新思科技 (SNPS) 2004 Q3 法說會逐字稿

Welcome to the Synopsys third quarter fiscal year 2004 earnings conference call. At this time all participants are in a listen-only mode. Later we will conduct a question-and-answer session and instructions will be given at that time. (OPERATOR INSTRUCTIONS). Today's call is scheduled to last one hour. Five minutes prior to the end of the call I will alert the conference of the time remaining. As a reminder, today's call is being recorded.

During the course of this conference call Synopsys may make predictions, estimations and other forward-looking statements regarding the Company. While these statements represent the best current judgment about the Company's future performance, the Company's actual performance is subject to significant risks and uncertainties that could cause actual results to differ materially from those that may be projected. In addition to any risks that may be highlighted during the conference call, important factors that could cause the Company's actual results to differ materially from those that may be projected in this conference call are described in Synopsys' quarterly report on Form 10-Q for the second quarter of the fiscal 2004, which is on file with the Securities and Exchange Commission, and in Synopsys' earnings release containing its third-quarter financial results, which is posted on its website.

In addition, Synopsys would like to advise you that financial and other statistical information to be discussed on this conference call, as well as the reconciliation of certain non-GAAP financial measures discussed on this call to GAAP financial measures, is now available on the Company's website and the earnings release containing its third quarter financial results, as well as in a financial supplement. The Web address for such information, including the earnings release, is http://www.Synopsys.com/corporate/invest/invest.html.

At this time, I would now like to turn the conference over to Aart de Geus, Chairman and Chief Executive Officer. Please go ahead, sir.

Thank you, operator, and thank you all for joining us. This is Aart de Geus, and I have with me Steve Shevick, our CFO.

I would like to address five points today -- our Q3 results; our view of the spending environment; the state of our products; what we plan to do going forward; and our outlook for Q4 and our preliminary outlook for fiscal '05.

Clearly, Q3 was an unexpectedly tough quarter for us. Revenue was 282 million, non-GAAP earnings were 33 cents per share and book-to-bill was approximately 0.7. These results were below our targets mainly due to like-bookings for upfront licenses and several contracts that customers unexpectedly pushed out in the last days or even hours of the quarter.

We're not satisfied with these results. Accordingly, we have taken a hard look at the market, at our customers' needs, our technology, and our operational execution. The result is that we are taking immediate steps to move to a more fully-ratable business model, starting with the present quarter.

Let me retrace our observations and conclusions in more detail for you. In terms of the spending environment, in July our customers became markedly more cautious about extending commitments and spending cash. This caution was triggered by some inventory buildup and a lack of visibility. Since then, a number of our customers have announced earnings below expectations or reduced forecasts. We experienced this development in their deciding to wait and see rather than buy incremental software or extend existing license agreements.

In addition, customers became more reluctant to part with cash for upfront licenses. Given that in our model, early payments are required for us to take upfront revenue, our quarterly terms (ph) business was well below the targets we had set for the quarter.

What about our technology? Last quarter, you heard me speak very enthusiastically about the new capabilities we were going to roll out at the June Design Automation Conference. We have delivered fully on those capabilities and my enthusiasm has not waned.

Galaxy 2004, our implementation platform, was released on schedule with products ranging from floor planning, synthesis and placement route, to extraction, timing verification and tests. With the June release, all these products are now on the same area, timing, power and test infrastructure, making it the most complete high-performance, lowest risk design system on the market.

Customer feedback both in the press and in our user forum confirms the bold capacity and performance improvement claims we made about this new technology. During the quarter we already saw its impact in strong benchmark results against both Magma and Cadence.

Moving forward, 2005 will be a year of further rollouts for Synopsys. Our strong integrated platforms will help customers transition (technical difficulty) to the complete solution they are seeking for. With the spending environment in a wait-and-see mode, however, we don't want to compromise the pricing of our new technology. That is why we have decided to immediately complete our transition to a maximally subscription-based model, where 90 percent or more of our revenue rolls off of our backlog every quarter.

This strategy has four compelling benefits. First, it will preserve pricing on our new technology because no single deal will make or break a quarter. Too often, upfront transactions at the end of the quarter become more about price than about value. Second, with customers conserving cash, subscription licenses will be more attractive than upfront payment alternatives, thereby allowing us to more aggressively pursue competitive opportunities. Third, customers are increasingly looking for complete solutions that include tools, IP, and services. Subscription licenses enable us to offer our customers complete solutions in a flexible manner to better match their current needs. And fourth, subscription licenses give us much greater visibility into our earnings stream and decrease end-of-quarter uncertainty. Overall, we believe that a greater focus on subscription licenses is best for Synopsys and for our customers.

Looking forward. With greater uncertainty around the spending environment, we are no longer counting on the recovery expected in the second half of the calendar year. In addition, we will be in the low part of our renewal cycle for the next five quarters with our global and strategic accounts, which make up to 2/3 of our business. These accounts have sustained us for the past three years while the smaller territory accounts, which made up half of our business until 2001, declined in the post-global era.

Over the last three years we have successfully established long-term agreements with the largest accounts and in many cases renewed their agreements on a rolling basis. In the process, we have become the primary supplier to many of the top 25 semiconductor and system houses, and built a noncancelable backlog of roughly $1.4 billion. These are valuable relationships and provide a great position to build on. Having put in place long-term contracts in light of the new technology rolling out and a shift in our license mix, we will let our new technology build momentum in the market and let utilization within the customer base drive incremental demand.

In the meantime, we have many opportunities. We will pursue incremental business by displacing competitors, by capturing new projects, by expanding utilization of existing licenses, and by selling upgrades in our own base. We also believe that opportunities in our smaller accounts have solidified and we will pursue healthy business there. Given the relatively low level of scheduled renewals, we expect aggregate bookings for fiscal 2005 to be materially lower than current analyst estimates, which is why we are providing very preliminary targets today.

Assuming no rollover of large orders from Q4 into Q1 '05, our preliminary view is that orders for fiscal '05 will be in the 900 million range. Of course, this number rebounds significantly as we move into '06 and a substantial number of large contracts come up for renewal.

Since we will book only a small number of upfront licenses, fiscal '05 revenue is expected on a preliminary basis to be approximately 940 million. Because we are making the shift immediately, Q4 revenue is expected to be in the range of 220 to 240 million. We will manage expenses aggressively while making sure that our customers are well served and are taking full advantage of the new technology rolling out. Non-GAAP earnings are expected to be between zero and 4 cents for Q4 and between 28 and 38 cents for fiscal '05. Steve will give further details in a moment.

