Exhibit 99.2
Transcript of conference call on Magma’s Acquisition of Mojave
February 25, 2004
5:30 a.m. PST
Operator: | Good morning. My name is Ametrius and I will be your conference facilitator today. At this time I would like to welcome everyone to the Magma Mojave Acquisition 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 period. If you would like to ask a question during this time simply press “star” then the number 1 on your telephone keypad. If you would like to withdraw your question press “star” then the number 2 on your telephone keypad. Thank you. | ||
Mr. Lazich, you may begin your conference. | ||
Milan Lazich: | Good morning. This is Milan Lazich, vice president of corporate marketing for Magma Design Automation and I would like to welcome you to today’s conference call discussing Magma’s acquisition of Mojave. On the call today are Rajeev Madhavan, Magma’s chairman and chief executive officer; Roy Jewell, our president and chief operating officer; and Greg Walker, Magma’s chief financial officer. We will be discussing yesterday’s announcement and field your questions at the conclusion of our remarks. |
Now I would like to review the Safe Harbor guidelines for today’s conference call. During the course of this conference call, including the question-and-answer period, we make forward-looking statements that may include statements about the features and benefits of Magma systems and the complementarity of the Mojave technology with Magma’s system and about the impact of the acquisition on Magma’s future financial results. | ||
These forward-looking statements are subject to risks and uncertainties that could cause actual results to differ materially, including but not limited to: problems that could be encountered with the integration of Mojave’s technology into Magma’s system; the integration of its operations and personnel into Magma; Magma’s ability to achieve anticipated synergies; potential delays in the planned development of new products; the high degree of competition in our industry, and the introduction by our competitors of new products or services; lower than expected sales due to weakness in the semiconductor or electronic systems industries, economic recession or terrorism; and unexpected difficulties we may encounter in managing our expanding operations. We refer you to the full discussion of risks and uncertainties associated with the company’s business set forth in the company’s Form 10-K for the year ended March 31, 2003, its Form 10-Q for the quarter ended December 31, 2003 and in subsequently filed reports. These documents identify important factors that could cause the actual results to differ materially from those contained in our forward-looking statements. Magma does not undertake any obligation to update forward-looking statements made today. At this time I will turn the call over to Rajeev Madhavan, Magma’s chairman and chief executive officer. |
Rajeev Madhavan: | Thank you Milan. Good morning everyone. By now you’ve likely read yesterday’s announcement that we are acquiring Mojave Design, a company founded by some very talented EDA veterans who have delivered outstanding technology in the area of design for manufacturing. This transaction enables Magma to address new markets with growing potential and to do so with a technology that we think is superior to any offering available on the market today. I will take just a few minutes to discuss what this means for Magma and our customers. | |
Mojave brings two key technologies to Magma. First, a next-generation verification technology and second, design manufacturability technology. | ||
For this transaction Magma will add the Mojave’s verification and DFM technology to our design flow to offer the most efficient, cost-effective and sensible approach to nanometer design, design implementation matters, design verification, and design manufacturing. | ||
Our view of DFM, or design for manufacturability, is very different from the key – from the approaches taken by other companies. As we have said before, Magma believes that manufacturing issues should be anticipated while the chip is being designed rather than having to repair it after the design process. | ||
We disagree with the approach that has traditionally been represented as DFM, involving forced layout printability enhancement. Today’s designs, especially in 90 nanometers, 65 nanometers and smaller geometries, need a different approach. With those geometries certain non-manufacturable patents cannot be corrected even with post-processing. |
We believe design implementation should anticipate manufacturability issues and reduce or eliminate the need for major post-layout correction, and that’s what we believe our acquisition of Mojave will let us do. | ||
We believe DFM starts as the design is being implemented. Over a year ago Magma started implementing several key features that improve manufacturability of the chips designed with our software. Improving a chip’s manufacturability includes implementing characteristics such as die size, the chip’s total wiring, and the total number of vias – or the connection between layers. In many recent conversations with customers I have consistently heard of Blast Fusion enhancements in version 4.1 that helped them improve all these characteristics. It has provided die size reduction of up to 27 percent, reduced wiring by 22 percent and reduced via count by up to 32 percent. Several customers have leveraged this design advantage into a DFM benefit using our latest software releases. Now we are looking at extending that advantage further with Mojave. | ||
