[ Back ]   [ More News ]   [ Home ]
January 08, 2007
Changes at MoSys
Please note that contributed articles, blog entries, and comments posted on EDACafe.com are the views and opinion of the author and do not necessarily represent the views and opinions of the management and staff of Internet Business Systems and its subsidiary web-sites.
Jack Horgan - Contributing Editor


by Jack Horgan - Contributing Editor
Posted anew every four weeks or so, the EDA WEEKLY delivers to its readers information concerning the latest happenings in the EDA industry, covering vendors, products, finances and new developments. Frequently, feature articles on selected public or private EDA companies are presented. Brought to you by EDACafe.com. If we miss a story or subject that you feel deserves to be included, or you just want to suggest a future topic, please contact us! Questions? Feedback? Click here. Thank you!

src="file:///C:/WINDOWS/TEMP/msoclip1/01/clip_image001.jpg" mce_src="file:///C:/WINDOWS/TEMP/msoclip1/01/clip_image001.jpg">

Introduction
In May 2005 the company changed its name from Monolithic System Technology, Inc to MoSys, Inc.  At the end of July 2005 the company announced the appointment of Chet Silvestri as chief executive officer.  Silvestri had previously been CEO at CEVA, CEO at Arcot Systems, Inc., COO of Tripath Technology, Inc. and president of the Microelectronic Division of SUN Microsystems, Inc.  Since then he
has recruited new VPs of finance, engineering, worldwide sales and most recently in mid October Raj Singh as its Vice President of Marketing and Business Development.  I had a chance to talk with Raj Singh recently.


Would you give us a brief biography?

Right before I joined MoSys I was a consultant for a while.  I was also involved with an EDA company that was doing C to RTL tools called Synfora.  Immediately prior to that, I was VP of Worldwide Sales for Virage Logic for just over three years from 2002 to 2005.  Prior to that I was founder and executive GM of a semiconductor company called 3D Labs.  I started it in 1994 and took it public in 1996.  Eventually it got
sold to Creative Technologies of Singapore in 2002.  Prior to that experience I spent six years working for Dupont in a variety of management positions.  Hopefully that gives you enough flavor.


I see you are a graduate of the University of Scotland with a degree in English Literature.  How did you end up in EDA in Silicon Valley?

That was a long and tortuous road.  I went into Public Relations and that PR company happened to have some technology clients.  I slowly drifted into the technology end of the business.  That’s the short version.  There is a longer version we can discuss over a glass of wine at some point.


According to the press announcement you are VP of Marketing and Business Development.  How would you describe the role of that position?

This company or in general?


Both!

In general and in the particular of this company Marketing certainly is the pull portion and the public face of the company as well.  The sales component tends to be more of a push.  Marketing is more demand creation, the pull, the visibility.  Other than all the day-to-day activities of product sheets, sales collaterals and sales force leads there is the more important business of Marketing of actually creating demand for the product and building an ecosystem that the product can live and
thrive in.


How does one build demand for a product?

In several ways.  Let me answer that in the particular rather than the general.  Specifically in MoSys’ case there does exist demand and a very good demand for the product.  In many cases what has happened in the past, the visibility of the company or the reach of the company has been limited and that has led people not to be aware of the potential of the technology.  Part of my role is to provide increased visibility for the company such that
more people are aware of what is possible with the product.


How do you increase visibility?

Currently we are evaluating various things including seminars, more press coverage, more outreach to existing and potential customers, partnerships and building an ecosystem.


What do you mean by an ecosystem?


Since most people who do SoC designs use multiple technologies and multiple pieces of IP both internal and external, having partnerships with people where there is a symbiotic relationship between our technology and their technology and where there can be optimized versions for certain kinds of interfaces I think helps both the designer and our eventual end customer.


How do you go about creating partnerships?


At a first pass you have to identify who would be an appropriate partner.  Then talk with them to build a relationship which is mutual beneficial.  Partnerships that are not mutual beneficial don’t last very long.


The major EDA companies have extensive partnership programs.  They bring to the table large install base, marketing clout, .. If you want to form a partnership with these companies, how do you persuade them that it will be mutually beneficial?

It is hard to answer that question in general terms.  If you were to take as an example, pick a large EDA company (there are only two or three such companies).  If you wanted a partnership with them, we have more to gain than they have.  You have to come up with what they have to gain.  Is it that their customers can design chips easier, with better power, and with whatever is the particular benefit happens to be for that particular partner? style="mso-spacerun: yes">  We have to investigate and prove that to them.