Let me give you now some color on the third quarter. From a geographic perspective, Europe and Asia-Pacific performed reasonably well, while North America and Japan were weaker than expected. From a product point of view, the order shortfall was felt across all major product lines. Services and floor planning standout as the strongest performers for the quarter, both exceeding plan.

Let me now give you some of our product highlights. Starting with our implementation platform, in Q3 we shipped Galaxy 2004, with an outstanding set of major enhancements. Not only do we have dramatic improvements in area, timing, power and test correlation among all our products, almost every product has delivered substantial run-time and capacity improvements as well.

To give you some comparisons to the year ago, design (indiscernible) is 3X faster with 3X the capacity. Physical (indiscernible) capacity increased by 2X, while run-time improved by 1.5X. (indiscernible) increased by 2X on 32-bit machines.

Quality of results improved for all products. In addition, our coverage of a wide range of low-power issues is a marked differentiator for Galaxy. In many ways, we see the Synopsys R&D engine hitting its stride, with new capabilities rolling out on schedule. Clearly, Synopsys remains the top choice for advanced chip design.

To give you an update on advanced chip activities, consider the following. Last quarter we saw 194 90 nm designs. This quarter we see 227. We are confident that we see at least 90 percent of all designs, since practically all of them use our sign-off tool. The 90 nm tape out count stands at 80, with 51 of them having used exclusively Synopsys placement route and an additional 16 using a mix of Synopsys and competitors' products. We presently track 29 65 nm and three 45 nm design. The first successful 65 nm tape out was done entirely with Synopsys tools and a second tape out with another customer is expected within the month.

Looking forward, we have a strong technology pipeline leading to Galaxy 2005. Our new floor planner, JupiterXT, was released this quarter, and is consistently winning benchmarks against Cadence. In Q3, JupiterXT orders were ahead of plan and are expected to grow rapidly, with adoption of (indiscernible) and inclusion in the start (ph) reference flow.

During Q3, TSMC announced its reference flow release, 5.0, which incorporates Galaxy, including JupiterXT, for designs at 130 and 90 nm. Synopsys is the only (indiscernible) supplier with a complete solution at TSMC. In Q3, we also announced validation of both our Galaxy and Discovery platforms on the 90 nm process common to IBM and Charter (ph) Semiconductor.

With the move to slower geometries, the emphasis on design flow manufacturing is increasing. Photolithography and yield optimization are big challenges for our semiconductor customers. Our installed basis of Hercules, OPC and fracturing software are all growing, and we expect customer interest to increase as 90 nm becomes more common.

During Q3, Semiconductor Manufacturing International Corporation, the leading foundry in China, adopted our OPC, which enabled them to obtain results and dramatically reduce mask synthesis turnaround time.

Looking now at the Discovery Verification Platform, we have made great strides there as well. With the May release of VCS, we now deliver native testbench capability, built-in coverage and complete assertions. These features make VCS both the highest performing and the most complete RCL verification solution in the industry.

We currently have over 20 competitive engagements with upside opportunity. In Q3, we recorded eight successful displacements for VCS and added 28 new logos for Discovery. Progress on system analog is strong as well. Over 40 vendors committed (indiscernible) roadmaps for support, which bodes well for its adoption, and we're tracking 25 engagements with customers working to standardize on system (indiscernible).

Moving to IP and services, our solutions group posted another solid quarter in design services, with year-to-date bookings and revenues up 77 percent and 12 percent, respectively. During Q3 (technical difficulty) new Serial ATA and high-speed on-the-go IP course. Meanwhile, our PCI Express core passed industry compliance and interoperability tests. We added three new logos to the PCI Express customer list in Q3 and we realized the industry's first PCI Express IP tape out by Realtek.

In DesignWare we announced AMBA Connect, which dramatically reduces design time and risk for AMBA BUS-based designs. We also announced and (indiscernible) partnership on verification IP around the next generation AXI BUS standard.

In summary, while we're not satisfied with our Q3 results, we do see an opportunity to put the Company on a stronger long-term footing by moving to a maximally-ratable business model now. Going forward, we are committed to meeting our financial targets, rolling out our new technology on schedule, and transforming ourselves into the key provider of complete, differentiated solutions in the broader emerging market of telephone (ph) infrastructure.

With that, let me pass it to Steve Shevick, our CFO.

Thanks, Art, and good afternoon. During this call I will discuss certain non-GAAP measures of our financial performance. These measures exclude amortization of intangible assets and merger-related and other special charges amounting to 34 million in Q3. Reconciliation of non-GAAP results to GAAP results appears in our press release and in the financial supplement posted on our website. Copies of our remarks will also be posted. Now for the numbers.

Q3 was a difficult quarter for us. Orders fell short of plan, with book-to-bill of approximately 0.7 versus an expectation of one or better. Revenue for the third quarter was 282 million, below our beginning of quarter guidance range of 300 million to 320 million and within the range we announced on August 2nd.

The brief explanation of why we fell short is that we did not book and ship our forecasted level of upfront licenses during the quarter. Though the pipeline was sufficient, too many opportunities fell out late in the quarter. Many opportunities fell out of the pipeline due to our payment terms requirements.

As I have explained in past quarters, in order for us to recognize revenue upfront, the customer must pay 75 percent of the license amount plus the first year's maintenance within one year of shipment. This is the most stringent standard applied by the major EDA vendors. As Q3 developed, customers became less willing to agree to these terms. As a result, we booked 32 million of upfront licenses in Q3 versus an internal expectation of at least (technical difficulty) million. Upfront licenses constituted 20 percent of total product orders versus an expectation of 30 percent.

Renewable licenses, which includes subscription licenses and term licenses, were 86 percent of software orders. Perpetual licenses were 14 percent. The average length of our renewable customer license commitments was 3.4 years. No customer accounted for more than 10 percent of orders in the quarter.

Upfront revenue for the quarter was $62 million, including more than 30 million from backlog. Upfront revenue accounted for 22 percent of total revenue, time-based licenses 58 percent, and maintenance and consulting 20 percent. Over 85 percent of revenue came off backlog in Q3 versus an expectation of 75 percent. The higher level is mainly due to the lower-than-expected level of turns revenue.

Compared to Q3 of FY '03, revenue declined by 6 percent. Time-based license revenue increased by 4 million, or 2 percent, to 164 million, while upfront revenue declined by 17 percent to 62 million. Services revenue declined by 15 percent to 55 million, with a 7 percent increase in consulting revenue being more than offset by a decline in maintenance revenue. One customer accounted for more than 10 percent of revenue in the quarter.

