( ESNUG 480 Item 3 ) -------------------------------------------- [03/31/09]
Subject: AtopTech, Magma, Apache vs. the Synopsys/Cadence package deals
> The economy has soured significantly since DAC. Because it's a relatively
> small start-up, is Atoptech still a viable option to replace SoCE?
>
> - from http://www.deepchip.com/items/0479-07.html
From: Eric Thune <eric=user domain=atoptech not calm>
To: John Cooley <jcooley=user domain=zeroskew not calm>
Hi, John,
I read the post on DeepChip about AtopTech's viability and wanted to address
it. First, let me say that AtopTech is doing well - even in this terrible
business environment. We recently announced 250% growth over last year and
our first 8 figure deal - hardly a struggling company. We've been operating
off of revenues and have plenty of cash on hand to weather the storm.
The economy is actually turning out to be more of an opportunity for us than
a problem. Many of our customers are wondering which vendors are going to
be around while other vendors arrogantly think they should settle for a sub
par solution because they are in good shape and won't even do a benchmark!
We've been growing because we have a better product that addresses our
customers' pains in physical design. Our software is lot easier to use,
we are focused on customer support, and continuously developing with new
technology. Once a customer starts using our tool, they love what it can
do for them. Tell your readers to bring us a test case and we'll prove it.
You'll never hear us make a ridiculous claim such as "any chip in two days".
Being a profitable, sustainable company is crucial. We are building a
company for the long term, and plan to be around for quite some time.
- Eric Thune
AtopTech, Inc. Santa Clara, CA
---- ---- ---- ---- ---- ---- ----
From: John Cooley <jcooley=user domain=zeroskew not calm>
To: Eric Thune <eric=user domain=atoptech not calm>
> We recently announced 250% growth over last year and our first 8 figure
> deal; hardly a struggling company. We've been operating off of revenues
> and have plenty of cash on hand to weather the storm.
Hey, Eric,
I love analyzing the doublespeak you marketing guys like to put in your
letters. You forget that engineers aren't like the general public; we're
numerically literate. So using your two clues above:
A. an 8 figure deal
B. 250% growth over last year
That's $XX,XXX,XXX for a 8 figure deal (I'm assuming) and not $XXX,XXX.XX?
Yes or no, please. What's missing from this supposed ten of millions
deal is the time period. Is it for 1 year? For 5 years? My guess is
that what you're really saying is a $2 million, 5 year deal. Not bad, but
I do like the deceptive way you made it appear like a $10 million deal!
Looking at 250% growth, I figure that means AtopTech pulled in $570 K last
year if you're now at $2 million and you do the algerbra. Again, not bad
growth, but you're a small company and initial growth percentages are like
that. Again, my kudos on making a $2 million deal sound monumental.
> The economy is actually turning out to be more of an opportunity for us
> than a problem. Many customers are wondering which vendors are going to
> be around while other vendors arrogantly think you should settle for a
> sub par solution because they are in good shape and won't even do a
> benchmark!
Oh! Oh! You're talking about how Synopsys won't do benchmarks, right?
I agree that their policy here sucks because Synopsys is trying to bully
customers into taking a monster package deals vs. having tools compete
on their technical merrits; but I'm not sure who you're implicating as
who will die... Hmmm... My guess is it's Magma because of it's low
stock price and Cadence because Gary Smith publically hinted that they're
getting out of the IC CAD business when Fister & Friends mass resigned.
> We've been growing because we have a better product that addresses our
> customers' pains in physical design. Our software is lot easier to use,
> we are focused on customer support, and continuously developing new
> technology. Once a customer starts using our tool, they love what it
> can do for them. Tell your readers to bring us a test case and we'll
> prove it. You'll never hear us make a ridiculous claim such as "any
> chip in two days".
Oh! Oh! I know that one, too! Rajeev said that 2 day claim about Talus
or Titan; I can't remember which. I'm thinking it's Talus. Am I right?
On ease-of-use, again you're targetting the wrong guys (Magma). Magma
tools are notoriously easy to use while you have to have a PhD to run
Synopsys Astro or IC Compiler (with all its myriad of public and secret
switches to use.)
> Being a profitable, sustainable company is crucial. We are building a
> company for the long term, and plan to be around for quite some time.
OK, that one I'll give you. It refreshing to see a start-up looking to
stick around vs. its maximizing cash flow and hype just to get bought.
Also, I don't see Synopsys/Cadence/Mentor/Magma looking to buy any new
P&R start-ups right now. They're just too invested in trying to sell
what they already have to customers.
- John Cooley
DeepChip.com Holliston, MA
---- ---- ---- ---- ---- ---- ----
From: Eric Thune <eric=user domain=atoptech not calm>
To: John Cooley <jcooley=user domain=zeroskew not calm>
> That's $XX,XXX,XXX for a 8 figure deal I'm assuming and not $XXX,XXX.XX?