This transaction is an opportunity to acquire leading technology at an early stage, making it easier to combine Magma’s existing design flow. We have said in the past that we are very judicious in our M&A process because we seek to preserve our integrated design flow where integration is warranted, and in other cases we prefer to influence key technologists in the development of technologies. We believe we are in a unique position to deliver new leading technology. So getting good technology early, when it is easy to combine with our existing products, is consistent with our philosophy surrounding mergers and acquisitions. |
As to the market opportunity this opens up, the combination of Mojave’s technology with Magma’s will allow us to address new markets that we estimate represent at least $500 million in annual revenue today. The move to nano-scale designs will only increase the role of DFM technology and presents great opportunity for the Mojave technology. | ||||
Magma is recognized a leader of manufacturability solutions within the design frame that span from RTL through silicon. By combining DFM with IC implementation, Magma will be well positioned to address the next generation of design at 65 nanometers and below. | ||||
In looking at this transaction our management team is pleased that we are adding such highly regarded technologists and technology to Magma. We anticipate great success from this acquisition in large part because of the talent we will be adding to our current personnel and because we will be delivering technology that will have a bottom-line benefit to our customers. We are confident that this is a significant step towards our goal of reducing IC design and manufacturing costs for our customers. | ||||
At this time Greg Walker will make a few comments about the financial aspects of this transaction. Greg? | ||||
Greg Walker: | Thanks Rajeev, and good morning everyone. Financial details of this transaction are covered in the press release we issued Tuesday and in an 8-K document to be filed with the SEC after the deal is finalized. Today I can tell you that: | |||
• | The terms of the transaction have been fully agreed upon by both parties. |
• | We expect the transaction to close in the June quarter, which is the first quarter of our fiscal year 2005. It will therefore have no impact on the current quarter or fiscal 2004 results. | |||
• | The deal consideration will be a mix of cash and stock – 50 percent cash, 50 percent stock. | |||
• | The transaction includes an initial payment of $25 million, once again 50 percent stock 50 percent cash. Subsequent orders-based payout, earn-out payments, will become payable upon the achievement of specific technology deliverables on an agreed-upon schedule. This will total as much as $115 million over the next five years. | |||
• | We believe this transaction will be dilutive in the June quarter, the first quarter of our fiscal 2005, by about 2 cents a share. This is dependent on when during the quarter the deal will actually close. For the remaining three quarters of fiscal year 2005 we expect the impact to be dilutive by about 1 cent per share during each quarter. Beginning in the first quarter of our fiscal 2006 we expect the impact to be neutral and then accretive thereafter. | |||
As to the structure of the deal, it’s worth noting that the bulk of the payments we are making are contingent on future deliverables in order of performance. For this reason we are satisfied that we have made a reasonable investment in technology that will enable us to address significant additional markets, while at the same time mitigating much of the deal’s risk. | ||||
At this time we’ll open up the call for your questions. Thank you. | ||||
Operator: | At this time I would like to remind everyone in order to ask a question, please press “star” then the number 1 on your telephone keypad. We’ll pause for just a moment to compile the Q&A roster. Your first question comes from Jay Vleeschhouwer with Merrill Lynch. |
Jay Vleeschhouwer: | Morning. Rajeev, could you be a bit more specific on the technology deliverables that you’re expecting out of Mojave over the next couple of years? Also you expressed a different view of DFM, which is fine of course, but you’ve had a reasonably productive working relationship with Mentor in DFM-type engagements thus far. So how do you see that relationship potentially changing as a result of this? | |
Rajeev Madhavan: | First let me start with the latter question about Mentor Graphics and the relationship. We do not see this transaction as having any effect whatsoever on our relationship with Mentor Graphics. It continues and we expect to provide solutions so which we address our customer’s benefits as we proceed now. | |