The geographic distribution of revenue among the regions was consistent with recent historical performance. North America contributed 56 percent of revenue; Japan contributed 16 percent of revenue; Europe contributed 16 percent of revenue; and Asia-Pacific contributed 11 percent of revenue.

For the quarter, 81 percent of revenue came from our core EDA products in design and verification versus 83 percent in Q2. 14 percent of revenue came from our emerging businesses in IP and design for manufacturing versus 13 percent in Q2. 4 percent came from services versus 3 percent in Q2. Aggregate non-GAAP operating expenses for the third quarter were 206 million, well below the bottom of our target range of 220 to 230 million. This was principally due to lower-than-expected sales and marketing expense and other employee expenses, including a lower level of expense for incentive compensation associated with the lower orders levels.

Operating expenses included approximately 5 million in legal and other expenses relating to the MoSys transaction. The MoSys litigation was settled in July and there's no further potential liability for the Company.

Our non-GAAP operating margin for the quarter was 26.9 percent, and for the year-to-date it was 27.6 percent. Other income for Q3 was 1.3 million, a little above our target range of -2 million to 1 million. Our non-GAAP tax rate was 31 percent, consistent with Q2. Weighted average shares outstanding for the quarter was 160 million shares, within our target range of 158 to 166 million shares.

Non-GAAP earnings was 33 cents per share, below our beginning of quarter guidance range of 35 to 40 cents per share and within the revised range we announced on August 2nd. Cash flow was one of the bright spots for the quarter, with non-GAAP free cash flow of 123 million. GAAP cash flow from operations was 130 million, CapEx was 11 million for the quarter, and acquisition-related disbursements made up the difference.

Cash and short-term investments totaled 699 million at the end of Q3, up 86 million from the end of Q2, due principally to higher collections and a lower level of stock repurchases than in Q2. During the quarter, we repurchased approximately 1.7 million shares of Synopsys stock for an aggregate of 50 million. As of today, approximately 212 million remains in our repurchase program; we expect to be in the market this quarter.

Q3 accounts receivable totaled 143 million, down approximately 78 million from the end of Q2, principally reflecting strong collections, the lower level of invoicing related to Q2 bookings, and the continued demand of customers for extended payment terms on subscription licenses. During the quarter, payment terms on time-based licenses lengthened compared to the last four quarters' average. DSO was 46 days, down from 68 days at the end of Q2. DSO is likely to increase as a result of the change in license mix since quarterly revenue will decline.

Deferred revenue at the end of the quarter was 415 million, down 37 million from the end of Q2, reflecting Q3 bookings and the timing of invoices associated with installment contracts in the quarter. Aggregate backlog was down from the April quarter, reflecting the book-to-bill ratio.

Headcount was 4388 at the end of the quarter, up from 4249 at the end of Q2, primarily reflecting employees added to expansion of our consulting capacity and targeted hiring in other parts the business. Headcount is likely to grow modestly in the fourth quarter based on offers already out to candidates.

Moving to our targets for the fourth quarter and the full fiscal year 2004. In both cases, revenue, earnings and cash flow are directly affected by our performance in Q3, and more significantly by the shift in our license mix.

For Q4, we have reduced our expectation of upfront revenue orders by approximately 85 million. This reduction directly impacts total revenue and earnings. To the extent that these orders are now booked as time-based licenses, they will go into backlog and should be taken to revenue in future periods. The license mix shift will also affect operating cash flow. Though we will continue to push aggressively on payment terms where possible, in general, we expect payments on time-based licenses to be more extended than what we have experienced on upfront licenses. These effects are reflected in our targets for Q4 and the full fiscal year.

For the fourth quarter our targets are revenue between 220 million and 240 million, of which we expect 90 percent or more to come from backlog, which is an immediate benefit of our shift to a more ratable license mix; total non-GAAP expenses between 222 million and 232 million; other I&E between -1.5 million and 1.5 million; non-GAAP tax rate of 31 percent; fully diluted shares between 150 million and 158 million, and non-GAAP earnings between 0 and four cents per share.

We expect time-based licenses to contribute 68 to 73 percent of total revenue, upfront licenses to contribute 7 to 12 percent of total revenue, and services to contribute 18 to 23 percent of total revenue.

Our targets for the full year are as follows. Revenue between 1.08 billion and 1.1 billion, versus our previous range of 1.2 to 1.23 billion; total non-GAAP expenses between 846 and 856 million; other income and expense between zero and 3 million; non-GAAP tax rate of 31 percent; fully diluted shares between 158 million and 162 million; and non-GAAP earnings per share between $1.01 cent and $1.05, versus 1.37 to 1.47 in our previous guidance.

We expect non-GAAP free cash flow in excess of $1.40 per share versus our previous guidance of at least $1.80 per share. And we expect TDLs to contribute 58 to 62 percent of total revenue, upfront license revenue to contribute 18 to 22 percent of total revenue, and services to contribute 17 to 21 percent of total revenue.

Orders for the year are expected to be below our previous target of 1.4 billion, and at this point in the year, we are not providing a new orders target. As we have said all year, we expect to receive a disproportionate share of our orders in the fourth quarter of this year. The timing of completion of large agreements currently in the pipeline will materially affect fourth-quarter orders -- and thus, full-year orders -- and negotiations on any of them could carry over into fiscal year '05. We will report orders data at the end of the fiscal year but won't comment on any individual transaction during the quarter.

For 2005, we are providing only preliminary targets. At this point, we are targeting orders of approximately 900 million, revenue of approximately 940 million, and non-GAAP earnings in the range of 28 to 38 cents per share. We will refine these targets next quarter, which is when we would normally give annual guidance.

The license mix shift has a meaningful impact on revenue and earnings in 2005. Assuming 900 million in orders, we estimate the impact of this shift from 30 percent upfront licenses to 10 percent upfront licenses to be over $140 million. We expect that 90 percent or more of each quarters revenue will come from backlog, as we will be far less dependent on current quarter revenue from upfront licenses.

On an annual basis, we would expect upfront orders to be roughly 10 percent of product orders, though this number should be expected to fluctuate from quarter to quarter. We will report the percentage each quarter but will not set quarterly targets, in order to preserve our flexibility to close business when it makes the most sense.

We will be vigilant on expenses and are considering our various options for expense controls. We will take a hard look at all projects during the fourth quarter, and will focus on those projects that offer the best opportunities for return. In summary, effective this quarter, we're at the beginning of another transition -- this time to an almost fully-ratable license model.