> ... My guess is what you're really saying is a $2 million, 5 year deal.
> Looking at 250% growth, I figure that means AtopTech pulled in $570 K last
> year if you're now at $2 million and you do the algebra. ... My kudos on
> making a $2 million deal sound monumental.
Hi, John,
Yes - decimal points don't count. While I can see how you derived these
numbers they are simply not correct. I'm bound legally so I cannot divulge
the terms of the deal we announced, but we have never and will never do
5 year deals. We have only seen deals like that when the big guys are
desperate to make their quarterly numbers.
> Oh! Oh! I know that one, too! Rajeev said that 2 day claim about Talus
> or Titan; I can't remember which. I'm thinking it's Talus. Am I right?
Yes, it was Talus. They stopped using it because it became a big joke with
their customers. AtopTech is more about under-hype and over-deliver.
> On ease-of-use, again you're targetting the wrong guys (Magma). Magma
> tools are notoriously easy to use while you have to have a PhD to run
> Synopsys Astro or IC Compiler (with all its myriad of public and secret
> switches to use.)
All one has to do is listen to Magma's financial conference calls about the
difficulties they had rolling out Talus and how their support costs were
out of control because of it. We have brought usability to a whole new
level. What is important is how long it takes an engineer to get to a good
result. If they need the vendor at their side to get these results I would
not call that easy to use.
> I don't see Synopsys/Cadence/Mentor/Magma looking to buy any new P&R
> start-ups right now. They're just too invested in trying to sell what
> they already have to customers.
The beauty of a profitable company is that you do not HAVE to be acquired
nor do you HAVE to raise money. If you are self sustaining, you can
continue to do business, grow, and build value. While the public market is
closed right now, it doesn't mean it won't re-open in the future.
- Eric Thune
AtopTech, Inc. Santa Clara, CA
---- ---- ---- ---- ---- ---- ----
From: John Cooley <jcooley=user domain=zeroskew not calm>
To: Eric Thune <eric=user domain=atoptech not calm>
> The beauty of a profitable company is that you do not HAVE to be acquired
> nor do you HAVE to raise money. If you are self sustaining, you can
> continue to do business, grow, and build value. While the public market
> is closed right now, it doesn't mean it won't re-open in the future.
Hi, Eric,
But that just sets you up to be squeesed out by Synopsys/Cadence every tool
needed for every step of RTL-to-GDSII package deals. All Synopsys/Cadence
has to do is roughly catch up to you and that will kill you in the long run!
For example, let's look at Apache. If Cadence got its act together and put
together a tiger team of ex-Simplex people that seriously focused some
killer R&D and killer Marketing in their space, I'd say Apache could be dead
within 12 to 18 months (or at worst, when the bulk of the Apache deals came
back up for renewal.) The SNPS/CDNS monopoly package deal is brutal to
compete against in the long run. One slip up and your little Atoptech start
up is history.
- John Cooley
DeepChip.com Holliston, MA
---- ---- ---- ---- ---- ---- ----
From: Eric Thune <eric=user domain=atoptech not calm>
To: John Cooley <jcooley=user domain=zeroskew not calm>
> For example, let's look at Apache. If Cadence got its act together and
> put together a tiger team of ex-Simplex people that seriously focused
> some killer R&D and killer Marketing in their space, I'd say Apache could
> be dead within 12 to 18 months (or at worst, when the bulk of the Apache
> deals came back up for renewal.) The SNPS/CDNS monopoly package deal is
> brutal to compete against in the long run. One slip up and your little
> Atoptech start up is history.
Hi, John,
The small companies are the engine of innovation in the EDA industry. If
industry stops supporting them, the pace of change slows dramatically. It
can be argued that the only reason the big guys improve their tools is
because of challenges from the smaller companies.
Do you honestly think Cadence or any other of the big guys haven't tried
to focus and beat Apache? Apache has done a wonderful job of not only
maintaining the gap between them and the big guys but expanding it. You
are right in that small companies have to stay on its game to survive, but
those who do reap the awards. The small guys that are successful are
technology driven, not marketing hype. Great examples are AtopTech, Apache,
ExtremeDA, and Oasys.
Bundled deals are bad for both the vendor and the customer. Vendors sell
the tools at super low prices so they can get more customer dollars. The
hidden costs of buying inferior tools comes to roost in the end. The
deals don't bring in enough money to justify the continued efforts in R&D
to make their tools leading edge and in the end the customer is the one
who is hurt since the tool either doesn't work or requires too much
support to make it work.
Why do leading edge semiconductor companies such as Broadcom choose the best
in class tools? Because it lets them execute better and get their chips out
the door faster. The so-called "Frankenstein flow" is a fallacy - industry
standard interfaces let users integrate their point tool flows seamlessly
and quickly - we do it every day. In most cases the best-in-class flows are
better integrated than the bundled flow.
