益華電腦 (CDNS) 2007 Q3 法說會逐字稿

Good afternoon.

My name is Katina, and I will be your conference operator today.

At this time I would like to welcome everyone to the Cadence third quarter 2007 earnings conference call.

All lines have been placed on mute to prevent any background noise.

After the speakers' remarks, there will be a question-and-answer session.

(OPERATOR INSTRUCTIONS)

I would now like to turn the call over to Jennifer Jordan, Corporate Vice President of Investor Relations, for Cadence Design Systems.

Thank you.

Ms.

Jordan, you may begin your conference.

Thank you, Katina and welcome to our earnings conference call for the third quarter of 2007.

The webcast of this call can be accessed through our website, www.cadence.com, and will be archived for one week.

With me today are Mike Fister, President and CEO, and Bill Porter, Executive Vice President and CFO.

Please note that today's discussion will contain forward-looking statements and that our actual results may differ materially from those expectations.

For information on the factors that could cause a difference in our results, please refer to our 10K for the period ended December 30th, 2006 and our 10Q for the period ended June 30th, 2007.

In addition to the financial results prepared in accordance with generally accepted accounting principals or GAAP, we also present certain nonGAAP financial measures today.

Cadence Management believes that in addition to using GAAP results in evaluating our business, it can also be useful to measure results using certain nonGAAP financial measures.

Please refer to our earnings press release for a discussion of nonGAAP measures and to both our earnings press release and our website for reconciliations of GAAP and nonGAAP financial measures used in today's discussion.

Mr.

Fister.

Once again, Cadence delivered excellent results in the third quarter.

Revenue was $401 million.

Operating margin was 30% and cash flow was $89 million.

Cadence continues to focus on aligning our Technology Solutions with our customer needs.

As a result, we deliver and received more value and have strengthened our market segment share position.

The fact that Cadence offers the industries most complete and robust technology with segmentation that is tailored customer demand is contributing to our success.

Customers are increasingly adopting our innovative and end solutions and leveraging our unique software distribution mechanisms.

For example, we introduced two years ago a form of technology delivery called an EDA card which provide, excuse me, proving increasingly popular with customers.

EDA card is a form of contract that enables customers to draw down product licenses from a self serve real time system so that they're able to better match their project license demand with purchase license capacity.

The infrastructure for EDA card is proprietary and provides us and our customers with an excellent visibility into their project needs and utilization.

Customers also have the ability to draw down licenses of differing time durations each priced accordingly until they exhaust their purchase capacity.

Our initial experience has been that customers draw down the licenses faster with EDA cards than with traditional contract types.

Just as this is unique delivery system helps our customers manage their most complex business productivity challenges, our innovative Technology Solutions help customers manage their most complex design challenges.

Within the core businesses, as an example, we examined our Digital Implementation Platform end-to-end, and delivered the Cadence Low-Power Solution, a way of preserving power intent throughout the digital flow that is unmatched in the industry.

This differentiation is fueling sales in Encounter platform both competitive wins and expansion into existing accounts.

And we see Electronics America using Encounter platform to implement the ARM11 MPCore, is one of the world's highest performance low power processors.

Utilizing Cadence's Low-Power Solutions including RTL Compiler and Encounter GXL.

NEC achieved overall performance in timing performance and lower power.

Advanced technologies in Encounter GXL such as statistical static timing analysis and optimization allow the customers to accurately account for the effects of processor ability in leading edge 45 nanometer designs.

When (inaudible) adopted this solution for SoC designs including Encounter's timing system GXL with statistical static timing analysis.

During the third quarter, design with Physical was also added to Logic Design Team Solution.

Design with Physical gives Logic Design's access to the interconnect information in their logic synthesis environment, dramatically improving physical predictability and increasing the quality and accuracy of high speed, low power designs.

In customized C we experienced continuing momentum for the high-end GXL versions of Virtuoso 6.1.