Does everything boil down to technology in the final analysis?

No!  I wish.  In fact it is rarely so.  I was having this conversation the other day with a dyed in the wool technologist who said that it was all about the technology.  I said that if that were the case, we would all be using some form of CPM today and not Microsoft Windows.  Rarely does the best technology win out against all comers.  There are multiple versions
of things where late comers have become more established in the market than the pioneers who end up with arrows in their back.  It’s certainly not the case.  I think an optimal solution delivered at the right time crisply executed and supported correctly is the one likely to succeed.  It may or may not be the best technology.  We would hope that is was the best technology.  But technology by itself will not win the game.


You have been around this industry for a long time.  Over the last half dozen years, what has changed in the EDA arena and how does this affect the role of Marketing?

One interesting thing that has happened is that the number of new design starts is actually reducing.  I think that this is causing some interesting ripples in the business because as the number of design starts reduces, the amount of tools that are purchased may change as well.  And that will probably change.  If you look at that and say that the number of design starts are reducing, designs are more complex, certainly advanced technology are getting more and more complex.  There is a need for tools that didn’t exist before and a need for innovation that didn’t exist before.  As an IP provider we look at people doing design starts.  It terms out that the amount of memory being used (MoSys is inherently a memory company) is actually increasing.  It is getting bigger, it is getting denser.  This is a good thing for us.  The stage we find ourselves in, the design starts getting more complex is a
good thing.  The fact that they are reducing is a more difficult situation for mainstream EDA vendors than it is for us.


MoSys is an IP company.  Taking terms from well known books, has IP reached a tipping point?  Has it cross the Chasm?

Laugh.  There are different kinds of IP of course.  In the memory IP space, I think it has crossed the chasm.


How do you reach that conclusion?

The number of people including IDMS who are doing their own memory design has reduced dramatically in the last five years, particularly in dense memory design.


What is the reasons for that?

Several reasons.  One, it became really hard for single entities to be able to invest in the infrastructure of memory design and be able to get good ROI on that.  When generic products could be purchased an IP vendor who could defray the cost over multiple sales, the ROI just didn’t make any sense.  You have seen that happen in multiple places.  When I was aboard we had people like Freescale approaching us saying can you do this for us because we
can’t possibly do that for the same price you can it to us because we are going to do that once and you would sell that to 10 people, to 20 people.


How do companies like Freescale overcome their not-invented-here attitude?

What tends to happen is that there is a class of products that they will go outside for and there will be some projects that they will carry on doing which are unique to that particular IDM or semiconductor company and gives them an edge over their competition in that particular marketplace.


Nothing in life is risk free.

Correct.


If a company brings in IP from the outside, it will hopefully reduce development cost but there may be misinterpretations of specs or other problems.  How does a company like MoSys convince prospects that those problems are manageable and that they will help users overcome them when they arise?

Two answers to that.  One is that the sorts of business relationships we do are somewhat different from other memory IP companies.  We have essentially two streams to our business.  One of which is a marco program.  You come along and buy a block from us.  That block comes with silicon report and silicon verification data.  You have some comfort that at the process node and at the flavor you are going to use that it has been tested and it works.  We have an extensive SE post-sales network that insures the integration of that block.  That’s one part.  The other part is that we have a number of IDM customers who are technology licenses where they actually get the technology and some sample macros.  They are then able to use their people to do their own design internally.  The support mechanism for that is extensive. style="mso-spacerun: yes">  It is public knowledge from press releases for example that Nintendo and NEC are technology licensees of MoSys.  That’s how that relationship works and it works very effectively now because they can see what they are doing.  They can see the circuits.  They can see how it is done.  They can work with us.  We have engineers on staff working with them to get the finished product to the chip.


Do you have a sense of how the business is split between these two categories?

I do.  But I don’t think that it is public knowledge.  So I am not sure that I am liberty to disclose that.


How do you go about putting a value on IP, a combination of licensing fees and royalties?

Are you asking me how we go about pricing IP?


That’s one measure of value.

Do you have another in mind?


The models I have seen have been seller’s cost plus desired profit, …

Laugh.  Our IP gets used for interesting reasons.  One is people want a very dense memory and they do not want to use DRAM because of the cost and complexity involved.  So they come to us.  That’s one reason.  The other reason is that people want to shrink an existing chip and they can easily do that without going to the next technology node.  There is
inherent value to that.  The third one is that they want to do the next version of the chip, add more logic but keep the die size the same.  If you decrease the memory size from 6 to 1, then you can have more space on the die to add more logic.  So we have some internal mechanisms to try and figure out what the values of those are and come up with a way to try to monetize what we charge.  By the way our customers without exception are very happy to pay for what they receive and pay royalties as well.


