Attendees:
Mark Hahn, Cadence (Chair)
Tom Dewey, Mentor
Ed Martinage, Cadence
Greg Schulte, Cadence
Vikas Sharma, IBM
New action items:
Who When What
---------- ------ --------
1. Mark 4/27 Send example of how the DAC demo subset
would be extended with typical values
2. Mark 4/27 Send modified version of processor design
to IBM
Open action items:
Who When What
---------- ------ --------
1. Mark 4/20 Close on DAC demo participation from
Synplicity, Altera
2. Bob 4/20 Determine whether an NDA is required for the
-> 4/27 library data
3. Mark 3/2 Add operating conditions to the taxonomy
-> 3/30
-> 4/13
-> 5/4
Closed action items:
Who When What
---------- ------ --------
1. Mark 4/20 Provide a list of BuildGates commands that
accept min/max values for best case, worst
case operating conditions
Next Meeting:
The next meeting will be a teleconference on 4/27
from 9-11 am (PDT).
Details:
1. Review action items
No progress on the library NDA question, because Bob is
on vacation.
Synplicity and Altera will decide whether to participate
by the end of April. Synopsys has indicated that they will
not participate directly, but they don't see any problem with
demonstrating translators between DCDL and the Synopsys design
constraints format, if the translator is developed by a company
licensed under the TAP-in program. Cadence is considering whether
there is enough time to do this.
2. Discuss DAC demo
There are still some problems in translating the IBM library to
TLF. These will be revisited when Bob returns.
IBM has a translator from Einstimer commands to PrimeTime commands,
but this is for a previous version of Einstimer and would not be
easy to to adapt for the demo.
Velocity can consume the Synopsys design constraints format, but
doesn't have plans to read the DAC DCDL subset. We talked about
whether it would make sense to include Velocity in the demo and
agreed that doing so doesn't demonstrate additional support for
DCDL.
We talked about typical values, as used in delay calculation for
SDF min:typ:max triplets. Mark proposed adding a -typical flag
and expanding the number of possible values from 4 to 6 for
environment conditions that may be used in SDF. Vikas had some
concerns about Einstimer's ability to support this, and Mark will
send an example to clarify.
Einstimer uses absolute value adjustments for multi-cycle path
constraints, rather than clock period multiples as in DCDL.
Vikas is working through issues in converting between the two.
Cadence has implemented most of the first draft of the DCDL subset
in both BuildGates and Pearl. Waiting for the TLF to test Pearl
on the demo design.
Thanks,
Mark
-- Mark Hahn phone: (408) 428-5399 Senior Architect fax: (408) 428-5959 Cadence Design Systems email: mhahn@cadence.com