We saw our first tape-outs using the new platform.

At major North American Communications Company included a 45 nanometer tape-out using the new platform is moving their custom analog design flow to Virtuoso 6.1.

The speed, capacity and versatility of Virtuoso Multi-Mode Simulation in handling the verification of complex designs drove several competitive wins during the period.

For example Realtek, a leading Taiwanese designer of embedded systems chose Virtuoso Multi-Mode Simulation as its preferred solution and Renesas also adopted Cadence's wide variety of mixed signal design products for Multi-Level Simulation to layout as well as RF and digital SiP.

At CDNLive!

Silicon Valley our team provided additional insight into our strategy for the manufacture ability adjacency.

The GXL versions of Encounter and Virtuoso integrate Chip Optimizer and CMP Predictor.

Our collaborative with, excuse me, our collaboration with and subsequent acquisition of Clear Shape extends Cadence's capability with design-side lithography analysis and Optimization Solutions.

Together, these capabilities, which are already certified by TSMC at 65 and 45 nanometers provide a design and implementation flow for what you design is what you get in terms of manufacture ability at advanced process nodes.

And in manufacturing sign-off the acquisition of Invarium gives us leading edge optical proximity correction for OPC.

These kinds of technologies to address the next generation lithography challenges of 45 and 32 nanometers.

During the quarter we added over 100 million gates of Palladium upgrades within it's existing accounts and booked Palladium sales in a number of new accounts requiring system level verification.

LSI selected Cadence Palladium hardware and their networking and storage groups used Palladium to help with verification challenges including the verification of software and hardware interaction in advanced high speed interfaces.

Additionally leading customers are using our Incisive Software Extensions or ISX with Palladium for hardware, software, co-verification and validation.

And as projected, we released the SoC functional verification kit which provides a pre-built verification environment extending from the black to the system level targeted especially for today's future edge wireless and consumer designs.

It's this type of leadership that has compelled many of the largest semiconductor companies in the world to deepen their relationship with us, making Cadence their primary design solutions partner.

After a full and exhaustive evaluation process, NXP, named Cadence its primary strategic partner for EDA technology.

NXP has endorsed us as the clear Low-Power Chip Design Solutions leader and the only Company with the depth and breadth to deliver for NXP, the fully integrated front-to-back analog and digital solutions they require.

And like us, NXP shares an acute interest in developing solutions at the system level.

I'm excited we'll also be working and investing together to develop new technologies for the next generation of electronic system level designs.

Our consistent performance as a result of strong customer response to our technology and to our strategic road map.

And now I'll let Bill speak to the financials.

Thanks Mike.

The results for the Company's key operating metrics for Q3 were total revenue up 9% year-over-year.

NonGAAP operating margin of 30%, improving 200 basis points from Q3 of 2006.

And operating cash flow of $89 million.

GAAP earnings per share for Q3 were $0.24, compared to $0.14 in Q3 of 2006.

NonGAAP earnings per share for Q3 were $0.33, compared to $0.26 in Q3 of 2006, up 27% year-over-year.

Total revenue for the third quarter was $401 million, compared to $366 million in Q3 of 2006.

Product revenue was $274 million, maintenance revenue was $96 million, and services revenue was $31 million.

Revenue mix by geography in Q3 was 41% for North America, 25% for Europe, 22% for Japan and 12% for Asia.

One customer accounted for 14% of revenue.

Estimated contract life on a dollar weighted average basis approached four years, driven primarily by one large customer.

In the quarter, approximately 50% of our product business was represented by ratable licenses.

This was lower than our historical rate because of a higher mix of term contracts.

An important factor which contributed to the higher level of term business in Q3 was the EDA cards that Mike talked about.

Cadence offers both term and subscription EDA card contracts.

The volume of EDA card contracts has grown rapidly over the past several quarters and has reached $1 billion cumulatively.

Based on our recent experience, I expect the ratable mix for Q4 to be in the low 50s and for the year in the low 60s.

