( DAC 02 Item 26 ) ---------------------------------------------- [ 9/10/02 ]
Subject: Sagantec, RubiCAD, Q Design/Marple, Numerical OPC, Avanti Taurus
WHAT'S IN A NAME?: I like how the olde Marple Technologies pulled a fast
one by renaming itself "Q Design Automation" and claimed it was a new EDA
vendor at this year's DAC. Oh, well. I've had girlfriends tell me that I
have the maturity of a 12 year old sometimes. I wonder if that's the same
thing and that makes me 12 now. Anyway, this Marple game isn't all that
important in the compaction/expansion business anyway. This market is
roughly 65% Sagantec, 30% RubiCAD, and 5% "other". Marple's under "other".
"Have used Sagantec. Doesn't recognize some analog requirements such
as guard rings. Seeing as we only migrate analog blocks, it did save
time overall."
- John Webster of Intel
"Not sure what Rubicad does, wasn't impressed at all."
- Jai Durgam of SiImage
"31.0 Compactors/Expanders
Rubicad, Sagantec and Q Design Automation make compactors/expanders,
which can be used to modify layouts for new layout rules, or take a
loose layout and make it as small as possible. Sagantec has a new
engine that supports P-cells from Cadence IC Craftsman, and does a
lot of things aimed at final tweaks for increased manufacturability
(via redundancy, using extra space to minimize crosstalk, etc.)
Rubicad emphasized how they correct timing and signal integrity
problems. You input GDSII, design rules, and a list of critical
signals and it adjusts widths and spacing.
The Q Design Automation tool bolts onto Cadence Virtuoso and shows up
as extra menu items. They have a separate standalone tool.
Antrim also sells a tool to help people port a design from one process
to another. It saves processed versus raw SPICE results and is
integrated closely with the Cadence environment."
- John Weiland of Intrinsix
"The compactor wars continue. A pseudo new player is the DAC 'newbie',
Q Design Automation. This is a very nice repackaging of the Marple
technologies company is a new shiny high profile image. Their 2-D
compactor actually works and is not terribly complicated to use (i.e.
you do not need more staffing resources to set up the tool and run it
than you would layout resources to move the design). 2-D compaction
has some good technology fits for IP companies. I think the OEM ASIC
clients still can't utilize it as they don't have new design targets to
migrate the layouts with and Q Design is not currently large enough to
run a major service biz.
Sagantec is still in the game. They seem to be doing well, living off
of just a few clients, but who isn't now a days. I think the biggest
limitation on Sagantec tools is capacity. If you architected something
scaling and automated DRC repair with compaction their solution will
work, however most designs are too big to fit into thier tools.
Rubicad has a new thrust on their compactor -- create new correct by
construction designs and you do not have deal with all the post layout
clean up and analysis. They have only been pitching this story for 10
years and between the giant crowds for the comic they had in the DAC
booth and everyone else now saying it is a good thing - people now
believe it. Biggest drawbacks on Rubicad owning this market on high
capacity automated cleanup:
1) The Rubicad tool is very versatile (i.e. complicated to setup)
for introductory users. The people running the tools have to
understand what the intent of design rules are to translate the
info into correctable instructions. Most users just know how
to read rules, not what they mean or why.
2) Rubicad needs to read and translate standard DRC control files
(Hercules or Calibre) rather than making their customers write
their own. Most places do not have staff to recode decks they
download from the fabs so they do not know if they are setting
the Rubicad tools right.
Rubicad solved their historical major problem. They actually have a
competent and personable sales force in now (rare for the compactor
industry) so they actually have the ability to identify a business
model and ROI basis for their tool use. Please contact Rubicad quick;
you never know how long competency in the EDA industry will last.
It appears that Prolific has won the cell generator race. I did not
see anybody else with real offerings in this area that have solutions
that work (i.e. 4 terminal devices, multiple isolation formats, etc.)
for the 90 nm and below processes. If Prolific has enough money to
keep the lights on between business development and the huge legal
bills for the people they are going after, Prolific has a good thing
going. The rest of the players in this niche are still targeting
Cadence SE style cells that do not work well with Magma/Avanti/Plato
style modern APR tools."
- Pallab Chatterjee of SiliconMap
"35.0 Mask Related Tools
When feature sizes really get small, you need to use phase shifters on
every other opening in the mask. The problem is that "every other
opening" might be clear when dealing with long parallel lines, but in
typical layouts where lines appear and disappear and have turns and
branches, there is no way to always do that. For example, a critical
net is not allowed to have a "T" in it (a "T" has three areas touching
it, so there's no way to put shifters on every other one). You can
have a "T" in your design, but the features can't be a minimum size.
The tool from Numeritech (Numerical Technologies) and also the Taurus
tool from Avanti will check for illegal geometries and add phase
shifters. The Numerical Technologies tool is also integrated into
Cadence Assura. I didn't get to check out Mentor for similar tools,
but I know they have a tool for phase shifters, too. All of these
tools are only dealing with polysilicon (gate level) at present. Since
poly is typically used only within cells, this means that you can run
their DRC on individual cells, and if the cells are clean it's unlikely
you'll have problems at chip level.
Dealing with metal levels is another story. It's not clear to me who
will solve this problem -- routers when they design the metal,
compactors/expanders like Rubicad or Sagantec after the routing is
done, or someone like Numeritech, Avanti or Mentor when you go to make
the mask. Numeritech is working with Cadence and their tool is now
integrated into the Cadence framework.
The Shearwater Group sounds like a brokerage house but actually sell
EDA. They include Pinebush Technologies (plotter software), K2
Technologies (GDSII viewing and tapeout automation) and Xyalis
(pattern fill and other tools).
Artwork Conversion Software, Inc. sells mask related software like fast
GDSII and MEBES viewers, plotting software, and IO planning software.
Ubitech sells some Samsung software for intelligent pattern fill that
tries to prevent timing problems. I don't know if this actually works
but it would sure be nice; metal fill occurs so late in the process
that having another iteration at that point is disastrous."
- John Weiland of Intrinsix
|
|
