( ESNUG 481 Item 2 ) -------------------------------------------- [04/23/09]
Subject: Kathryn on Jasper vs. Magellan, IFV, 0-In, RealIntent, IBM RuleBase
> I read that Jasper is now doing design services. Usually it's a bad sign
> when an EDA start-up switches to consulting. What have you heard?
>
> - from http://www.deepchip.com/items/0480-01.html
From: Kathryn Kranen <kathryn=user domain=jasper-da hot calm>
Hi John,
I think your reader confused the Design Activation Services we announced
earlier this year as a product change for Jasper.
Make no mistake. Jasper is a focused EDA software company.
The services we have ADDED to our standing product portfolio was for one of
our brand new products: ActiveDesign. It's a behavior-based RTL analysis
system that automatically generates waveforms (without a testbench or
simulator), and indexes design behaviors into a database for knowledge
sharing and reuse. Sound a bit perplexing? Yeah, we know, so we added a
service to offer users an easy "on-ramp".
Now regarding Jasper's overall financial health:
- JasperGold, our flagship formal property verification tool, enjoyed
hockey-stick growth this past year.
- JasperGold had a perfect 100% renewal record for 2008, meaning we
didn't lose a single customer, and no customer reduced their
time-based-license amount by a single dollar!
- We also landed several new top-tier accounts in 2008. Watch for
press releases on at least two of them in the next month or so.
The next topic to discuss is CASH. We manage our business to cash break
-even. We modified our SW business model over a year ago; thankfully
long before the current financial crisis. The math is pretty simple,
really. We know we have to generate enough sales to fund Jasper's operating
expenses, and that our business is fueled by applications engineering (AE)
bandwidth to proliferate deployment and help customers achieve their
targeted-ROI goals. So, we make sure the revenue at each account is in line
with the prorated "share" of our AE bandwidth required to support that
customer. For example, if we expect a customer to consume 10% of our total
AE resources, we'd better be sure the deal we negotiate will generate at
least 10% of the revenue required for us to break even. Likewise, our
products have to deliver enough differentiated value to justify our selling
price! We use a "targeted ROI" approach to show customers that Jasper is
able to reduce their labor costs and mitigate risks. This natural balancing
act keeps our R&D engine strong and our business healthy. (We've invested
over 100 man-years in research and development to date, by the way.)
Regarding operating CAPITAL, Jasper raised $7M in Series D venture financing
in November of 2008. Closing a round in this financial climate was no small
feat! Due to our strong fundamentals, we attracted a great new lead
investor, and all existing VCs participated in the round. This was no
"recapitalization" - i.e. no shareholders were wiped out. The common
shareholders (like me) are happy. The company now has the cash reserves to
control our own destiny.
The other topic in this financial discussion is managing EXPENSES. Despite
our blockbuster growth in 2008, Jasper (like any prudent company) set
conservative financial targets for 2009. As one of my VCs puts it, "Flat is
the new up!" Our 2009 operating plan allows us to be self-sufficient even
in the event we have a flat year (although I expect to grow). First we
reduced management bonuses and some commission plans. Late last year, we
laid off a few employees (one developer each in Brazil and Sweden and 4 US
employees). It was hard to see those good people go, but we had to do it.
For critical positions, like field AEs, we're "surgically" hiring again.
We've added another European AE and one more in India this year.
- Kathryn Kranen
Jasper Design Automation Mountain View, CA
---- ---- ---- ---- ---- ---- ----
From: John Cooley <jcooley=user domain=zeroskew hot calm>
To: Kathryn Kranen <kathryn=user domain=jasper-da hot calm>
Hi, Kathryn,
Thanks for following up. Sounds like things are going exceptionally well
for Jasper despite a hurting world economy. I just have a few questions
from your response:
You said 100% of your JasperGold customers renewed, yet JasperGold has
"hockey-stick growth this past year". For such growth you'd, of course,
need 100% renewal plus a lot more. How steep a stick are we talking here?
A light 5% or 10%? Or something meaty and respectible like 20% or 30%?
Earlier this month EDAC reported that CAE sales were down 17.6 percent
for Q3/08 and down 24.4 percent for Q4/08. It's ugly out there, yet
you're reporting healthy numbers? How can that be?
More specifically how are you competing against discount package deals for
VCS bundled with Magellan at Synopsys? Or Cadence NC-Sim bundled with
IFV? Or 0-in bundled with ModelSim at Mentor? Why aren't their discount
price-points killing you?
How are you competing against non-bundled formal players like RealIntent,
IBM RuleBase, Averant, OneSpin 360MV?
I'd love to hear your thoughts on each of these specific rivals, Kathryn.
- John Cooley
DeepChip.com Holliston, MA
---- ---- ---- ---- ---- ---- ----
> You said 100% of your JasperGold customers renewed, yet JasperGold has
> "hockey-stick growth this past year". For such growth you'd, of course,
> need 100% renewal plus a lot more. How steep a stick are we talking?
> A light 5% or 10%? Or something meaty and respectible like 20% or 30%?
From: Kathryn Kranen <kathryn=user domain=jasper-da hot calm>
Hi, John,
Your 5% to 30% guesses are way off base. Jasper GREW over 100% last year,
as in our 2008 annualized bookings were more than 200% of our 2007 bookings!
> Earlier this month EDAC reported that CAE sales were down 17.6 percent
> for Q3/08 and down 24.4 percent for Q4/08. It's ugly out there, yet
> you're reporting healthy numbers? How can that be?
