( ESNUG 496 Item 2 ) -------------------------------------------- [12/15/11]
Subject: 62% of EDA users fear Synopsys STA/PNR/Synth/SPICE/DRC dominance
KEEPING YOU LOCKED IN: With the Extreme-DA and Tekton acquisitions, many
users see this as Aart using his wallet to keep his monopoly on the static
timing analysis market -- thus forcing ALL chip designers worldwide to
have to at least buy *one* tool from him. (Smart strategy that!)
And Aart seems to be trying to do something similar by buying market
dominance in the digital P&R, RTL synthesis, and SPICE -- assuming the
Justice Department approves this LAVA-SNPS merger.
"I'd like to hear what you think of this SNPS-LAVA merger. Is it
good news for you? Bad news? Neutral? Why do you think this?"
Here's the tool-talk breakout of how the users commented to the "Why?".
This SNPS-LAVA merger is BAD overall because:
gives SNPS monopoly in STA : ######## 16%
gives SNPS dominance in digital P&R : ########## 19%
gives SNPS dominance in synthesis : ## 4%
gives SNPS strength in LVS/DRC : #### 7%
gives SNPS dominance in SPICE : ##### 9%
we depend on LAVA database : #### 7%
"I'd like to hear what you think of this SNPS-LAVA merger. Is it
good news for you? Bad news? Neutral? Why do you think this?"
---- ---- ---- ---- ---- ---- ----
I suspect Magma's new STA tool Tekton could be a major reason. PT is
"running out of gas" with the memory and runtimes issues. Since timing
closure is the most compute intensive portion of the entire chip design
flow, this is a major issue with bigger designs.
I hear rumors that Qualcomm replaced PrimeTime with Extreme-DA and then
Magma Tekton. With this LAVA aquisition and it's recent purchase of
Extreme-DA, SNPS has gobbled up the two most promising next generation
STA tools. I'm sure they hope to migrate their customers to the next
generation STA tool, probably Tekton based, and keep their de-facto
monopoly on signoff STA.
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
This isn't good news for us. With Goldtime and now Tekton in the hands
of Synopsys, no competition for STA doesn't appear very healthy. We've
learned the importance of having viable alternatives -- it just got a
lot harder in the area of STA. What's to stop Synopsys from tripling
their price next time our contract is up? (A different vendor in another
niche did exactly that to us this year, and we had to suffer the pain of
switching to an alternative.)
It'd be very disruptive to try and make such a switch on short notice
in the STA area. Cadence ETS was (at least last year) is even slower
than PT, had trouble with some of our constraints, and isn't script
compatible with PT.
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
This merger is good for Synopsys but bad for Magma employees and users
of Magma tools, especially those who really like Magma tools. I think
Synopsys acquired Magma just to kill the company, just like they did to
Extreme-DA's GoldTime. It's to preserve their STA monopoly.
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
BAD NEWS: I think SNPS is killing the EDA market. Two acquisitions in
few months: maybe Aart is scared about how good and how rapidly
competition can evolve while SNPS cannot. My feeling is they buy the
competitors to kill them: for instance, of the three STA tools they
have bought, which one will remain? The same for the SPICE: Saber,
Hspice, Hsim, Finesim. And what about ICC vs Talus? Etc. Etc.
I'm curious to see what they keep and what they discharge as there is
too much technology overlap.
Few years ago, if I'm not wrong, I remember Aart said the "time of
acquisitions" was over, but apparently he was just joking to not scare
the others BIG EDA vendors. ;-)
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
This is not good news for EDA customers, who now have one less credible
source of mainstream design tools.
I would describe Magma as an innovator that displaces tired EDA technology
with technology designed for the latest design challenges. For example,
you mention SPICE. I would add Tekton (vs. PrimeTime). Magma is doing
the job that Synopsys should have been doing, but hasn't.
Given the choice between investing in its own products or buying somebody
else's, Synopsys did the latter. It's an admission of innovative failure
and of a failure to respond adequately to customers' needs. Once Magma
is subsumed by Synopsys, it's likely that SNPS innovation will grind to
a halt, too.
The Synopsys move also reinforces the recent responses to De Geus' claim
that a total Synopsys toolset is more valuable than best-in-class EDA
tools. Aart has just bought some best-in-class Magma tools.
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
I think the one bit of good news that may come out of this is if Synopsys
can make use of the Tekton and, to a lesser extent, the QCP engines. PT
has become a dinosaur and is woefully inadequate when used to verify
designs in advanced technologies which require multiple signoff scenarios.
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
Other than for marketshare, I don't see why Aart bought Magma. The two
have enormous product overlap.
Synopsys Magma
P&R IC Compiler X - Talus Vortex
Static Timing PrimeTime X - Tekton
Logic Synthesis DC Ultra X - Talus Design/RTL
Full Custom X - Custom Designer Titan MSDP, ADX, etc.
DRC/LVS IC Validator X - Talus qDRC
DRC/LVS X - Hercules Quartz DRC/LVS
Extraction X - Star-RC QCP, QuickCAP
Lib Char X - Liberty NCX SiliconSmart
Yield/Litho Yield Explorer, etc. X - Camelot, etc.
SPICE HSPICE/HSIM FineSim
Fast SPICE CustomSim FineSim Pro
X - denotes tool Aart will most likely kill off. Note no X's for any
SPICE tool. SNPS doesn't End-of-Life them.
Anyway, this merger is good for Synopsys; not good for EDA users.
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
Bad. Backend PNR will be 80% Synopsys, 15% Cadence, 5% Mentor.
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
PNR: Talus will die. It was decreasingly competitive in the last few
years anyway. Synopsys just bought its customer base.
FineSim is superior and will replace Synopsys SPICE in-house.
Tekton will be integrated into PrimeTime.
Overall very bad for chip design companys. One less EDA supplier to
leverage on deal negotiations.
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
Bad.
We have a policy to have two P&R solutions and used ICC and Talus until
today. There are benefits because of this competition such as contract
negotiation, better support and so on.
And also we can reduce the risk to rely on one company.
Now we have to ramp-up another P&R tool and remove either ICC or Talus.
I realize now Synopsys is VERY kind to their rival Cadence.
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
In a few short months Synopsys has made 2 major acquisitions which have
almost 100% overlap with their current existing product lines in static
timing analysis and digital PNR. It was interesting hearing Aart try
to dodge that question during his Earnings conference call.
If anyone wants to believe any tool not named FineSim or SiliconSmart
will continue on, then I've got a bridge in NYC that I would like to
sell you. Aart's approach is to: sue them, buy them, slowly kill their
product off and attempt to transition their customers onto SNPS
mainstream products, by telling the industry you will "merge their
technology into the Synopsys best-in-class products".
- [ An Anon Engineer ]
---- ---- ---- ---- ---- ---- ----
I think this is very bad news for the EDA industry. I loved SiliconSmart
and its support. Their sincerity came through. I'm worried Fart De Geus
may just trash most of the good stuff.
- [ An Anon Engineer ]
Join
Index
Next->Item
|
|