As we saw in 2001, revenue will initially dip as orders that were taken immediately to revenue instead go into backlog, and then will grow both as orders grow and as layers of subscription revenue build on top of each other.

The benefits of the shift are enormous. We believe that this change will enable us to offer customers technology in terms that more closely match their needs; will lead to an improved pricing environment; will help us roll-out our new technology in a more planned manner; will give us greater visibility into our earnings stream; and will decrease end-of-the-quarter uncertainty.

On orders, we are heavily influenced by market conditions and the renewal cycle. Though we will experience a very low renewals year in 2005, the cycle should improve in 2006 and beyond. Based on current expiration dates and expected timing of renewals, plus other reasonable assumptions, we expect to see orders grow over 25 percent from fiscal 2005 to fiscal year 2006, with additional growth in the years beyond.

Operator, we'll now open for questions

(OPERATOR INSTRUCTIONS). Sumit Dhanda, Banc of America.

A couple of questions. Hopefully you can hear me clearly on the cellphone. First off, any guidance on free cash flow for fiscal '05, if you could give us a ballpark number there based on new license mix? And the second question I had really was on your radically changed outlook for '05 -- even up until three or four months ago your expectation was for a much better '05, and even though you hadn't given formal guidance of $1.90, earnings per share was something you talked about. Exactly what has changed for such a drastic reduction in order expectations going into '05. Frankly (indiscernible) by this.

Sumit, it's Steve. Let me answer the cash flow question first; then I'll turn it over to Art. On cash flow -- at this point, since we've really only given preliminary guidance on '05 we're not going to give a cash flow number. In general, as I said, the payment terms on subscription licenses tend to be more extended than on upfront licenses. So you would expect to see payment terms lengthen out a little bit, which would affect cash flow. And we'll give some more guidance at the end of next quarter.

Sumit, on what has changed, I think the biggest change really occurred in July where we saw that the customer base around (indiscernible) became markedly more nervous about the future, maybe triggered by some sense of inventory buildup, some uncertainty and non visibility of '05. But definitely much more cautious in terms of making longer-term commitments. And given that we are looking at a part of the renewal cycle that has fairly few renewals in the next five quarters, it is clear that pushing on the accelerator with the customers would have entailed significant price concessions, which is something we most definitely don't want to do at the very moment that we have just rolled out a lot of strong technology unless we're going to rollout some more. And by the way, the same comments apply to the upfronts there even more so. Our customers are supposed to pay with cash early on, and that is something that right now I think is difficult to get.

Let me ask a couple of quick follow-ups. What is your conviction that this is really (inaudible) of losses? I understand the caution that your customers exhibited in the month of July; that's evident in most semiconductor company reports. But this change (indiscernible) dramatic to attribute to just one month of caution on the part of your customers. And the second question I have for you is you talked a lot about the visibility associated with the subscription model and your change to a fully subscription model here, but clearly at least this year it's not helped you in any way or form. So what gives you confidence that this is the model that's going to help visibility going forward?

Two comments. The other thing that, obviously, changed is we missed the quarter. And in the history of Synopsys, we have never missed a quarter; we do not take something like that lightly at all. We will use this as a catalyst to revisit every nook and cranny of the Company to see if we can execute well. But we also looked hard at the market around us and saw that it's tightening up. We have been able to build a very strong noncancelable backlog with many customers, so in that sense they are solidly linked to us, if you like. And we have technology that has high value. Going forward as you well know, we were the company that led the way towards going much more ratable a few years ago. We saw that for a long period of time that was a major advantage in negotiations, and coming into the year we thought that we would be able to do about 250 or plus million in upfront. That is clearly not desirable or potentially not achievable unless one gives major price concessions. So that is really what has driven us to make this change immediately.

And the share losses?

I'm sorry; I didn't hear the second question.

(inaudible) none of this is attributable to share losses (multiple speakers)

Sure. Well, for the share comparison, of course, we will need to play out a little bit through the model change. But one of the things that we are absolutely seeing is that from a benchmarking point of view, we're doing quite well. And so, this is especially true since the release of Galaxy 2004, and from that perspective, we think that competitively we are in a strong position. All the more reason to make sure that we not only hold onto our customers but take advantage of the subscription model going forward. Next question, please?

Jay Vleeschhouwer, Merrill Lynch.

Aart, in looking at your bookings number for next year, 900 million, that total number is less than the pre-Avanti Synopsys product-only number in 2001, which means that after four years you'll have made no bookings progress -- in fact, regressed significantly. So why shouldn't we infer from that some portfolio or product mix issue, notwithstanding improvements you have made in the products. Doesn't it fundamentally suggest an overdependence on areas that really aren't growing, and consequently you're not therefore making bookings progress.

Okay. The first aspect to this is really that all of the orders and all the business is really now very much a multiyear business. So we're in a cycle that is 3 to 3.5 years, and depending where one is in the cycle, the bookings will be much higher or much lower. A good example actually will be between '04 and '05, depending on where some of the large deals that are right now aimed at for Q4 end up -- depending on which side of the quarterly boundary they are -- they will have a big impact on one or the other year. And so from that perspective these fluctuations are not abnormal.

From the perspective of the overall portfolio, that's a very valid question. And clearly, when we acquired Avanti, our first objective was to have a complete portfolio (indiscernible) design, and with very, very few small holes, we, I think, have by far the most complete portfolio. The second objective was to make that portfolio become an overall integrated platform. That has taken us a couple of years and took a lot of work and it's only really in Galaxy '04 that we feel that we have that. And the benefit of a complete portfolio really comes to bear once the products work well together, and that is the place that we have achieved now. So from that perspective, I think we're looking forward with a fairly high degree of confidence.

But to follow-up on the point from your prior questioner -- at one point you had sustained $1.90 estimate for next year, admittedly, with higher percentages upfront. But you must have had some assumptions of larger deals with existing customers. In other words, you're now saying that the low point in terms of renewals, but that would seem to not be consistent with what it would have taken to sustain the $1.90 number for next year.

One of the things that made us suddenly bullish early entering the year was that we saw a consecutive growth of the semiconductor industry doing quite well, and two things really did not materialize. One is that -- the strong sense that the second half was going to be very strong is now suddenly in question, and the other thing is that the percentage of R&D expenditures on the semiconductor side that's flowing to EDA has been not as strong as hoped-for. And so from that perspective we want to be a bit more careful.