With respect to Mojave and what they’re doing, again we are not in a position to answer exactly the technology deliverables but it’s fair to say that this calendar year – by the end of this calendar year we expect to announce technologies based on (unintelligible) as well as research and development projects that we have being working on within Magma. The first phase of it will be accomplished this calendar year and there will be subsequent releases obviously that we have targeted for. | ||
But it will, as we have said, redefine the design for manufacturability to include manufacturing verification and implementation aspects of things within a comprehensive solution, a comprehensive flow that we can provide to our customer space. |
Now we do not think this has any effect whatsoever, again to the second portion of your question, to the Mentor Graphics portion of the technology that you raised here. | ||
Jay Vleeschhouwer: | OK. A follow up for you Rajeev. As you know it often takes years for new, perhaps even superior, technology to establish itself in the marketplace and displace existing technology. And you saw that yourselves with Blast Fusion. So how do you foresee the adoption or displacement or penetration of this new approach or technology developing over the years given, again, the establishment of existing solutions? | |
Rajeev Madhavan: | OK, EDA is all about technologists and which technologists have the ability to deliver those technologies. I mean, yes, there is a gestation period in adoption of any new technology in this space. However the quality of what you can deliver and the quality of how you can back it up is dependent on the technologists. | |
We are in a very unique position with Mojave with three great technologists there, and a whole bunch of others, and the whole team which is very, very solid in having delivered a multitude of EDA technology which have been accepted and have become industry standards, and have gone on to become industry standards. | ||
And it is fair to say that yes, there will be issues and there will be, you know, time to reach the level of market that Blast Fusion has achieved today. But also remember that Magma has a better sales channel today than it did when we started out with Blast Fusion. We had one or two channels and within that channel we achieved substantial bookings for a new product in the first year with Blast Fusion. |
So I think, you know, the prospect for this is even higher. We are more optimistic than that. And we continue to believe that our channel is in a position to actually be able to demonstrate these technologies to a multitude of customers today than it was when we were launching Blast Fusion. | ||
Jay Vleeschhouwer: | OK, just two last ones. You highlighted the importance of manufacturing costs and yield enhancements and so forth. Is there going to be some way where users of the new system, once it’s completed, can predict the return on investment or cost or yield impact as early in the process as you are anticipating with this new technology? In other words, is there some actual ROI impact that you will be able to more accurately offer – show customers with this? | |
And then lastly for Greg, is it accurate that shortly after Mojave was founded when that team left Synopsys/Avant! that you had put – that is Magma had made – some investment in the company? | ||
Rajeev Madhavan: | No, there’s no investments in Mojave from Magma or anybody, any of the execs at Magma. That answers the second question there. | |
Let me talk about the first question, which is getting an idea of the yields very early on in the flow. Absolutely, “yes” is the answer to it. We as a company are a believer of giving early feedback. You know our early silicon predictability with respect to timing closure has been a very important piece for Magma and we continue to enhance that to add other key factors in this particular case (unintelligible). | ||
Jay Vleeschhouwer: | Thanks Rajeev. |
Operator: | Your next question comes from Rohit Pandey, Credit Suisse First Boston. | |
Rohit Pandey: | Hi Rajeev. Does the company have a beta (unintelligible) or something it acquired, or how far away are you from a product? | |
Rajeev Madhavan: | We do not want to comment on exactly when the products will come out. As I said we expect to have a product in production this calendar year. | |
Rohit Pandey: | But did Mojave have a beta at least, ever? | |
Greg Walker: | No they are not – this is Greg – they’re not at beta yet. We’re expecting betas probably some time in the fall. | |
Rohit Pandey: | OK. Yeah, that’s all I had. Thank you. | |
Operator: | Your next question comes from Bill Frerichs with the D.A. Davidson & Company. | |
Bill Frerichs: | OK, Rajeev to go back to Jay’s first question about the relationship with Mentor. When these guys started the company they advertised it as the “Calibre Killer,” possibly because the market penetration of Calibre. Now that plan has changed. Can you confirm that this is not a design rule checking tool? | |