Total cost and expenses on a nonGAAP basis for Q3 were $281 million, compared to $280 million in Q2.

Our nonGAAP operating margin in Q3 was 30%, compared to 28% in Q3 of 2006.

We are on target to achieve a 30% operating margin for the full year 2007.

Quarter end headcount was approximately 5300.

Total DSOs in Q3 were 112 days, compared to 93 days in Q3 of 2006.

The higher DSOs are directly attributable to more term business where customers pay over time.

We expect DSOs to be at mid-90s at year end.

The quality of our receivables remained high with receivables 90 days past due less than 1%.

Operating cash flow for Q3 was $89 million, compared to $60 million in the third quarter of 2006.

For 2007, we expect to generate operating cash flow of approximately $450 million.

Capital expenditures in Q3 were $19 million.

For 2007, we expect normal capital expenditures of about $75 million, plus $21 million for work on the new engineering building.

Cadence repurchased 12 million shares of common stock at a cost of $251 million in Q3.

Approximately $155 million remains under our current stock repurchase authorization.

Cash and cash equivalents were $936 million at quarter end.

Now I'll turn to our outlook for Q4 and the year 2007.

For Q4 we expect revenue to be in the range of $465 million to $475 million.

GAAP EPS should be in the range of $0.34 to $0.36 and nonGAAP EPS in the range of $0.45 to $0.47.

For the year 2007, we expect revenue to be in the range of $1.622 billion to $1.632 billion.

GAAP EPS should be in the range of $0.94 to $0.96.

And nonGAAP EPS in the range of $1.34 to $1.36.

Other income and expense for 2007 should be in the range of $40 million to $45 million.

I expect that we will attain our primary operating metrics for the year of growing the top line, improving operating margin and increasing cash flow.

Looking beyond this year, we should be able to grow the business profitably while expanding operating margins.

Operator, we'll now take questions.

(OPERATOR INSTRUCTIONS) We'll pause for just a moment to compile the Q&A roster.

One moment we are still compiling the roster.

We're still compiling the roster.

Your first question comes from Jay Vleeschhouwer with Merrill Lynch.

Thanks, good afternoon.

Bill, I'd like to ask about a couple of the assumptions behind your fourth quarter revenue outlook and your license outlook.

First, are you expecting as we've now seen in the last five quarters, a double-digit percentage revenue customer?

And secondly for the year do you expect, as you've been expecting before, to be able to grow backlog by the end of the year?

Sure, Jay.

Let me cover your second question first.

I do expect that we will grow our backlog for the year.

In terms of large customers, I think we have seen this year that we have had the ability to close large contacts or contracts as part of our business and our pipeline.

I think we've done a good job of that.

It's the ability to have strong technology and to manage our pipeline and I do think that we have the ability to do that going forward, although I don't want to project how those things close in a particular quarter.

All right.

With respect to some of the customers you named, NXP was of course in the pipeline for the year.

What's different this time with the renewal?

You were their primary EDA vendor already, so what's new about it?

And you also mentioned NEC America.

Historically you had been largely exposed to NEC through NEC Europe, so are you now suggesting that you're broadening your exposure and share at NEC more globally?

Yes, absolutely.

It's a pleasure to have been a long-standing part of the supply dynamic to companies like NXP and I think from the press release, Jay, and from what you can see we've definitely expanded our relationship front-to-back including up to stack in the ESL.

That's a tremendous vote of confidence in the product direction that we're on that competed for lots of little points of light with that-- with our technologies.

And the -- I think that with other companies, you see the concept of run rate it's not just a figment of our imagination, it's a reality.

The companies like NEC see the value of Low-Power and the end-to-end dynamic there, and to play it more broadly, those are market segment share gains, as I said in my comments.

So I think a lot of the tests of the metal is not just being a part of the supply dynamic, but more and more an "integral partner", with these companies and increasing our ability to help them across all their -- all the pieces of the solution, [G eccentrically] and technologically.