Finally, I would add to that that as we have done a number of deals with customers locking them in, right now they are fairly careful in committing beyond that -- not because they will not use our technology, but because they've learned some pretty tough lessons in 2001, and one of them is to make sure that you stay close to your cash. And so the upfront transitions have not -- are not -- are something that we need to pay a lot of attention to.

Two last ones, then. You didn't say anything specifically about cost structure or physically reducing your cost structure, or any other kind of management changes. I had thought that you might have more to say about that on the call tonight in light of the revenue and bookings outlook reduction. And then lastly, this is now the third model change in effect in the last four years with many of the same arguments being offered in terms of visibility, price protection and so forth, and yet on more than a number of occasions over the last 4 years you have digressed from your model. So what is your (indiscernible) now with this iteration that you won't digress in the model and, for instance, allow upfronts to reach back into the stream?

First regarding the cost transition. One of the things I did say is that we would pay a high degree of attention specifically to the expenses. I also said that we want to make sure that we support our customers well at the moment that we are rolling out a lot of technology. And so we will engage in time management (indiscernible) of the company to look at how we do that most effectively, while at the same time making sure that we build on the strength that we do have.

Regarding the model change, I think looking back there's no question that going to a ratable model has, all in all, served us well. And recall that in this process we did build about 1.4 billion in backlog that is committed and noncancelable, and has actually given us a very strong position to weather any potential storm that could be happening in our customer base. At the same time, there is no question that as we entered this fiscal year, we had higher hopes for the degree of upfronts. And all the way into Q2 we were able to get payment terms with customers that supported that. That started to change, and therefore we decided to take action.

Bill Frerichs, DA Davidson.

First, Steve, what is your revenue recognition cutoff for subscriptions in a given quarter? When do they have to book by in order to take them?

We are on a daily revenue system, Bill, so we effectively get as many days from when it comes in until the end of the quarter on subscriptions.

But from the last day you get one day?

Yes.

Okay. Aart, going back to your earlier comments, I presume -- reading between the lines -- that this quarter that we're currently in could still be a fairly big bookings quarter for you, right?

We have certainly said so. And yes, as you know, there are a number of large deals that we have planned for this quarter. We want to make sure that we take advantage of the very fact that we cannot be held up at the end of the quarter to get good terms, but to be objective and clearly to close the business this quarter.

Finally, thank you for the extraordinary update on the backlog number. I take that that's a current number through the end of last month?

Yes.

What is the aging of that backlog, Steve?

We haven't generally given that out, but our profile is generally close to 50 percent in the first year, 30 percent second year.

Harlan Sur, Morgan Stanley.

A question for Aart. Design consulting services activities for you, as well as for Cadence, is up pretty substantially. Design starts are up based on your numbers. Tape out activity is certainly increasing. And it seems to me that the semiconductor companies are transitioning to 90 nm at a fairly good pace, and companies that have product schedules to hit market windows over the next 18 to 20 months on 90 nm, to me it would seem, cannot push out tools purchases. So maybe you can help me better understand why the near-term perturbations in the semiconductor environment are having such a drastic impact on your business?

Certainly -- again, let's make a distinction between revenue and orders. On the revenue side, clearly the issue was the upfront percentage in our business. On the order side, there the story is more complicated, because our customers make multiyear commitments. And so the customers that you are referring to and (technical difficulty) very well because we do a majority of the designs -- of the advanced designs with them, have our tools for the next year or two. And than the question is at which point in time does their renewal come up? As the confidence of a customer increases or decreases, they tend to move renewals up or down a little bit, make more or less large commitments going forward. But in general, they will most definitely not let any of their tool subscriptions lapse because they are absolutely vital.

You also mentioned that the service business tends to be strong. I think that is for two reasons. One is that many of our customers have been reducing forces in '01 and '02, potentially beyond where they should have. And secondly, because in a number of situations design support is a necessity in order to be successful with some of the most advanced designs. And there we really become a member of their team.

So is it fair to say that from a design start activity, from a process technology transition -- whether it's to 90 nm or 65 nm or 45 nm -- you haven't seen any sort of letup in the acceleration there?

I think it's fair to say that. I would caution, though, that the number of designs at 90 nm compared to 130 or 180, no matter what, is still small. These are very advanced and these are the people that absolutely spend most money on a design and work closely with us. Now, if you take for example the 90 nm tape outs, of which there were 80 so far, we -- and when I say we, I should really say -- our customers with singly our tools -- there's 51 of those. So the majority of that. So we know them quite well. So I think that transition is actually well on track. 90 nm does not appear to have any of the endemic problems that 130 nm had, but at the end of the day, what will govern the massive transition is really the pricing schemes between 90, 130 and 180.

Raj Seth, SG Cowen.

Aart, I just want to go back again to this '05 bookings number; maybe you can help me understand this a bit better. When you were outperforming in '03, some of your critics suggested that your business was driven, obviously, on the back of some Avanti renewals, and that you effectively were pulling some renewals forward, given the opportunity to enter discussions because some Avanti stuff was renewing. At the time you pushed back pretty hard and said no, this isn't in fact happening. And I guess I'm curious with your '05 not being a strong renewal year, coupled with the comment you made about when renewals happen, customers are not likely to let natural renewals lapse, why shouldn't the conclusion now be that in fact that is what happened, that some of the renewal activity was simply pulled forward? And if your view on the '05 renewals schedule hasn't changed, I still can't understand how you were giving guidance that was effectively suggesting bookings above 1.4 billion for '05, and it's now sitting at 900, assuming the average duration of these deals hasn't changed. I just don't get it.

First comment is that the bookings that you get with a customer are influenced somewhat by the condition that one is willing to give, but not extraordinarily so either. In other words, customers renew when they are sort of ready to renew. When we acquired Avanti there was no question that many of our customers felt that the technology was in good hands, that that company then had a future, and therefore -- and, by the way, that some of the technology that Avanti had stand-alone was very strong. And so we were able to renew a number of the deals that otherwise may have disappeared or gone to competitors.

At the same time, if you look at the timeline of renewals, as the -- for lack of a better term -- the mood or the level of caution goes up and down, customers dial up and down a little bit their appetite for future commitments. That doesn't mean that their renewal cycle is really off. There is no customer that has a viable business that would ever let any of these deals lapse. That's just unthinkable. So typically all of the renewals tend to be a certain period of time before that would occur. And it is not so much that period of time that changes all that much as more the level of commitments that see some variability.