Rajeev Madhavan: | Bill, again I want to reiterate. Our view of DRC, LVS, etc. is, as we pointed out – DFM in our view will encompass a number of things. It will encompass implementation, it will encompass verification, it is going to redefine the reticle space, the manufacturing space of manufacturing all the way to the implementation space and verification space. So in a unique way what we are doing is a combination of technology. |
Now you can say that these kinds of technologies do not affect DRC, LVS, but that is not the case. This is all a revolutionary way of launching technology. When Magma came out with (unintelligible) it did have an impact on some businesses, yes longer term it has an impact on some businesses. | ||
(Unintelligible), we are dramatically changing the way – it is not going to be a DRC tool, it’s not going to be an LVS tool, it’s not going to be a manufacturing tool, it’s not going to be just an implementation tool. It’s the concoction of all of those to form a flow that we are talking about. | ||
So in that sense does it compete? Yes. But it’s a very radically different way of doing the products than what is out there today. And so we are taking a very comprehensive approach on solving the solution and not just thinking of it as a design rules checker. | ||
Bill Frerichs: | OK. And then to go on, you mentioned in your opening remarks that future versions would have an IP implementation aspect to them. Is this related to your effort with PDF solutions to get some yield learning information into the higher part of the design flow? | |
Rajeev Madhavan: | No, you know, I don’t recall and I don’t think I made a comment on anything related to IP here. We have said in the press release for PDF about us working together with PDF but no, today’s press release and today’s, or yesterday’s, announcement are all not really connected with what we are doing with PDF today. |
Bill Frerichs: | Are these – are they parallel tracks? | |
Rajeev Madhavan: | Well, you know, there are very different teams within Magma working on these things and so we have a relationship of working with them and we have a relationship as I said early on working with Mentor Graphics and, you know, mutual customers drive the need for us to work with these third-party vendors. We are absolutely happy to do so. | |
Bill Frerichs: | OK, great. Thanks very much. | |
Operator: | Your next question comes from Garo Toomajanian with RBC Capital Markets. | |
Garo Toomajanian: | Thank you. Rajeev, you have talked about in the past all the benefits of the unified data model and I’m wondering if there are any risks perceived or otherwise in having verification technology that would be based on the same data model that is used for the design implementation. | |
Rajeev Madhavan: | First of all Garo, and I want to – I have not made the comment that this particular needs to be absolutely integrated with the data model. We have not made such decisions at this juncture. You know, we will comment on it as the products evolve. So let me just leave it at that and we will see when the products come out how we (unintelligible) solutions. | |
Garo Toomajanian: | OK, thank you. | |
Operator: | Your next question comes from Erach Desai with American Tech Research. |
Erach Desai: | Hi. I guess the first question – I don’t know if Greg alluded to that or said that and I joined late – was the milestones that you guys have or for the earnout, does that include bookings? Did I catch that at all? | |
Greg Walker: | It’s actually almost like a double trigger, Erach. There are technology milestones that have to be achieved. If those are achieved, then the earnouts are based on the orders levels achieved subsequent to that. If you don’t deliver the technology even if you make the earn out, the orders are still – the earn outs are not earned. So it’s a double trigger. | |
Erach Desai: | OK, that makes sense. And, you know, as much as Rajeev’s going to not love this, but I want to circle back to the Mentor relationship issue because I think for the last couple of quarters, whether explicit executive endorsement or sort of people flying by the seat of their pants in the field, but effectively the Magma and Mentor combination for the physical design and physical verification flow was being pitched. And some people have been successful with that. And I frankly just don’t see how Mentor, if I’m sitting in Mentor’s shoes, can look at this as not being competitive down the pike. | |
Rajeev Madhavan: | Let me, I mean, I want to reiterate what I said. We believe Mentor will continue to – especially Calibre an outstanding tool, an outstanding tool that will continue to be a significant player if not dominate the traditional DRC LVS market. And we continue to do – to partner with Mentor Graphics. | |
Now I want to reiterate one more thing about Mentor Graphics. We actually do work with Mentor [on] more than DRC and LVS. We are working relationships in the area of test automation, we are working relationships in the area of mix signals, we do continuously work together in solving our customers’ needs. |
I mean, it is – we have actually affected, both of us together attacked common customers and made sure that they have an entire Magma/Mentor flow. That has been achieved in the last quarter at a couple of sales sites and we are very happy that has occurred and we will continue to support that. We will continue to enhance that. Both teams will continue to market and solve that. | ||