- Eric Thune
AtopTech, Inc. Santa Clara, CA
---- ---- ---- ---- ---- ---- ----
From: John Cooley <jcooley=user domain=zeroskew not calm>
To: Eric Thune <eric=user domain=atoptech not calm>
> The small companies are the engine of innovation in the EDA industry.
> If industry stops supporting them, the pace of change slows dramatically.
> It can be argued that the only reason the big guys improve their tools is
> because of challenges from the smaller companies.
I'll grant you this one, Eric.
The only reason why the old Design Compiler ever improved was because two
synthesis start-ups, Ambit and Get2chip, had forced Synopsys R&D into
revamping DC. Otherwise the R&D folks at Synopsys would have rather worked
on something more glorious. (It's about status. Who wants to brag that
they redid someone else's 10 year old code vs. being able to boast that
you're the "father" of a new Synopsys tool?)
> Do you honestly think Cadence or any other of the big guys haven't tried
> to focus and beat Apache? Apache has done a wonderful job of not only
> maintaining the gap between them and the big guys but expanding it. You
> are right in that small companies have to stay on its game to survive, but
> those who do reap the awards. The small guys that are successful are
> technology driven, not marketing hype. Great examples are AtopTech,
> Apache, ExtremeDA, and Oasys.
I think Cadence never honestly tried to go after Apache. From what I
heard, Mike Fister was too preoccupied with advancing sales as much as
possible to prep for CDNS going private with Blackstone. It was all about
advancing sales NOW, not long term marketing nor any real focus on crushing
Apache. Hell, Fister even pulled Cadence out of DAC to save pennies!
In my opinion, the main reason why Apache is still around today is purely
because Cadence was (and still is) too caught up in internal politics and
palace intrigues that restrict its ability to seriously focus its Mktg,
Sales, and R&D on Apache. Cadence is a sleeping giant that, if it really
woke up and chose to, could wipe Apache off the map in one 12 to 18 month
concerted, hyper-focused attack.
This problem isn't just an Apache-Cadence thing. Synopsys could also kill
off Apache using the power of focused R&D, marketing and package deals.
Nor is it an Apache thing; frankly Synopsys or Cadence, if they so wanted,
could kill off Sequence or Prolific or Atrenta or ChipVision or Jasper or
Tuscany or SpringSoft or Nangate or Avery or Calypto or any other small
EDA start-up just by its focused R&D, killer marketing and package deals.
(BTW - There are some preemptive defenses that a small EDA start-up could
do against a concerted SNPS/CDNS attack, but most of the small EDA start-up
CEOs lack the special skills & insights required to enact such safeguards.)
> Bundled deals are bad for both the vendor and the customer. Vendors sell
> the tools at super low prices so they can get more customer dollars. The
> hidden costs of buying inferior tools comes to roost in the end. The
> deals don't bring in enough money to justify the continued efforts in R&D
> to make their tools leading edge and in the end the customer is the one
> who is hurt since the tool either doesn't work or requires too much
> support to make it work.
I have to respectfully disagree with you on this, Eric. The whole reason
why Cadence and Synopsys eagerly go after bundled deals is because it makes
a lot MORE money for Cadence and Synopsys.
1.) Their dominant tools (Virtuoso, PrimeTime) will always sell
no matter what. By tacking on the sub-par tools (Synopsys
"Orion", Cadence ETS) in a package deal, these sub-par tools
at least get SOME revenue coming in. Let me say that another
way: on their own the sub-par tools always lose against the
dominant player in a head-to-head competition, so it means
the sub-par tools won't ever bring in any cash; they'll
always be money sinks until they beat their rivals.
2.) Once a customer is locked in, you can control the customer.
For example, years ago Cadence made a big deal that it
kicked Synopsys out of Freescale. (It was all over the
press and Fister yarped about it in his financial calls.)
As long as the customer does OK, because he's locked in,
you get to ride his success, too.
There's a reason why Synopsys is scrambling to become everyone's "Primary
EDA Vendor"; it's the same reason why Microsoft has done so well. Having a
monopoly position means having a sweet, sweet life. Just ask Bill Gates.
But you are right on one key point: EDA customers dumb enough to drink the
SNPS/CDNS package deal Kool Aid often don't realize how duped they were.
For less money, they get a mess of mixed quality tools which SNPS/CDNS has
zero incentive to improve (aka Design Compiler before Ambit) plus users are
left stuck debugging their untried one-company-only flow. Good luck trying
to get Cadence ETS or Synopsys "Orion" to work; I'll stick to PrimeTime
and Virtuoso until something proven better comes along, thank you.
Also, actively volunteering your company to go under the sole control of a
monopoly (or duopoly) supplier is just plain bad business.
- John Cooley
DeepChip.com Holliston, MA
Join
Index
Next->Item
|
|