Mike in your remarks you referred to CDNLive!

and some of the new technologies you are working on there.

But clearly you've acknowledged yourselves that you're behind in DFM for example, it's still a small percentage of your business.

When do some of the things that you showed last month, based on internal technologies, Clear Shape and all the rest, really begin to become more material?

I mean at what point do you run the risk, so to say, of just being too far behind your peers in this growing category?

Yes it's a good question.

As I mentioned in the first of the kind of subthoughts, the-- a lot of our DSM is correct by construction are integrated into the design so that you do it right, as opposed to trying to add it in later, test it in, OPC it, whatever the heck people try to do.

And that's a-- to me a historical artifact of the approach.

And already in the GXL versions of Encounter and Virtuoso we have those technologies implicit in there.

For the other elements of modeling, which is the ability to use the same model in the design and in the manufactured process, that's where our collaboration and now acquisition of Clear Shape, has amounted to impact for over a year and the Invarium stuff is just cool technology coming along, it's absolutely in the same dynamic, modeled abstraction.

And consistently throughout next year we'll see those things kick in monetarily.

For the most part now we're in trial with some very specific customers, who have very specific problems and we referred to in the last call about some memory technology-- some memory companies.

Those are guys that are out already at 45 nanometers.

To go back and do a bunch of DFM stuff at 130, that track's already been laid.

So a lot of that's a forward-looking thing as people move aggressively into 65 nanometer production or certainly 45 nanometer tape-out and production.

I like our chances, I think it's a correct by construction world.

You have to have the implementation flow to be able to marry with the manufacturing sign-off and the manufacture ability, and so far I'm happy with our progress with the customers that we're in trial with.

Yes Jay, I'd just like to add a little color to that.

With the correct by construction particularly in digital in GXL and in custom in Virtuoso in GXL, you will see additional strength in those two product lines as the result of our success with manufacture ability, but we're not going to be able to pull that out.

So at the high end of our line, and I think we've talked a little about this in the past, that's going to be one of the ways that you're going to see us out-perform traditional growth in digital and in custom is because of the strength we're going to bring from manufacture ability up to the high-end of our design capability, in addition to just taking some market segment share there.

All right, finally the duration of the-- on average went up in the quarter, which you attributed to one deal.

Cynopsis saw a long deal with Intel of course, five years was unusually long.

And you've otherwise been almost completely at three years, up until now.

So what was the requirement in this particular case that the customer wanted a longer deal, which enlarged your average?

Could this be a spreading phenomenon after all, notwithstanding your previous statements that three years wasn't necessarily in your interest or the customer's?

Sorry, longer than three years wasn't necessarily the thing to do.

I do think it is going to be unique to the customer situation, but I think the one thing when you do have let's say primary status with a customer, in some cases they do want a little longer footprint so they can establish that degree of stability.

But I do also think that as I think about this particular situation, we still have the ability to increase our footprint there, albeit a longer contract.

Because there's additional value we can bring to the table, which we will be able to grow our run rate at that customer over the term of that contract.

So we're not done there and I think that's the thing that we really are trying to focus on, both with our model that we've talked about with having our AMTs work with customers and just the way we're trying to sell the new technology.

Yes what I would chime in, is any time that we look at a contract life that's longer than the three year average that we're approximately operating on, the trick that we're trying to do is not have a limitless access to technology going forward.

It's -- that's what's unhealthy about the long, long, long contracts.

And in this case, while we can't tell you a lot of the details of it, we do have a nice balance with the customer and also giving them a lot of confidence in the kind of time line commitment that they have to the technology, the technology base.

And look here, things like ESL are great footprint extrapolations, that system designs space is fertile based on what we can build on top of our verification technology.

And I think that's a good example of the kinds of things that will allow the customers to evolve and allow our technology footprint to expand.

Thanks a lot.

Sure.