Again, I want to very much, you know, be very blunt. We believe Calibre will dominate the traditional DRC LVS market and we will continue to partner with Mentor Graphics. | ||
Erach Desai: | OK. | |
Operator: | Your next question comes from Dennis Wassung with Adams, Harkness. | |
Dennis Wassung: | Thank you. A couple of quick questions. First, and I think Garo sort of asked this question differently, and it sounds like it’s still early to decide, but I’m just curious as to how this affects sort of your single executable philosophy that you’ve had up until now. Obviously if you’re not quite sure on the single data model this may be sort of a moot question. But any comments there? | |
Rajeev Madhavan: | You know, I’m not at a position of wanting to talk about the technology in this particular case. We have a fairly good idea of how we are going to integrate this technology but at this juncture it is too early to be talking about that. Once the transaction is completed and we get closer to a release cycle we will update you about the products. But we have a good idea of what the functionality is that we are going to be integrating and what is it that we don’t have to integrate. And you will see that in (unintelligible). |
Greg Walker: | Yeah, there will probably be a mix of both. | |
Rajeev Madhavan: | It is fair to say that even in our acquisitions like Silicon Metrics that we did we talked about the need for no integration, we talked about CSI where we said we’re bringing in the people and making them write on the Magma data model. In this particular case there is a mix and so we will comment on that as the time evolves. | |
Dennis Wassung: | OK, fair enough. And last question – you talked about a $500 million-plus sort of addressable market that this sort of breaks you into. I’m just curious if you could break that down a little bit more. I’m assuming that’s things like OPC, resolution enhancement, physical verification. Anything else I’m not thinking of there? | |
Rajeev Madhavan: | No I think you nailed a few things exactly there. It’s got, you know, basically it’s based on numbers that our competition has said in some of those spaces that they believe that’s where the market is. You combine all the three things that you just said, that’s more than $500 million there. | |
Dennis Wassung: | Very good, thank you. | |
Rajeev Madhavan: | Thank you. | |
Operator: | Your next question comes from Steve Rosston with Glynn Capital Management. |
Steve Rosston: | Yes please, Rajeev. A couple of questions. One is, can you talk about the integration of the teams and how they’re going to report through the Magma structure and the roles of the people? And then I have a follow-up question after that. | |
Rajeev Madhavan: | Let’s let Roy have the answer to the question here. Roy? | |
Roy Jewell: | Hi Steve. I think that what we’ll do is, anything that ends up being integrated as part of the implementation flow will report in to the current management team that we have today. The base technology that Mojave is bringing in here will be maintained as a separate business unit where they’ll have control over the development and deployment of that product through our channels. | |
So I think you can look at it as we’re going to do nothing to compromise the delivery of the implementation flow. That will still be maintained under the current management team and the new group coming in here will be maintained as a separate business unit so they’ll have control over their future. | ||
Steve Rosston: | And you’ve worked on all that reporting structure and who goes where at this point in time, no doubt. | |
Roy Jewell: | Yes. | |
Steve Rosston: | Thank you very much, I’m all set. | |
Operator: | Your next question comes from John Barr with Buckingham Capital. |
John Barr: | I just wondered, Greg, if you could just clarify on the guidance. So if you’re talking about 2 cents in the June quarter, it sounds like there is some one-time transaction-related expenses in there. | |
Greg Walker: | Well, it’s a couple of things. One is it’s a little unpredictable right now because the actual stock price that will go into the calculations for the share count is determined based on trading averages leading up to closing. So we don’t really know exactly how many shares are going to be involved and that will swing the dilution a little bit. | |
When we look at the ongoing expenses of the group plus the dilutive effects, it’s right in a rounding between 1 and 2 cents a share and we think that the biggest impact is going to be in the June quarter as far as that and that will be about 2 cents. And then going forward it actually rounds down to about 1 cent. So it’s right on the borderline. Our objective for Roy and I from an operational standpoint will be to manage this after the June quarter such that we see the impact (unintelligible). | ||
John Barr: | And how many people are you bringing on that we should add into the expense model, or how much should we increase our expenses? | |
