( SNUG 05 Item 13 ) ---------------------------------------------- [12/20/05]
Subject: PhysOpt/Astro vs. Blast Fusion vs. SoC Encounter
THE MAGMA PROBLEM (Part I) -- My jaw dropped when I saw the first cut of this
survey data. Suddenly it became extremely clear why Synopsys and Cadence are
so deeply hostile to Magma. If *I'm* seeing this, *they're* seeing this.
"In the next 12 months as an old PhysOpt/Astro user, for our
production P&R flow my company will most likely (choose one)"
keep using old PhysOpt/Astro: ##################### 52%
upgrade to new IC Compiler: ######### 21%
switch to Magma: ####### 18%
switch to Cadence: #### 9%
"In the next 12 months as an old Magma Blast Fusion user, for
our production P&R flow my company will most likely (choose one)"
keep using old Blast Fusion: ########################## 65%
upgrade to new Cobra: ############# 32%
switch to Cadence: 0%
switch to Synopsys: # 3%
"In the next 12 months as an old Cadence SoC Encounter user, for
our production P&R flow my company will most likely (choose one)"
keep using old SoC Encounter: ################### 48%
upgrade to new Encounter GXL: ############## 35%
switch to Synopsys: ### 7%
switch to Magma: #### 11%
The Big Ugly here are those "switch to" totals:
switching away from Synopsys: ########### 27%
switching away from Magma: # 3%
switching away from Cadence: ####### 18%
I thought to myself: "Oh, man, that's just *nasty* for Synopsys & Cadence!".
Scratching my head about that 27% for Synopsys, I started looking at the
individual survey responses to see what was up. It was then I noticed "Hey,
that guy is a long time Magma user who's trying to pull off that he just
switched to Magma!". Not right nor fair. A second time around I recompiled
the data for these 3 questions moving the known Magma Fanatics out of the
"switch from Synopsys" bin and into the "keep Blast Fusion" bin (where they
rightfully belonged.) Here's the real stats:
"In the next 12 months as an old PhysOpt/Astro user, for our
production P&R flow my company will most likely (choose one)"
keep using old PhysOpt/Astro: ###################### 54%
upgrade to new IC Compiler: ########## 25%
switch to Magma: #### 11%
switch to Cadence: #### 10%
"In the next 12 months as an old Magma Blast Fusion user, for
our production P&R flow my company will most likely (choose one)"
keep using old Blast Fusion: ########################### 68%
upgrade to new Cobra: ############ 30%
switch to Cadence: 0%
switch to Synopsys: # 3%
"In the next 12 months as an old Cadence SoC Encounter user, for
our production P&R flow my company will most likely (choose one)"
keep using old SoC Encounter: ################### 48%
upgrade to new Encounter GXL: ############# 33%
switch to Synopsys: #### 9%
switch to Magma: #### 10%
I checked. There were no Synopsys Fanatics nor Cadence Fanatics that I had
to correct for.
Anyway, even with correct data, it still sucks to be Synopsys and Cadence.
switching away from Synopsys: ####### 21%
switching away from Magma: # 3%
switching away from Cadence: ####### 19%
I looked at the rest of the survey for some insights. I found:
"What 2 or 3 things about Synopsys would you DEFINITELY change?"
6. too much general arrogance: ############## 14%
7. sales & sales arrogance: ############# 13%
10. drop Magma lawsuit: ########## 10%
Here are a few of the "arrogance" complaints in context:
- "We are using DC. I must admit that our Synopsys sales person is great.
George Taglieri sends emails or calls to find out how things are going
without the sales pressure that I get from some of the others. Prices
are high, terms are poor, but usability and quality are OK. The
condescending attitude of 'our way is the only way' gives Synopsys a
big minus. Our foundry uses Astro. If there was an opportunity to
purchase a PD tool I would opt for Magma."
- "We have been dealing with many companies - EDA, IP, Services. No one
has a sales force that is as arrogant and unwilling to help than
Synopsys. They consistantly use the 'take it or leave it' approach."
- "DC is dominant. Solid tool. Aging a bit, but that's not a bad thing.
Local support is great! SolvNET is good, really good. SNUG excellent.
Sales is 'Helpful friends', ot that I always take my friend's advice.
Arrogance of some individuals must go!"
- "DC is the industry standard, the 900 lb gorilla. All others are
wannabes. This is good for Synopsys because other than PrimeTime,
they don't have great share in the backend flow. Our backend support
is for Cadence and Magma through OpenChoice and MagmaTies. Synopsys
is a dying dinosaur. A dinosaur's brain is a long way from the rest
of the body. It takes a long time for the brain to realize that it's
dying. Synopsys is the least customer-friendly, most arrogant EDA
company. Simply put, they are jerks and proud of it. See prior
dying dinosaur comment..."
Here are a few of the "drop Magma lawsuit" comments in context:
- "Use Design Compiler. Don't use the others. Sales is 'Helpful
business acqaintances.' Stop the lawsuit against Magma. Keep our
application engineer where he is! He is really pretty good!"