Lastly, you're absolutely correct that with the Avanti opportunity came the opportunity for us to be strongly in the (indiscernible) and (indiscernible) domain. And the very fact that we are able to renew many of the existing Avanti deals well helps us hold solidly onto the installed base at the moment that we then have to spend about 24 months going into primarily integration mode. And one of the keys to the success going forward is the very fact that we're done with that, and Galaxy '04 now provides capabilities beyond that.

I acknowledge what you're saying about customers dialing up or down a little bit, I think that's already used on the size of renewals. But if you were at 1.4 billion in '03, we are talking about -- I don't know -- high single-digit or maybe even double-digit growth into '05. 1.5 billion to 900 million certainly isn't a little bit. You would think that you had maybe a better understanding of those renewal schedules even a couple of quarters ago. But maybe I can take some additional questions off-line.

One thing maybe you can touch on, Steve. You mentioned inventory. Can you expand a little bit on that? Is there a lot of design software out there that is effectively shelfware that people aren't using right now, and you are waiting for -- I think you used the word utilization -- are you waiting for utilization to get to the point that they have to buy more? Can you expand a little bit on that?

Sure. I think it was in our script, but what -- we have been tracking utilization rates in our key customers as best we can, and they tend to vary all over the place with some customers having a need now; other customers getting to the point where they will need to buy; and then still others having some extra software that they'll need to grow into. What we are looking for is to see the -- watching the overall level of utilization grow to drive incremental purchases.

And where do you think utilization is in your installed base right now?

I think giving a broad number across the whole base is almost impossible, because we have estimates for each customer. So by the time you add all the estimates together, I think the margin there may be too great to really give reliably.

I think it's fair to say that we sense that there is a gradual increase of the utilization rates. And that is good for us because that is exactly the type of thing that brings customers back to the table. Once they reach sort of the 80, 85 percent point, they will start talking to us with or without renewal timeframe.

Even order of magnitude -- are we at 70 percent, are we at 75 customer? Where are we.

The reason you would find that is a little puzzled is because there's a number of accounts where the customers are -- would not communicate any of that. But I think you're in general in the right ballpark.

Raj, let me just add one thing on the '05 number. Remember, the '05 number was -- actually was not couched as guidance; it was an assumed growth number off our original '04 bookings number. We started out at 1.45 and we said assume you have 6 percent growth. At the time we gave that estimate we thought it was reasonable, based on what we saw in the customer base in '04, and then continuing on into '05 in both the customer base and the environment. Clearly, we're short on '04, which is not a great base for jumping up to that number in '05.

Rich Valera, Needham.

Thank you. I would like to go back, and at the risk of beating a dead horse here, talk about your -- when you had strong renewal cycles, it sounds like you had the last three quarters have been weak quarters for you bookings-wise, certainly down substantially year over year. Presumably, the fourth quarter might be strong with a substantial renewal there, and then you're talking about four or five additional weak renewals quarters. So I'm trying to understand, I guess, when did you have the strong renewals quarters, other than a period in the first half of '03 when you did very well. You can just hop on that.

Fundamentally, the strong renewals were in '02 and '03. Typically, renewals are anchored in a quarter against one or two of the larger deals. And once you don't have that, then of course you need to do many much transactions from an order point of view. And again, orders are on a 3 to 3.5-year cycle, where the reality is one sense to renew much more often if one does it incrementally. Customers don't have to do that if they don't want to. They have often done it in the past. Our sense is that right now they're waiting things out a little bit.

Steve, can I assume you -- I think based on your comments -- that you won't be giving quarterly bookings data?

For next year?

For next year.

We'll report at the end of this year what happened (technical difficulty); for '05, I don't think we'll be giving quarterly bookings data, but we'll probably give some way to help you quarter-ize it.

To the degree you do, you sort of run the risk of being held up at the end of the quarter anyway, right? Isn't that sort of the issue you have always faced with giving (indiscernible)? So I'm just wondering how you could reconcile the issue of being held up at the end of the quarter, since that is the goal of transitioning to your new model?

I will take that as a suggestion.

Finally, just one comment on the earlier comment that -- moving away from an upfront model towards subscription helps preserve the (indiscernible) of your technology. Since subscription actually enables remixing, which in essence allows you to sort of give away future technology, while as upfront doesn't allow you to give away future technology, it seems that in essence you're doing the opposite there of actually moving to a model -- unless you actually curtail the amount of remixing you do. If you could just comment on that?

In a subscription you can limit the amount of new technology that customers have access to -- only a small amount, and they therefore get -- use it as a way to broaden out the use in their portfolio as a way for people to adopt new technology sooner and then drive incremental need as they use up all the technology they have access to. It's also possible to drive subscription-type treatment through payment terms as well, in which case you wouldn't give out any new technology.

Just one final one. Steve, in your closing comments you said something about 25 percent growth in '06? Could you just repeat that?

25 percent growth in '06.

Did you say something about 25 percent growth?

I said if you look at scheduled expirations and anticipated renewal cycle, plus some other reasonable assumptions that you have to put around it, we would expect from '05 to '06 25 percent orders growth.

(OPERATOR INSTRUCTIONS). Garo Toomajanian, RBC Capital Markets.

Aart, can you tell us if you're thinking about any changes in your product roadmap, or more or less emphasis on certain products as a result of either competitive pressures or the changes that you have seen in the marketplace?

If at all, we want to accelerate the roadmap that we are on, because we invested a lot in the integration and alignment and correlation among the products. The key right now is to take advantage of what we have just built, which is a much integrated solution that works well together and that's very complete. So from that perspective there are no fundamental changes, of course. We want to make very sure that we stay on the course and deliver what we have planned.

Okay. But would you see yourselves putting maybe more emphasis on some of those higher growth areas such as DSM and IP?

That's a good question. On the DSM side, we have -- we are engaged right now in a number of fairly big benchmarks and evaluations, largely because DSM decisions tend to be big for the purchaser and have a long time frame in terms of utilization. So as these materialize that is certainly an area that we could accelerate. On the IP side, we have just rolled out a number of new cores, specifically the Serial ATA, the high-speed on the go. And our objective will remain to (technical difficulty) portfolio as fast as we can, because we do see that there's good opportunities straight there.

Art, last quarter you said that you're seeing increased urgency of customers, and that is part of what has been driving some of the strength in services. Did that diminish in the quarter at all, and what does it mean for services going forward?

No, I cannot say that that diminished. I think customers are still very much in the mode to do the chips as fast as they possibly can. I do see a little bit of a shift in mood at the executive level about worrying now about the market rather than the technology. But when dealing with the engineering crews, they are on track to get to the market as fast as possible.