Greg Walker: | Right now the team is between eight and 10 people, and so that’s a good starting point. And then there will be some growth to that team as we progress over the next year or so. | |
John Barr: | OK. And then just for Rajeev, the division here, what you’re looking to accomplish and then trying to understand the impact on current DRC and reticle enhancement markets, it sounds as if you’re trying to move things to correct by construction, or “design it right and not have to do as much fixing after the fact.” Is that a simplistic way of looking at it? |
Rajeev Madhavan: | It’s a very simplistic way but it’s very accurate in that we believe in the philosophy that it’s got to be correct by construction. With the chips that we’re talking about it’s impossible to correct it as a post-layout reticle enhancement or printability enhancement techniques. That’s way too late to catch any of the issues. | |
Hence this is basically a correct-by-construction methodology that Magma has embarked on for timing and (unintelligible) and other issues being standard. | ||
John Barr: | OK, great. Thank you. | |
Operator: | At this time I would like to remind everyone, in order to ask a question please press “star” then the number 1 on your telephone keypad. | |
Your next question comes from Jennifer Jordan with Wells Fargo. | ||
Jennifer Jordan: | Yes, good job gentlemen on this acquisition. It seems that most of my questions have already been answered and John asked the one that I was most interested in. | |
So my understanding is that this is a product or a set of products that you’re going to use as you go through the implementation flow so that people can as they’re doing their layout and placement make decisions that relate to yield and manufacturing rules as well as RET and therefore allow you to have fewer corrections as you go down the line. | ||
Rajeev Madhavan: | Yes. |
Greg Walker: | That’s right on it, Jennifer. | |
Jennifer Jordan: | OK, and Greg, with regard again to the comment you just made about you and Roy and the operational management, in the past with your acquisitions you’ve been able to manage to – that by the time the acquisition closed you were not – it wasn’t necessary to lower your guidance. So I assume that’s still your goal here. | |
Greg Walker: | Yes, that’s still our goal. | |
Jennifer Jordan: | OK, great. Thank you. | |
Operator: | Your next question comes from Jay Vleeschhouwer with Merrill Lynch. | |
Jay Vleeschhouwer: | Thanks, just a follow-up. I just wanted to verify a comment that was made about Magma at date last week. Rajeev, is it accurate that your current installed base of licenses for Blast Fusion is approximately 1,500? Is your base now that large? | |
Rajeev Madhavan: | I actually don’t have an accurate number for you here Jay but we’ll look it up and give the data offline. We could take it – but that probably is around right but we just – I can’t confirm that. | |
Jay Vleeschhouwer: | That’s fine, we can do it offline. | |
Greg Walker: | This is Greg. I think the last time we talked about it I think we might have said 1,600, but let us verify. |
Jay Vleeschhouwer: | OK, that’s fine. Secondly, can you perhaps discuss the acquisition and your plans for DFM in the context of what your current largest customers are doing with respect to DFM? For example, Infinion, TI, others – what are they doing today? And give us perhaps an example of how they might be able to incorporate or move down this path that you are describing, given that each of them is among your largest customers. | |
Rajeev Madhavan: | Yeah, I think we see certainly our largest customers – the issues of doing post-layout corrections. And for example things like OPC and phase shift has been solved to a certain degree by over-design. What I mean by over-design is basically making your (unintelligible) cells bigger, padding it in such a fashion that when two cells are placed next to each there is no phase shift issues coming in. | |
I mean, you can always over-design and let – the benefits of (unintelligible) in technology that you get, which is sort of what we observe as a solution today for a lot of the 90-nanometer nodes that are coming up and running. | ||
But as we are involved in some of the 65-nanometer processes, there is a request from most of these customers that such techniques – which are over-designed with respect to power, over-designed with respect to cells, over-designed with the whole implementation – has to stop. Because if you continue to do that the gains achieved by going down to deeper and deeper geometries is lost by having to do more padding in the design process. | ||
We think these things – by moving it upwards we can actually be effective in taking away the guard banding that is done in order to protect the things that never happen or happen once or twice. And so we’re just solving for that. Today is an over-defensive tactic which is applied. |
So what we are gaining in our experience in 90 nanometer in bringing up the cell nodes that we’ve done is very, very interesting, and now we’re learning the approach right now in solutions in many of these places are brute force. And from that brute force we need to move towards something a lot more intelligent to allow for more benefits to come from the shrinking to deeper and deeper geometries. | ||