- "We initially used Magma Blast Create for our VHDL synthesis as we were
using Blast Fusion for the place and route. Unfortunately we found it
had so many problems we dumped it completely in favour of Synopsys DC.
Synopsys local support is pretty good.
SolvNET web support: Useful
SolvNET article quality: Generally good
try SolvNET before phoning : Y
I've not attended Synopsys Users Group as yet but it must be better
than the user group Magma attempts. Rate our Synopsys sales as
'Helpful business acqaintances'. Synopsys should concentrate on
their customers and not taking rivals to court."
- "DC is probably superior. Also benefits greatly from legacy; aka this
is not an easy tool to switch, and there's a lot of user knowledge.
Dump the lawsuit with Magma, make up with Rajeev, and apologize to my
customers, and tell them I'm going to win their accounts based on
technical merits, and not FUD."
Here are a few of those 11% of Synopsys users who are switching to Magma.
- "We exclusively use DC - as it is very well understood by all our
designers, has good run times, and can handle our design size. Our
Synopsys local support is truly awesome! They are very pro-active in
identifying any application questions or issues. Some good things get
presented at SNUG, but it's a high tax on time. We will continue to
use the PhysOpt/Astro for the majority of designs. The initial pricing
on IC Compiler is steep, and will likely put the decision to adopt it
in our Finance Department and out of our Engineering Department.
PhysOpt/Astro pricing issues have also forced us to adopt Magma."
- "I'm based in Minneapolis but work corporate-wide between my company and
Synopsys. My local support is OK, but they can't address most of my
issues, which require deeper development support. Sales used to be
'Helpful business acquaintances' or even friends, but since our contract
has gotten 'old' it's degenerated to 'Department store cashiers.' I've
worked closely with Synopsys for 15 years. I feel Aart is one of the
best CEOs in the industry. Intelligent, and a high degree of integrity.
The company has gotten too big & lethargic. It's become an amalgamation
like Cadence, making it difficult to stay energized. I've gone from
being in awe of them, to seeing that they are on-par with the developers
in my company, to almost dreading having to work with them when problems
come up. They are not nimble or hungry like they used to be. Now it
seems like getting anything done is a major pain in the butt. Switch
to Magma."
- "I am using Blast Create for synthesis, but I liked Design Compiler a
lot better. Probably need some time to adjust. SolvNET is the best
self-help among all EDA tools. I've tried Cadence and Magma, and they
are not as good as SolvNET. The SNUG is a lot more organized than
MUSIC. This is the first year that I got to go to MUSIC. One of the
presenter was a no-show, and the technical content was not as good as
SNUG. As a company I rank Synopsys 'Far Better', but the company
decided to go to Magma for cost-saving."
- "We believe DC is still the best. We use DC. Local Support is so-so
to mixed. SolvNET is useful, article quality generally good. I rate
SNUG excellent. Sales is 'Helpful friends'. PhysOpt/Astro now.
X - switch to Magma. Find or buy a better physical tool."
Pricing, perceived weak technology, and a feeling that Synopsys had become
too bureaucratic were the main reasons for users switching to Magma.
Overall these results are bad news for Synopsys in the backend market. Yes,
from the stats Synopsys is going to retain 79% of its customers. Cadence
keeps 81%. But Magma keeps 97%! Not good. Vegas thrives on a 2% house
advantage; Synopsys is screwed with an 18% margin to Magma. Arrogance was
understandable when Synopsys completely owned RTL synthesis, but it bites
them in the ass in a 1/3, 1/3, 1/3 P&R market. ("Look, Ma! No monopoly!")
And even though Magma started the lawsuit, to the customers Synopsys appears
like it's desperately using the courts to kill Magma because it can't
compete with Magma technologically. (How the heck did *that* happen!?)
I called up Gary Smith at Gartner/Dataquest to see if his data matched mine.
"My data shows all same things you see, John. You're only duplicating my
work," Gary joked.
"We see Magma picking up more users seats in the customer accounts. Cadence
and Synopsys aren't being completely replaced, though, they're just losing
individual seats to Rajeev," reported Gary. "I've seen cases where they've
gone 90% Magma seats with the remaining 10% stay Synopsys seats."
"On that small 3% 'switch away' rate you found for Magma, the engineers who
use Magma tools love them. The only reason why they'd stop using Magma is
because a Pointy Haired Boss forced them to switch back," said Gary. "You
can layout a design in considerably less time using Magma compared to
Synopsys or Cadence. It's a quicker design flow."
"Magma's also easier to use, which means you can have more junior engineers
doing your layout," he added. "But engineering managers aren't completely
kicking out those slower, more expensive, harder to use Synopsys seats yet.
The Synopsys tools are the only ones that can close timing on the most
difficult designs. For the bulk of 65 nm and below, the battle is between
Magma and Synopsys."
"Synopsys had a major sales force arrogance problem back in the mid-90's,"
added Gary. "When Aart came in, he made a major effort to eliminate the
arrogance. It's a shame to see it showing back up in your survey."
"Funny thing how Magma sues Synopsys and Synopsys comes out as the Bad Guy,"
concluded Gary. "No one ever wins an EDA lawsuit except the lawyers."
Index
Next->Item
|
|