Your next question comes from Harlan Sur with Morgan Stanley.

(Inaudible) calling in for Harlan.

One question I had, just trying to get a read on the tone of your end customers.

If we look at our semiconductor universe, we see some mixed bag of results there.

Some companies are starting to struggle, maybe their business is kind of slowing down a bit, whereas other companies are still doing very well.

Wondering if there's any change in the general tone of your customer base when it comes to their planned EDA spend, especially going into the 2008 year?

Thank you.

Yes, think productivity.

What I find and it comes from some -- my past experiences is a heightened focus on productivity.

Some of that is to allow them to spend money in other places, like hire more software people.

That's not necessarily bad for us.

They rely more -- we and they rely more intensely on the value of our automation.

And also one of the reasons why we push for strong adjacencies up into that space predicate on what we do with verification, one of the examples I used was hardware, software co-verification at a big semiconductor company.

So productivity is also the thing that drives these-- some of the rational around the consolidations because they can operate more efficiently, and that's a training dynamic as much as anything else.

I think sometimes people get over focused on just as the core spending go down, people spend a lot of money on things besides licensing technology.

They've got to train the people and deal with (inaudible) all those kind of things.

So that's what we're responding to.

The other two focuses that we have that do resonate nicely are complexity in management and time-to-market.

In the consumer cycle, if you miss it, bad things happen, Christmas doesn't wait and all those kind of euphemisms.

And so across those three with a focus on productivity I think we'll cover most of the customer universe.

I see big companies and small companies alike.

[John] the only thing that I would add to the color is I do see with our experience with customers in Q3 is that they are holding onto their cash a little tighter.

And I think that's probably just reflective of some of that uncertainty that you're seeing out in that customer base as well.

And so that's probably adding a day or so to receivables.

But I think that being said, we're also just watching carefully R&D spending and at this stage I don't think we've seen anything from customers that's giving us an indication that they're going to dramatically change their R&D spending patterns.

And so I think we're watching carefully and we'll look at that in Q4 as we get ready to project what our business is for 2008.

Okay great thank you very much.

Your next question comes from Mahesh Sanganeria with RBC Capital Markets.

Hi, guys.

This is [KC] calling for Mahesh.

I have a couple questions for Mike and then a few for Bill.

Could you comment on the impact, if any, of the change in the ratables, both for this quarter and for the year?

Sure.

KC the-- I think one of the things that we have seen is that particularly with our experience with the EDA cards, we have the ability to work with customers to give them some more flexibility to, as Mike said, match their design needs with their purchasing to better align that.

And then given that we have worked on this model slowly, over the last two years, that it really has picked up steam for us in the last quarter, and just by reference we've seen our business with some of these term cards triple from Q2 to Q3.

And yet that allows us to really expand our run rate with these customers.

Because the experience that we're having is that customers, because they have the flexibility of buying more when they need it, are actually utilizing their technology at a rate that's about 47% faster than under a traditional time based contract.

And so what we see is the ability to increase our run rate and I think that is what's going to drive the mix going forward.

It'll be unique to customers.

We have some who still prefer to access new technology.

But we are seeing a trend for more of buying the strong technology as it's delivered and so I think we're seeing the view of that is reflecting in more term EDA cards and that's what is reflecting in our Q4 or my Q4 forecast for ratability.

Yes the tension in the system is always the discussion of access to future technology and consumption rate.

The reason that we got into this is to be able to have visibility into that.

The contention sometimes with the purchasing element as opposed to the developers is how fast or if they can strain the ability for the engineers to take the technology and this is a great demonstrable of sometimes it's more valuable to give the guys some runway and let them go.

And it's consistent with almost anybody whose a developer.

I'm an old developer, so the -- I think there's a -- it's a responsiveness to people wanting to spend the money in the way that they want to spend it and having very sophisticated mechanisms to do this.

It's very neat differentiating technology that we have in the computing systems to allow real time access to it and measure the usage models and not have a debate about it.