How do you verify how many chips a customer sold with you IP?

I’ve been here about two weeks and my understanding is that we just take it on faith now.


Is that typical in the IP industry?

No!  People do lots of audits.  Both Artisan and Virage have vigorous audit programs, at least for the foundries.


Is there any difference in how MoSys conducts business based on geography?

There are cultural differences.  But we do not have a different pricing mechanism.


What about the ability to protect your IP in different geographies?

That’s a good one.  Certainly there are some markets where you work with not to the end customer.


Are there any differences beyond language that impact the way marketing is done?

Let me give you one example of that.  In the US when you issue a press release, you just put it on the wire.  For a really important release we may hold a press conference or call.  In Japan press releases have to be physically walked to the editor’s offices.  There are companies that specialize in that.


How about China?

China is interesting. A number of Chinese technical publications pick up on US news stories.  A number of them look to greater China region to see who else is publishing stuff.  There is a large and growing mainland China based press.  In fact I will be there in two weeks time.


Does the European technical press largely piggyback what appears in the US press?

Yes and no.  Some people do.  Some people look at the local scene.  One of the things we are looking to do in the future is press releases in multiple languages, particularly for major releases.  This is not something MoSys has done prior to this.There are number of good publications in Germany that are very efficient.  France is another
and the UK of course. 


How important are tradeshows, conferences and the like to a company like MoSys?

Again, it is the visibility issue.  You have to look at ROI and say “Is the lead generation and the visibility producing the right result?”  It is part of the total mix.  You need presence, you need visibility, you need speaking engagements, and you need PR and contributed articles.  You look at all of those as a holistic thing.


In the case of generated leads, you can track the sales results.  How do you measure visibility?

In the age we live in one possible way we can measure visibility and I will admit upfront it is not completely accurate is by web traffic.  There are other ways.  If people are writing about the space you are in, do you get mentioned and how often do you get mentioned.  Are you the main person who gets quoted?  Are you quoted at all?  Are you a sidebar?


Trade shows and publications provide demographics about their attendees and readers.  But that doesn’t address the issue of whether the target audience will visit your booth or read your ads.

That’s true.  Presuming if one picks the tradeshow carefully and you have a product that is of interest to people attending that show, you are going to get some percent of the people to come by.  If you pick poorly, then the attendance will be low.


What role does Marketing have in determining product direction?

Quite a lot, certainly in this company.  Build it and they will come” might have been a great idea at one point.  Understanding what the market dynamics are and where there is a potential for producing a product based upon the skill set and technology of the company is a very important thing. You want to build things that have value for your customers.


How do you determine what the product characteristics should be?

Mostly by talking to customers, which I am currently engaged in doing quite a lot.  The last two weeks have been like standing in front of a fire hose.


They think that you will correct all the problems or what they perceive as problems.

Time will tell.


Sometimes when you listen to customers they say they want x and y but then someone else introduces something completely different and your customers beat a path to his door.

The classic marketing example of listening too much to your customers was the Edsel car.  They said that’s what they want, not that’s what they would buy.  You have to be careful in what questions you ask and listen carefully to the answer and parse what they really need.  You are right.  People often do not know what they want.  Either because they can’t articulate it or haven’t
though that one out.  Yet we can listen to what customers say, we can look at the marketplace.  We can look at gaps in technology and say in this process there is a gap that nobody even thought of.


Years ago at Applicon we hosted mechanical customer and prospect visits where they would ask for many advanced features across a broad application spectrum.  Yet Autodesk became an overnight sensation when they introduced Autocad which had limited functionality, difficulty human interface and poor performance but with a price the square root of what we and others were offering.

Disruptive technology will capture mind share.  That’s an interesting analogy because I know Carol Bartz and Carl Bass (former and current Autodesk CEOs) very well.  They have a different problem today where they have reached a certain size.  Where no matter how many acquisitions they have done in the last 10 years, none have made money for them.  The core product carries on.  They can’t get out
of the core product.  They are stuck in their own success.


That was the same problem Lotus had for many years.

That’s correct


Until they came up with Lotus Notes.

They were a one trick pony for years.  People perceived them as that.  This is it.  That’s what they do.


Who do you see as competition for MoSys?

For people who want the amount of memory that we are good at, typically 2 Mbit and above, the alternative is embedded DRAM.  The number of places and foundries where embedded DRAM is available is very small.  In a lot of instances we are the only game in town unless you are an IDM, in which case you might have an internal DRAM program.


