( ESNUG 346 Item 5 ) --------------------------------------------- [3/16/00]
Subject: ( ESNUG 341 #1 ) PhysOpt, Gambit, Scheme, PDEF, & Avanti Saturn
> Perhaps my knowledge of the basis of PhysOpt placement is incorrect. Did
> this placer come from technology developed by Gambit? What kind of
> utilizations do you typically achieve, or, what kind of transistor
> densities do you usually achieve?
>
> I was also wondering what the total placement times were for the block(s)
> in question. I can see how the INITIAL placement for PhysOpt or a variety
> of other placers would be better than Avanti. Avanti's initial placement
> frequently is not that great. But in our evaluations (admittedly some
> time ago) the Avanti placer achieved the best final placement in the
> shortest amount of time.
>
> - Nick Summerville
> Ford Microelectronics Colorado Springs, CO
From: Bob Prevett <prevett@nvidia.com>
I'm not sure where Synopsys got their placer from. We achieved
utilizations ranging from approximately 85% to 95%, depending on the block.
We found that the timing and congestion driven placer for PhysOpt took
longer than the Avanti non-timing-driven placer. Sorry, I don't have data
comparing run times for PhysOpt against the Avanti timing-driven placer.
> Keep me anon. I work in the Methodology Development group at LSI Logic.
> I read Bob's review of Synopsys PhysOpt in ESNUG and am very interested in
> finding out more of your issues in working with an Avanti backend. Were
> there issues with getting the PDEF into Avanti? Was it translated via
> SCHEME? Did both cell placement and global routing information transfer
> from PhysOpt into Avanti? Did PhysOpt have a good understanding of
> blockage due to memories, power rails, etc? Also, he didn't mention
> whether his "old" flow utilized Avanti Saturn for physical re-synthesis.
> Did it?
>
> - [ An Engineer At LSI Logic ]
Well, I'm a design engineer, not a layout engineer, but I'll tell you what I
do know about this issue. We had to write Scheme scripts to be able to dump
out a version 3.0 PDEF file for PhysOpt to read in, since the Avanti PDEF
files were dumped out in version 2.0. An apps engineer from Synopsys helped
set this Scheme script up for us.
PhysOpt uses congestion analysis in addition to timing analysis to generate
the placement. However, global route information/estimates used by PhysOpt
for the congestion analysis are not dumped out. So, only placement
information, in the form of PDEF files, was transferred between the Avanti
and PhysOpt tool sets. All routing was done in Avanti Apollo.
Before running the PhysOpt placement, we used Avanti to do the floorplan for
the layout partition. This includes blockages due to memories and other
custom macrocells. Also, any blockages needed to relieve routing congestion
issues in known problem areas are also added at this floorplanning stage.
This floorplan with the various blockages is read into PhysOpt via the PDEF
file.
We did experiment with Avanti Saturn in the old flow; our results we not
that great at that time. However, that was almost a year ago, and it's
likely that Avanti has improved this tool since then.
- Bob Prevett
NVIDIA Santa Clara, CA
|
|