Jay Vleeschhouwer: | It seems that much of the impetus for adoption in the current DFM and DRC and OPC markets and all those other acronyms has been the endorsement or use by the foundries, TSMC and so forth. That seems to be an important part of the strength for example that Mentor has in the marketplace. So how do you see your technology strategy here ultimately playing into that? I mean, how important is foundry adoption, the role of foundries in production, particularly in Asia for the success of your strategy? | |
Rajeev Madhavan: | I think in their adoption of foundries of Blast Fusion and all the solutions that we have done and will continue to be coming out that it’s a key factor in our success. There’s no doubt that is an essential portion of it. | |
But remember, some of these issues is going to be controlled by the designers and not just the fabs going forward, because having that knowledge of manufacturability is going to be essential for you to make a decision of whether it is going to be economically feasible or not. | ||
Today, you know, to a certain degree that has been shielded from the designers. As time proceeds that decision will also be one of the factors that the designers need to make in allocating what the foundries are. |
So Magma as a company obviously will continue to work with the foundries as well as have to work with the customers, and it is basically the two of them driving each other, and in between Magma, to enhance our software that makes all of this possible. | ||
Jay Vleeschhouwer: | And do you foresee that at some point there would be some kind of reference flow, methodology or endorsements, the kind of thing we often see press releases about from EDA vendors, would pertain to what you’re trying to do here? | |
Rajeev Madhavan: | I’m thinking that there will be that but I think that will be quite a time away because it takes time as you – I’m referring you back to your first comment, it takes time to adopt these technologies. | |
For example, Blast Fusion as you’ve been seeing a stream of reference flows, announcements in the last three or four quarters – that has happened while it took two years of adoption before that achieved. | ||
So, but, that is not essential for you to make sales, because sales are done at the customer’s site as well. That just enables you to actually be able to accelerate your sales in a much faster fashion. So I think the same period of gestation and same period of adoption goes through every tool. Mojave will be no exception. | ||
Roy Jewell: | OK, this is Roy. Let me say one thing. We talked about the time it took for Blast Fusion to become adopted in the industry. I think that we have some advantages here because Magma has developed credibility in the semiconductor industry of delivering the leading implementation flow for next-generation process technology. So there’s an expectation on the customer base, in the customer base, that what we’re delivering will be advanced. |
The second thing relative to working with the foundries is this team has a history of doing this at Avant!, before that with ISS, so that they have got a history and a credibility with the foundries already in terms of building a tool that gets adopted. | ||
And then finally if you look at what we’re doing here, just as Magma pioneered the concept of correct by construction and moving analysis in as an integral part of the implementation flow, we’re now looking at design for manufacturability, and it’s been relegated in the past to think like OPC, which by name is Optical Proximity Correction. It’s an after-the-fact correction of deficiencies in the manufacturing technology. | ||
We’re again moving that up into the implementation flow where it needs to be addressed early on and not a correct-by-repair type of mentality. So I think this is very consistent with what this company has done since its inception. | ||
Jay Vleeschhouwer: | Thanks, Roy. | |
Operator: | Your next question comes from Rich Valera with Needham & Company. | |
Rich Valera: | Thank you. I just wanted to circle back on the DRC LVS questions that have been asked. I just wanted to be clear. Do you see a role for DRC LVS tools as they exist today, say two or three years down the road? Because it sounds like you really plan to subsume a lot of that functionality into what you’re doing with Mojave. So I just wanted to get your thoughts on that. |
Rajeev Madhavan: | So let me answer that question, Rich. Even today while some of our competition is selling tools such as Silicon Ensemble and tools which are standard placers and standard routers—you know, part of our competition is a router, a placer, and a technology – even today they are being able to stay in the market and compete. | |
Yes, technology wise it may now be a leadership position. So it takes two to three years. DRC isn’t going anywhere. DRC will be there. LVS will continue to be there. I think for the foreseeable future it will be there. It’s a question of how all this affects – we are not clear about saying whether DRC will be eliminated entirely or not. | ||
