( ELSE 06 Item 18 ) -------------------------------------------- [ 06/23/06 ]
Subject: Magma Mojave Quartz DRC/LVS
SOMETHING WENT WRONG -- No, this has nothing to do with the Synopsys-Magma
lawsuit. A little over 13 months ago (and right before DAC 2005) Mojave
splashed onto DRC/LVS scence with a very powerful ESNUG user benchmark.
ESNUG 445 #12: Mojave Quartz DRC Pummels Calibre in User Benchmark (I)
ESNUG 445 #13: Mojave Quartz DRC Pummels Calibre in User Benchmark (II)
ESNUG 445 #14: Mojave Quartz DRC Pummels Calibre in User Benchmark (III)
And then a little over 6 months ago I released the results of my annual
Synopsys Census and in it was the surprizing news about where DRC/LVS
users *expected* to be in the future:
Magma Mojave: ############################## 76%
Cadence PVS: ######## 21%
Synopsys Hercules: ##### 12%
Mentor Calibre: ### 8%
- from http://www.deepchip.com/items/snug05-15.html
But look at the user quotes below and you'll see they're only *interested*
in Mojave Quartz DRC/LVS -- there's no tsunami of ex-Calibre customers
yarping about how they had just switched over to Mojave. (In fact, I have
yet to find *one* user seriously talking about switching over.) Worst yet,
check out the Mentor Calibre section of this census, you'll find a backlash
of established Calibre users jealously defending Calibre. So far, Mojave's
"threat" to Calibre has been an awful lot like the massive build up of hype
that preceeded the "Star Wars: Episode I - The Phantom Menace" movie.
People were expecting the next Luke Skywalker or Hans Solo; instead they
got stuck with Jar Jar Binks. "Monsters out there, leaking in here. Weesa
all sinking and no power. Whena yousa thinking we are in trouble?" Mojave
had everything going "right" for them for a while, but then something went
wrong along the way.
The only Magma demo I saw was the Mojave physical verification demo. It
certainly looked new and exciting, and the DRC run times for large chips
were very impressive, however we currently don't have a requirement to
deal with such very large chips (as we are primarily involved in smaller
custom analog/mixed-signal designs.) But I will be looking forward to
seeing how the LVS and parasitic extraction end of things work out for
them in the future.
- [ An Anon Engineer ]
I saw a demo of the Mojave DRC/LVS tools. It was very impressive, but
there is no LVS tool yet. When they have an LVS tool, we will seriously
consider it. However, DRC without LVS doesn't really do me any good
because you can't really buy Calibre LVS standalone.
- [ An Anon Engineer ]
We were impressed with Mojave Quartz DRC. It has dramatic run-time
from parallelism across the network. It's still relatively new, and
I expect Magma to work toward having it handle more complex DRC rules
over time.
- [ An Anon Engineer ]
The thing that interested me about the Magma 2005 DAC demo that I saw
was the new Mojave physical design rule checker. Scalability is the
major strength of this checker over Calibre. As long as it lives up
to what it was advertised, e.g. 16 CPUs with 15x performance, this is
going to shake Calibre's market position. However, at the time of
demo, the LVS was still not available. It's hard to say whether it
is easy to debug, especially the power-ground-short type of problems.
From the company overall, the major thing that concerns me about Magma
is the support. This is probably common problem for the EDA companies
grown to this size. However, it is simply unbearable explaining some
complicate bugs on the phone to some inexperienced tech support. And
even before you open your mouth, they ask you to file a testcase in
their bug tracking system, the Molten. Besides, for sometime, they
used to be the performance leader. But I had seen some other tools
out-perform them now. e.g. the NanoRoute from Cadence is a faster and
better router, the GUI from Sierra and other emerging companies which
uses QT instead of X or Java as the engineer.
As of the strength of Magma over the other vendors, I think it's still
their unified database, which saves you all kinds of resources and
effort to convert back and forth between the databases used in
different tools. Besides, the open database which can be accessed via
their Mtcl is really powerful thing. Automatic clock tree synthesis
which also gives you the options to customize is also something
superior than Cadence or Synopsys.
Compared with the Synopsys and Cadence, Magma is still a young company,
although some aging signs have shown. I believe they move quicker than
these giants, too. But the complaint from users like me and many others
is that we always hope is that Magma had better fix their bugs before
putting in more new features.
- [ An Anon Engineer ]
Calibre currently dominates this market. It does no RCL extraction, nor
self inductance nor mutual inductance. Calibre ADP now supports in-die
variation. It handles metal fill capacitance as well, and is also
integrated to Advance MS so it can do SPF for digital portions of your
circuit and various formats for the analog portions. It integrates into
the Cadence schematic environment.
Magma says their new Mojave tool, which can run on multiple machines
(not just multiple processors) is 40X-50X faster than Calibre.
- John Weiland of Intrinsix Corp.
Mojave Quartz-RC:
Haven't tested it at smaller geom, but at 130/90 nm, it correlates very
well with Star-RCXT. Biggest issue is runtime. (3x too slow) Speed
up the core algorithms and also multi-thread it. We're Fusion users and
the "signoff in the loop" is an intriguing sell... we'd love to dump
Star, but performance just isn't there yet.
Mentor's Calibre DRC/LVS:
As history has proven, competition is usually a good thing for the
market, and possibly from multiple people. They now have competition
and are realizing it. Hopefully they can turn PowerPoint slides
into C code.
- [ An Anon Engineer ]
Index
Next->Item
|
|