I mean, it's a database discussion then with the customers and that's our benefit.

Okay.

As a quick follow-up, can you comment on any changes to your revenue predictability because of a decreased ratable model.

Sure, [KC].

One of the questions that I've had in the past is what do we expect to come out of backlog in 2007.

And that really hasn't changed.

I still expect that about 2/3 of our revenue for the year is coming out of backlog.

And so that predictor of 2007 stays the same.

Now, if our experience continues to be what it is in Q3 and in Q4, and we're going to have about, let's say 50% of our business coming from these term contracts, then of course the math will show that over time we're going to have 50% coming out of backlog versus what we're seeing today of about 2/3.

And that's a trade-off that we're willing to make because we can see the ability to increase our run rate with our customers and it's something that we'll factor into our pipeline, albeit the math would show that you don't have that same predictability out of backlog, but we think it enhances our ability to sell more often to these customers and so I think we'll get it back in terms of a richer pipeline over time.

Absolutely and the one thing I would add, look it's a debate for the value of the technology.

Anybody who goes out and does extremely, extremely long deals is I think diminishing the value.

We've been very forth right to say we'd like to demonstrate the value to our customers and I think we'll get some return on that.

And this is just an ever increasing focus for us to try to do a good job at maximizing the value of the fantastic technology we deliver.

Great.

I've got two quick questions on the balance sheet.

Going forward, how should we view debt (inaudible) and also how should we view income from interest going forward?

Sure.

In terms of the debt on our balance sheet, as many of you know our original convertible, a portion of that will be callable this summer.

So the $230 million is now short-term and you can see that in the classification.

So you would expect that that would be retired.

And I think we'll look at the trade-offs to adding additional leverage on the balance sheet.

I've been asked that in the past and I think it makes sense for software companies and one particularly that generates good cash flow to have reasonable leverage.

So we will look at the timing and the availability of the debt markets to do that.

But I think we'll just watch that and see.

But our inclination is a little more is better.

In terms of other interest and expense, I think I gave a forecast there that is in the 40 to 45 for the year, so I think that's probably the best estimate we have right now and then again, we'll look at next year when we talk about that in January.

Okay.

Thank you.

You're welcome.

Your next question comes from Terence Whalen with Citi Investment Research.

Great thanks for taking my question.

This one relates to cash flow in the $450 million target for this year for cash flow operations.

What amount of receivable sales in the fourth quarter is implied to hit that $450 million target?

And then I have a follow-up.

Terence, I think that's going to really depend on what cash that we're going to be able to get from customers in the fourth quarter.

We don't have a specific target.

It's really the mix of business that we get.

And that is something -- we're feeling that we do have -- we're about the same level we were last year in terms of nine months and cash from operations.

It's a little over $200 million plus.

So yes we expect to have a good cash flow quarter in Q4 that goes along with our history and you could expect that we will be selling some receivables.

Those are things that the financial institutions value and we think it's something that we can get a good return on by using that cash in our business.

So don't have a specific number for you.

But it will be a reasonable amount of our normal cash from operations like it has been in the past.

Okay great.

Then another question related to, related to the license mix.

As you transition the business into next year, relying more on up front versus ratable, do you expect the frequency or the concentration of 10% plus customers in a quarter to increase versus 2007?

And then secondly, what changes are being made in terms of the approach based on the sales force perspective to customers with this model change and any other operational changes that go to the model change mix between ratable and up-front?

Thanks.

Terence, I don't know that we're going to see an increase in the number of large contracts.

I think there's a certain number of large customers in the world, all of which I think we serve.

And we work closely with those customers to meet their technology needs.

So I don't know that we're going to see anything dramatic in terms of a change there.

In terms of the working with the sales force, one of the things that we have and we alluded to earlier, is we have very good visibility into the consumption of the technology as does the customer.

And so that allows us to work very closely with the customer to help get them technology that they need when they need it.

