sense given
the merger. The current expectation is that the conclusion
will be that it does make sense.
2. Update on SLDL relationship
Mark talked about the presentation on DC-WG he did at the
SLDL workshop in Switzerland. SLDL is looking at defining
system level constraints in the near term, and is interested
in establishing a formal relationship with DC-WG to ensure
consistency or collaborate on this.
3. Review progress on action items
Jim talked about the status of IBM potentially donating
the Einstimer constraint language as a second strawman.
It is still under consideration within IBM, but most people
Jim has talked to so far are in favor of the donation, and
he is optimistic about it. Compared to the Ambit language,
the Einstimer language is fairly similar but may provide
some incremental value.
4. Continue discussion on operating conditions
We completed the review of Jim's write-up, with most
of the discussion centering on noise. Jim and Dan felt
that noise management will probably move from a library
guardbanding approach with appropriate assertions in the
library to a design constraints. This would allow
designers to manage the risk versus performance tradeoff
more directly. Mark argued that there would still have
to be assertions in the library to ensure that the cells
functioned properly, but that these assertions would be
directly tied to the cells' limits, rather than including
a guardband.
A possible noise constraint would be the number of simultaneously
switching outputs. We agreed to defer further discussion
on noise constraints because it isn't within the scope for
timing constraints.
5. Discuss boundary parasitics section in taxonomy
Steve Grouts was unable to attend and we were nearly out
of time so we postponed this item.
Thanks,
Mark
-- Mark Hahn phone: (408) 428-5399 Architect, Deep Submicron Business Unit fax: (408) 428-5959 Cadence Design Systems email: mhahn@cadence.com