But it’s fair to say DRC will grow more, implementation will grow more, and this assumes what and how is going to be a very interesting thing to watch. | ||
Rich Valera: | How about just specifically, understanding, you know – legacy technology is going to stay around for a while. But specifically within the Magma flow, if the leading-edge Magma flow, what’s your view? You know, as opposed to what the industry might sort of … | |
Rajeev Madhavan: | Our view historically has been, and this continues to be the case, that if a chip is done and at the end of the chip there is some timing violations that we leave, or for that matter DRC violations that we leave, which is correctable, which is not because your tight constraints cannot be met. | |
But if it is correctable and which is actually, you know, you can detect it and you cannot detect it in Magma but you can detect it in another system, we treat it as a bug. We treat it as a bug for DRC LVS. If you find that we can – our system does not detect it and someone else’s verification tools detects it, we encourage our customers to file upon which we treat as a very, very (unintelligible) bug. |
So that has been the thing about this company from day one. And that is not today’s issue. It has been from day one that we treat at the DRC LVS that we cannot detect has been detected by (unintelligible) tool, there’s a bug in our system. | ||
Now that does – does that mean that people will not run DRC LVS? No, I would encourage them to run DRC LVS because the cost of creating a reticle is such that it’s very high. You need to run it in order to verify that you are now ready to spend the $1 million or $1/2 million dollar to go to reticle. | ||
Rich Valera: | That’s – you’re sort of implying that you don’t plan to develop functionality specifically like the current DRC LVS functionality. | |
Rajeev Madhavan: | I am not implying anything of that nature. I think our product specifications will come out as it gets all out and from such point. And I’m not going to give that – I’m just talking more about the philosophy related to the acquisition. It doesn’t mean anything with respect to we won’t develop a particular functionality or we will develop a particular functionality. | |
Rich Valera: | OK, fair enough. Thanks a lot Rajeev. | |
Operator: | Your next question comes from Jennifer Jordan with Wells Fargo. | |
Jennifer Jordan: | Yeah, Rajeev, just a follow-up on that thought. It seems to me that – or is it safe to say that there is a certain degree of unknown about what the best |
configuration will be because as you continue to move down process geometries, the complexity of what you’re trying to solve increases the gap between the steppers and the design feature size continues to increase so that you may end up finding that it’s all able to be correct by construction or you may end up finding that you have to continue to do both stages? | ||
Rajeev Madhavan: | I think in the short term both phases will be required but within the – we do not see anything that prevents more and more of automatic correction to be done during the design frame, and we do not think that there is a reason why you have to do post-layout manipulation. We think that is not necessary going forward. We think that is an artifact of how people have done tools in the past and has nothing to do with how tools should be done going forward. | |
Jennifer Jordan: | So and that – another way to look at it then is to say that this is an approach that allows you to work with the existing methodology until such a methodology is ready to change. | |
Rajeev Madhavan: | That is right. | |
Jennifer Jordan: | Rather than trying to butt heads with it immediately. | |
Rajeev Madhavan: | I don’t think, you know, no new technology that comes in eliminates and I’m saying that DRC LVS is here to stay, it will be here to stay for a long time. I mean, the cost of making these chips are such that independent checking is essential. |
I mean, Magma will tell you today that we are functionally right. Yet every one of our customers will run formal verification or simulate the design at the end to make sure it is correct logically too. | ||
Even if that step doesn’t catch any problems, it has to be done because the cost of these things are – it’s like buying an insurance policy. And you do need to get an insurance policy running something at the end to ensure that it is correct and that Magma, what it sees is indeed correct. | ||
So I do not see that – DRC LVS will not as a functionality cease to exist. How it all gets subsumed and (unintelligible) subsumed etc., market and time will tell. | ||
Jennifer Jordan: | So – more of the way it gets used. | |
Rajeev Madhavan: | Yes. | |
Jennifer Jordan: | OK. | |
Greg Walker: | One way to think about it is we’re not out to eliminate the checking entirely. We’re out to eliminate the corrections. | |
Jennifer Jordan: | Uh-huh. | |
Operator: | At this time there are no further questions. | |
Rajeev Madhavan: | Thank you. If we have answered all your questions, we’ll end the call at this time. Thank you for joining us today. | |
Operator: | Thank you. And thank everyone for participating in today’s Magma Mojave Acquisition conference call. You may now all disconnect. |
END