Which is they appreciate it and it helps our sales force help them.

So I think we all have better knowledge.

And the other thing that I would just emphasize is this is a refinement of just the use that we see of customers with contracts.

I don't see this as any kind of a model change, just to be clear from my perspective.

Yes for the structure I'm out in front with Kevin a lot on-- with the customers.

What we do with the sales force specifically is we use the techniques that we have to reinforce the value and opportunity for run rate continuous business with the customers.

You can tell that by the testimonials that come quarter-after-quarter where you see like names over and over again and this is something that is a mind set with some of our field people, because a very old fashioned mindset would have been I got the contract and now I don't go back and look at it again for three more years.

There's business every quarter with expansion of the license usage that we have and/or new technology areas that are geocentric.

Because most of these big customers, as Bill said, are operating in all different places of the world and that's the refinement we've made that many people may talk about that's what we do.

And I think the team's doing a nice job there and so far we've got some more that we can take that.

That's why something like an EDA card is such a neat mechanism because it reinforces the mechanism that we use for our sales teams, our application and engineering teams to have that conversation with all aspects of the customer.

And your question is a good one that we work -- the sales force works closely because they can assist the customer to utilize that technology faster.

And I think that's the one experience that we have now seen over the two years that we have slowly introduced and then got acceptance of using these cards is that customers are actually utilizing the technology as I mentioned, 47% faster.

So it allows the sales force to work with them to get that new technology into those customers and to have additional selling opportunities and essentially grow our run rate.

Great.

Thank you.

(OPERATOR INSTRUCTIONS).

Your next question comes from Sterling Auty with JPMorgan.

Hey, guys.

Couple of questions.

Want to dig deeper into these EDA cards.

So if they're a term structure, that means they're locking into a specific product and a specific version that they can actually take down.

That's correct, right?

Yes, Sterling, they actually have a choice, so we offer both.

But with the ones that we have seen them adopting the fastest is the term type card.

That's correct.

So it is locked into a fixed price book.

Okay but you talked about that you've been able to give the customers the flexibility.

What do you mean by giving them the flexibility then?

What I mentioned, what they can do is they can draw the technology down faster.

They pay more money to do that.

It allows them to match a short term capacity with a demand that they might have.

And they're basically making a trade-off of whether they have a long-term commitment or not.

To me it's a little bit like renting a car.

If you know you're going to have a car for a year you might do something different than if you're going to rent it by the week.

You pay more to rent it by the week.

Well that works out really good for you when that's all the need you have.

And (inaudible) some of this is a little bit of the tension or the nominal conversation you have between the users of the technology and maybe even the purchasers of the technology.

And this is a real good objective way to be able to have that discussion without or/and constraining the access.

The things have a fixed duration so they expire and in that respect they're really no different than a regular contract so it's kind of use it or lose it and it works out good.

Yes, let me give you another example, Sterling, because it may help.

Say a traditional contract the customer would license 10 seats for three years.

All right.

They would have those 10 seats, they'd use them for three years.

If they get a card, they can start off with 10 seats and then as they see a project need, they can increase that and start going to say 15 or 20 seats and generate more capacity and burn through those seats faster.

So let's say at the end of two years they have burned through that technology, used it when they needed it and now are coming back for some additional technology.

So that's a way that customers can utilize the technology faster to meet their needs.

I guess I'm still a little bit confused.

So are you saying that out of the 10 seats per year type of thing and you're kind of using up the number of chips that you've got and then come back and purchase more versus if it's subscription you subscribe for 10 seats you can only use 10 seats at a time over the three years?

Generally, yes.

In my example, if they only had a contract that allowed them to get 10 seats, that what's their capacity would be.

Another way to thing about it and this is a good way to have a different discussion about remix, historically I bet people would have gone and tried to have an endless debate about I want infinite amount of remix opportunity.

Now you take that discussion out of play because people can buy the technology they want for a spontaneous demand.