Dennis Wassung, Adams, Harkness & Hill.

A couple of quick questions. Steve, I think you alluded in your prepared remarks about the potential for some of the larger deals that you have expected in Q4 to slip into fiscal '05. I am just curious what is driving that. Is that just expectation of further push outs here and just further weakness in the market, or (indiscernible) those comments, please.

I think having made the shift now to a highly ratable model, where there is a large transaction that's very important to us, we don't want to be constrained necessarily by particular quarter boundaries if we don't have to be. So these kind of transactions are big and complex, and sometimes having a quarter end can be a compelling event that helps drive both sides. But in this case I think we would like to hold the line fairly strongly on what we expect to get in the way of value for the technology, and that might push it across the quarter boundary. I don't expect any of the deals in mind to go away entirely.

I guess sort of following onto that -- that point. From a competitive perspective, you are talking about not caving on price here, and how -- how was that affecting the competitive decision? If there is a deal that's going to happen and you are not going to cave on price, how is that -- how do you give us comfort that this isn't opening the door more to competitors, or you're not seeing more pressure from competitors? How is that -- I guess -- what is the dynamic there?

I think I would say we are seeing competitive pressure; it's a very competitive environment as we speak. At the same time, with customers you have no pricing power if you have no differentiation. And so if that is the situation -- and on top of that you're looking at the last, say, 10 hours in a quarter, that's when typically the negotiation is fairly one-sided. So from that perspective, we have, I think, two positives -- one is the move to the ratable model, which removes that imminent timing pressure; on the other hand is the very fact that the technology has gotten stronger, and that actually we can do quite well. But at end of the day, customers buy what is best and that is the way it should be.

The last one here. Talking about that there's a shift to a fully-ratable model here, up until just a couple of quarters ago you were in predominantly a ratable model, in that probably 80 percent-ish range. And then you reintroduced the term license about two quarters ago. Has there really been that much of a change just in the last two quarters of the customers' desire to go back toward term licenses, and now retracting back to more of a ratable payment model? Or, I guess, was that more of a Synopsys-driven event to try to pull some of this revenue forward in the last two quarters, and any other comments you have there? Thanks.

Dennis, I think the shift from -- we made earlier in the year was the shift away from perpetuals towards term licenses, both of which have upfront revenue, but a term license for us is a better outcome because it's finite and the customer has to come back to renew. What we have seen now going through the year with term licenses, and then also with perpetuals, to the extent we still have them -- we still had about an over $300 million target for upfront licenses. And it's the payment terms on both of them that became very difficult to attain with customers. I think that is what is driving the shift towards more time based or subscription-type licensing.

Rohit Pandey, Credit Suisse First Boston.

Thank you. Aart or Steve, what you're saying here is that what you're observing at the end customer is an industry-wide problem. So do you think it's the (indiscernible) your competitors, or are you saying that your guidance was too optimistic?

In hindsight, obviously, if we had seen the picture the way we see it now, we would not have given this guidance. But in hindsight there are a lot of things one does differently. Regarding the overall picture, I would be surprised if our customers didn't see some of the same things happening, because we all talk to the same customers, they exhibit the same behavior. And so over time, I think we will see.

And, Aart, in your 15, 20 years of experience in EDA, have you ever seen an inventory buildup in semiconductors impact EDA revenues?

Just to be clear, I don't think that inventory buildup impacts EDA at all. I think that what we are saying here is why are people suddenly cautious? My own sense is that the inventory questions are just that; they are the type of questions that come up when people become nervous and they start to watch more carefully, and for good reason because clearly in 2000 inventory was a big, big issue. A don't have the impression that it's as big an issue now. So don't read me wrong, I don't want to put any blame on anybody's inventory here. We misjudged how the quarter would come out. And in July things did turn significantly, but caution is absolutely the word.

Given that it's been better to (indiscernible) all of us to understand your fiscal '05 bookings guidance, two explanations (technical difficulty) our best explanation that perhaps you're anticipating market share (indiscernible) you tighten up the pricing, or you're trying to give (indiscernible) explanations -- but they're not easier to understand. Do you think you have a better explanation which we can understand?

Obviously, we've tried to explain it now a couple of times, and I can see that we're not totally successful. Let me try one more time, which is that -- we are looking at the picture from two or three perspectives. One is our technology; how are we stacking up competitively, and all the indicators are that certainly with the recent technology we are competing well; secondly, on the basis of what are the renewal cycles, and there heading down a very thorough analysis looking forward on our renewal cycles. As you can imagine, we have done a lot of work since the end of the quarter. We have concluded that the big renewals are all in the '06 timeframe, not in '05. And with the exception of one big one that sort of straddles '04 and '05, that has direct impact on the order level; and third, we do see a market that has turned more cautious, and therefore is unlikely to start renewing more ahead of time or in larger volume than what is strictly needed. In other words, we'll be driven much more by utilization rates. And although those are favorable to us, they are drivers that sort of change gradually.

Do you intend to keep the term licenses, or are you doing away with that?

I expect that we'll have some term licenses in the mix, Rohit.

For fiscal '05 cash flow, you said you would not give any guidance as of now?

At this point, I'm not going to give a preliminary number, no.

Erach Desai, American Technical Research.

I have -- I guess a lot of the questions have been asked, gentlemen. So let me -- two perspectives in terms of the bookings guidance for '05. Whether you want to look at it from a standpoint that in '00, '01 Synopsys stand-alone without Avanti was about a $900 million bookings level. And you don't seem to have -- and what you are guiding to for '05 is essentially no progress for the 900, 950 level (indiscernible) is essentially 4 or 500 million below what you were previously guiding. It's either Cadence is making inroads -- it doesn't look like that from their numbers; it's either Mentor (ph) is making inroads. I don't think that is true. Maybe Magma is a $500 million company. It doesn't seem like right away. So it's either competitive or it's either environmental, and I guess that's what we're all struggling with.

Erach, I can sympathize with that struggle because we have gone through exactly the same line of thinking. A couple of comments. First, I think one should not overstate the 900. You have variability among multiple years and some of that is unavoidable. I can see that 900 would not feel like we've made great progress, but the reality is we actually have made quite a bit of progress in terms of settling in on the market with an offering and a large backlog.