There are two reasons:
Firstly, Jasper delivers net cost-savings. We are willing to substantiate
our positive ROI by collaborating with customers on "targeted ROI" goals,
setting a bounded timeframe for results, and supporting users to success.
(Hence the perfect renewal record.) By having the ROI discussion up-front
and often, we raise (customer) management visibility of the value our users
are achieving. As one customer said about Jasper, "Once you discover a big
vending machine that lets you put a quarter in and get a dollar out, you
use it over and over again." We make sure the results are good, and we
also make sure those results are visible!
Secondly, our technology is far superior to our competitors' offerings, and
the market is finally savvy enough to recognize the gap. Naturally, the
burden is on us to keep widening the gap. Over the past several years,
benchmarks indicate we are pulling away from the competitors. If the
customer is serious about formal, JasperGold wins. If they're just
dabbling, they take whatever is available in their big-vendor packaged deal;
and many of those have called us back later! Justifying our price is where
the "targeted ROI" discussion comes in.
> More specifically how are you competing against discount package deals for
> VCS bundled with Magellan at Synopsys? Or Cadence NC-Sim bundled with
> IFV? Or 0-in bundled with ModelSim at Mentor? Why aren't their discount
> price-points killing you?
All of our customers had 1, 2, or all 3 of the big vendor tools at the time
they engaged with Jasper. We haven't lost a technical engagement in a long
time, and we have competed head-to-head with all three of the big players.
We like it when customers have already experienced those other tools,
because then they see the superiority and value of JasperGold very quickly.
Our average evaluation cycle is 2 to 4 weeks. Sometimes the hardest part is
simply raising the customer's expectations of formal.
Jasper vs. Synopsys:
We hear from customers who chose Jasper over Magellan that JasperGold has a
much faster set-up time and less-bloated constraints. They say Magellan's
hybrid approach burdens the formal engines with baggage from the simulation
environment. One customer told us the Magellan environment added 18,000
flops to the proof complexity vs. 180 flops for the corresponding JasperGold
environment. That means the difference between proof convergence and mere
"bug-hunting". We won that account, and they later renewed in a big way.
We also hear that Magellan is a black box that blindfolds the user. An
evaluator told us that most of the time he couldn't tell if Magellan was
involved or if it was just VCS running -- in contrast to JaspserGold which
gives visibility into the proof process. That was an experienced formal guy
who spent months working with Synopsys on Magellan. My favorite line was
from a user who told my AE after the evaluation, "Basically, Jasper butt
-kicked Magellan in every department!" (His words, not mine.) The key
JasperGold differentiation vs. Magellan: capacity, visibility, and engines.
Magellan is positioned more as a step up from simulation than a serious
formal property verification solution.
Jasper vs. Cadence:
Cadence tried to lock in customers with multi-year deals, and until 2008
took the huge majority of revenue up-front. (There's a big blip in 2007 in
the EDAC MSS formal property verification category.)
For a while there, we saw Cadence trying hard at several accounts. These
days, they seem noticeably quiet. In one head-to-head competition last year
at a top tier account, Cadence actually abandoned the evaluation, telling
the customer "we're busy on other accounts right now".
I know Cadence disbanded the IFV corporate AE team. (We've seen lots of
their resumes.) The key technical differentiators for JasperGold over IFV
are Formal Scoreboard, "Visualize" automatic waveform extraction, and
engines. These are the key ingredients that let customers converge on
important spec-level proofs.
Cadence will steer the customer to low-complexity properties. Again, once
the customer becomes serious about formal, we are very likely to win.
Jasper vs. Mentor:
Mentor 0-In seems strong in clock-domain crossing checks (not our domain).
They have (or at least had) lots of Checkerware customers, but those
assertions were mainly used for simulation, and they are written in their
proprietary language. Differentiators are the same as above: Formal
Scoreboard, Visualize, capacity, engines.
0-In steers the customer to add lots of lower-level assertions, while Jasper
focuses on proving high-level properties that correlate to the spec (and
yield more tangible ROI). When we've competed head-to-head, we have won due
to end-to-end proofs on properties where 0-in did not converge.
> How are you competing against non-bundled formal players like RealIntent,
> IBM RuleBase, Averant, OneSpin 360MV?
We play in a different space of formal than Real Intent, so we never see
them in evaluations. A few of our users have Real Intent "Implied Intent"
and clock domain checks, so we seem to coexist.
IBM is not in the commercial EDA business. I've heard they have 3 (or 4?)
non-IBM "partner customers" who gained access to Rulebase years ago and are
supported directly by the handful of software developers. I don't believe
IBM is pursuing new customers. Most of their energy must go into supporting
their own server and processor development projects. JasperGold replaced
RuleBase at one very prestigious account, the only time we've encountered
RuleBase in head-to-head competition.
We haven't seen Averant in the market for years. I'm not aware of any
active Averant accounts. (Their website lists many "references" who no
longer use Averant. Some of the people listed left those companies
years ago.)
I know OneSpin is active in a few accounts in Europe that were users before
the spin-out from Infineon a few years ago. I'm not sure if they've closed
new customers since then. In 2008, we saw OneSpin in only one account (and
they had been disqualified technically before our evaluation started).
JasperGold won that engagement vs. two big-vendor incumbents! Meanwhile,
we've hired three AEs who were at OneSpin once upon a time.
- Kathryn Kranen
Jasper Design Automation Mountain View, CA
Join
Index
Next->Item
|
|