And design is niche that way.

I mean you get going along the thing and you get close to a tape-out and you decide you need a bunch of more licenses for some element of signal integrity analysis.

He would-- historically, he might have been ultimately constrained and so you say the engineers work harder or fail.

This gives them a very easy way to go and blip that technology up without having that debate and pay a little bit more for the pleasure or the privilege of being able to do that.

We both are happy and off you go.

Okay.

And what are the collection-- what's the collection structure in terms of the payment under a term based EDA card?

The payment structures are very much like a traditional term contract and of course we incent the field to try to get more cash up front, and we're continuing to look to do more of that.

But essentially it's payment over time or with incentives for our sales force, as you know they get commissioned when we get the cash, is to more and more get that up front.

But we allow both.

But with the jump in DSOs, are you seeing that most of the customers are choosing to pay annually over three years or what is the experience that you're seeing with these?

Well, I think the experience is about the same that we see with normal term contracts.

I don't think we've seen anything significantly different so far.

Okay.

And then you saw there's a jump-up in long-term deferred.

Can you give us a little bit more color there?

Yes, I think as I've described in the past, that's just our ability in some cases to get cash ahead of when revenue is recognized.

And so that's -- when you have increases in either short-term or long-term deferred, that's just about getting some of that cash before the revenue is recognized.

I wouldn't try to make much more other than that.

Okay.

And then last question, Mike, kind of curious what you're seeing out there in the marketplace.

It would seem to me that 2005, 2006 you guys had a field day I think taking market share away from Mentor and Emulation Space.

I think now [Voloche] has been out there for a number of quarters, seems to be gaining traction within the old, existing Mentor accounts.

Are you finding that it's harder to win some of those accounts now?

Well, I think it's -- that's a good observation.

Anybody will first try to farm their installed base and that was one of the reasons why I put in the prepared comments some of the momentum that we have around our Emulator families, it's not just a one hit wonder with us, we've got some great products across the acceleration all the way up through the Emulation.

We have a targeted program to go back and upgrade our base and doing very well.

And we have seen some competitive momentum in market segment share shift as well.

But it's not been totally the focus.

I mean, it's -- and we've got the next generation Palladium coming to debut next year.

And so we've got a very nice potpourri of systems all the way from Acceleration up to Palladium III and then the anticipation coming of the unified architecture.

So I like our progress.

I don't think that we've lost anything that we didn't expect.

We've been very targeted and the real value in the Emulation for us isn't just in the functional logic simulation.

It's in the software, hardware, co-verification.

That's why I think that testimonial that I rattled off was indicative.

And that is the strength of our total verification approach and story, aided with Emulation as a neat piece of it.

I think we have the most comprehensive offering and strategies that way and the stuff that we're doing with some of these leading edge customers is very reinforcive of that as a component of growth for our Emulation capability.

Sterling I would just add that we are continuing to grow that piece of the business.

It is growing for us year-over-year in both existing and new accounts.

All right.

Great.

Thank you.

Your next question comes from Benjamin Pappas with D.A.

Davidson.

Hi guys this is Ben Pappas calling in for Matt Petkun.

Just a couple quick questions, if I could.

You mentioned the relationship with NXP is expanding.

Is it fair to assume then that the overall revenue run rate will increase on an annual basis?

Yes Ben, this is Bill.

I did mention that I think as our relationship expanded, I do expect that we will over time increase our run rate as we deliver new capabilities to the customer.

Okay.

Fair enough.

And then one quick other question.

In the past I know you used to mention a percent of new bookings booked at subscription, do you have that number again?

Well the-- I think what you're referring to is what I talked a little earlier about in terms of the percentage of our business that's ratable, and in Q3 it was 50%.

Okay.

Great.

Thanks, guys.

Sure.

This concludes our conference call for today.

Thank you for participating on the Cadence third quarter 2007 earnings conference call.

You may all disconnect.