Secondly, regarding your competitive analysis, I think it reflects exactly what we see. I think Magma has gained some market share over the last few years, especially at the time where we were integrating the solutions. And I think we have also gotten some market share from Cadence. Our competitive position against Cadence feels very strong, and we actually hope for benchmarks every time we can get one because we'll do well. And with Mentor, the places that we touch are fairly limited; it's primarily in verification and in the DFM arena. In DFM we are both doing well; in the verification arena, I think there we see somewhat of a transition towards system (indiscernible). Both Mentor and us are very committed to that, and I think we're moving along well.

So I think those expenditures are still to come. So I think fundamentally we should not forget that the semiconductor industry went down dramatically, and initially EDA did not pay much of that pain. We probably see some of that today. But I also believe that we see a transition towards, increasingly, a solution. And we believe that our investments are all very conducive to a strong position as people move from point tools to more complete solutions. And that's what we are banking on going forward.

One sort of follow-on -- I don't want to put words in Jay's mouth, but I think the frustration perhaps that he might have been expressing -- I'm just going to use my words, which is -- given the low hurdle rate for turns, you have arguably had the most ratable business model in the EDA world, if not maybe the software world, for a company your size for sure. I think perhaps we're looking for some accountability in the executive sales management, and I think that is where perhaps some of Jay's comments might been directed. Again, that's my commentary. Steve, a question that nobody else cares about, but just from logistics. MoSys -- you paid them a $10 million breakup fee; where does that show up on the P&L?

It was in G&A and it was last quarter. I'm just checking -- yes.

It was the previous F2Q?

It would have been Fiscal Q2.

My mistake.

Erach, I just want to remind you for non-GAAP financials we took it out. So we took it out as an acquisition expense. It is in the GAAP numbers, though.

John Gray, Lehman Brothers.

Looking at the industry here, obviously, the semi cycle has done well in '03 and into '04 here. Aart, I would like to get your thoughts as we enter into a little bit more kind of troubling times, I guess, in the semi industry, if we truly are going into a downturn and R&D budgets start to get cut again, I would like to get your kind of expectation -- is this orders and revenue number for '05 kind of like a reasonable semi cycle, R&D budgets moving lower in '05? And your thoughts about the cycle here?

John, first, I don't want to be the person here calling the downturn of the semi cycle. What we're reporting is a degree of caution where people are looking forward with question marks. At the same time there are a number of people that are looking for the second half to still be okay from a semiconductor point of view. The caution has primarily to do with '05.

Secondly, if we are looking at our model going forward, the very statement I did make was that the ratable model plus the backlog position that we are in allows us to be very strong and weather any storm if there may be one. Again, I'm not saying there will be one; but if there is, we are in a solid position because customers are strongly committed; they need our tools, and we think we can do incremental sales on top of that. At the end of the day, though, it's clear that a strong semiconductor environment is (technical difficulty) and from a technology point of view, I have not seen any slowdown yet.

Thanks. Steve, as we go into this next quarter and also into next year, I'd like to get your thoughts kind of where we expect operating margins to go and when we would expect any OpEx cuts, in which quarters you're looking for that?

Operating margin for the year, John, will be in the 21 to 23 percent range. We will give guidance on '05 at the end of next quarter. We have given the revenue and the ultimate earnings, and there are a few things in between that could move it. So I think I won't give it on this call, though it will be substantially lower. I think you can count on that. As far as expense controls and when those might go into place, we're looking at those now and would expect through the quarter to make some decisions and have them fully effective by the beginning of Q1.

Operator, I think one last question.

Our last question comes from Jennifer Jordan with Wells Fargo Securities.

I don't want to beat a dead horse here because it seems like most of the questions have been answered, but one thing that occurs to me related to the renewal cycles is it seems like since we have gone to the subscription models, and in some cases longer contract durations, two things have happened. The first is that the fourth quarter has become more increasingly significant over time (indiscernible) all the players in the EDA industry. And the second thing that has happened is it may be that we've created greater annual cyclicality. And is there a way for us to get a handle on what that is going to look like going forward? You alluded to '06 being better, but it's really hard to understand how you couldn't anticipate that '05 wouldn't be, unless you were expecting to really pull a lot of renewals forward.

Jennifer, this is a very good observation that I have made for a while as well, and it starts very simply of orders moving towards the end of the quarter, because customers are trained to know that that is when you get the best negotiation position over your supplier -- if they are dependent on those orders. And then from the end of the quarter, they naturally start to migrate towards the end of the year, because there you have even more negotiation power and potentially (multiple speakers)

It's also a function of the way that you all book your quarters and then bundle additional contracts into those quarters; they tend to grow overtime.

So that is the third observation, which is that's not necessarily negative, because if you are a supplier with a broad portfolio and if you do believe -- and we do believe -- that customers are going to go increasingly towards complete solutions, naturally, things will tend to be bundled together in larger pools of technology. Now, to an earlier question, having subscription doesn't necessarily mean that you have remix everywhere. Remix is actually somewhat limited and is a function of contract negotiations. But what is desirable is to have broad agreements with the big customers, because those are the ones that anchor their commitment to you just as much as they anchor our commitment to them.

It raises two other questions. One is, is the fact that you have been doing these renewals that are three years long -- has that aggravated the situation of potentially having shelfware sitting out there that people are not utilizing? And the second question is are you seeing push back from people as they face their move to 90 to 65 to want to commit to shorter duration contracts rather than longer contracts? Because there's a substantial amount of risk in knowing whether what you've purchased is going to (indiscernible) process generation.

The problem with shelfware is only as much a problem as vendors start to give more and more software at lower and lower prices just to sort of get a deal. And we have definitely seen a few situations around us where people gave out free software just to be sort of in place when we were going to renew our solution. And so from our perspective, at the end of the day what matters most is really can you get your customer to be successful with the software? That's why we are watching utilization; that's why we are helping our customers be effective with the products, because (technical difficulty) they will want to select the product that they have been most successful with, notwithstanding whatever else shelfware may be there.

Finally, for Steve. Steve, you alluded to the idea that there are ways in which you can drive subscription type treatment of your term contracts through payments terms. That sounds just like plain old term licensing to me, not upfront term licensing. Are we going to see term licensing?

As I said, I think we could have term licenses in our mix. Some might be taken upfront and some might be taken overtime per payment terms.

Thank you very much. Thank you very much for attending this conference call. As usual, Steve and I will be available for follow-ups after this.

Thank you. Ladies and gentlemen, this conference will be available for replay after 5:30 PM today through midnight August 27, 2004. You may access the replay service by dialing 1-800-475-6701 and entering access code 741248. International participants, dial 320-365-3844. (OPERATOR INSTRUCTIONS). This concludes our conference for today.