So a combination of available foundries and price?

Yes.  Typically DRAM only exists at certain technology nodes as well.


What is the impact of customers and prospects going to 65 nm?

We have customers in the 65 nm space and we have silicon in that space as well.  It is actually very good for us.  The MoSys stuff scales very well to new geometry.


Where do you see growth coming for MoSys?

In several areas!  We have just barely tapped the available market for our existing product.  We do have some new as yet unannounced products coming out.


The reason why MoSys has barely tapped the available market is?

Mostly resources and probably management focus.  There is a completely brand new management at MoSys today.  There is a new CEO, new CFO, new VP of Engineering.  I just joined.


Your challenge is to make them aware of your technology and that the risk-reward ratio is favorable.


That’s correct.  It is a message that resonates very well with people.  I have been in on several sales calls with customers.  It’s a good message.  They get it.  They look at me and say “Wow, I didn’t know I could that.”


I have interviewed many executives who say that their primary competition is not so much another vendor but the internal design team.

We do not have that particular problem.  In fact with IDMs who have internal design teams all the better because we can license them technologies in turn the design team can pick it up and use it.  We just announced last quarter such an agreement with Fujitsu.  We provided them a technology license.  We provided them people on staff to help their design team develop their own macros.  We are not in
competition with them.  We are an enabler for them.


For other firms looking for IP, do they simply pick out IP of an Internet catalog?

Yes!  Literally, they say they want memory at this particular size and speed.


You said that they have been lots of management changes at MoSys.  What attracted you?

MoSys has some truly innovative technology that isn’t getting commoditized.  It has real value to our customers.  What really attracted me was that the management team that is being put in place by the board which is new and energized, it has a vision of where it wants to take the company.  That’s a good place to be.


Where do they want to take the company other than more revenue, profit and stock price?

If you look at the IP space, there are a lot of companies out there, some private and some public.  One has a choice.  There is going to be consolidation in the landscape.  A lot of people will accept that.  Perhaps, you would too.  The question becomes “Do you want to be consolidated or be a consolidator”.  I
won’t answer that question.


My lips were just forming that question.


Does MoSys sell direct?

Yes!


In all geographies?

Yes.


That is because?

It is mostly because it is a technical sale and it is a long sale cycle and partly for historical reasons.  It may be that there are places where we could benefit from having relationships with third parties.  Certainly looking to see if there were other possible sales channels that we could use to spread our technology and make them more acceptable to SoC designers.  That’s one of my current initiatives.

 

The conventional thing is to have a rep in certain areas.  But there may be design service partners and other related parties that could provide access to chip designers for our technology and who provide a value added service.  So it would not be just a pass through because then you might as well go direct.  But if there are people who could provide value added service then that make s a lot of sense.



Your website lists as partners a number of foundries.  In what sense are they partners?

Some of them have technology licenses.  All of them have to qualify the technology on advanced technology nodes.  We work with their sales force to talk to our mutual customers and prospects about the technology.



The top articles over the last two weeks as determined by the number of readers were:


 
Incentia’s TimeCraft Advanced On-Chip-Variation Timing Analysis Adopted by Faraday  Incentia Design Systems, Inc., an EDA software company that focuses on the timing and synthesis market, announced that Faraday Technology Corporation, a leading provider of ASIC design services and silicon IPs, has adopted Incentia’s TimeCraft advanced On-Chip-Variation (OCV) solution to improve its timing analysis efficiency and accuracy with Faraday’s advanced
Location, Level, and Cell-Based On-Chip-Variation (LLC-OCV) libraries.  TimeCraft is a full-chip, gate-level STA for timing signoff. It features comprehensive timing checking and reports, a built-in delay calculator, multiple supply voltage analysis, multi-task capability for multi-corner and multi-mode analysis, and advanced PLL handling, which enable the analysis on a large variety of design styles. TimeCraft has the runtime and capacity advantages that can dramatically reduce timing verification turnaround time.


 
Cadence CEO Michael J. Fister Joins SynapSense Board of Directors  Fister, named Cadence CEO in May 2004 after a 17-year career at Intel Corporation, joins the SynapSense Board of Directors with CEO Peter Van Deventer and Chief Technology Officer Dr. Raju Pandey; Corley Phillips, Managing Partner of American River Ventures; and Rodrigo Prudencio, a Partner with Nth Power.


 
Bluespec Begins Volume Shipment of SystemC Synthesis
ESEComp is the first ESL control logic synthesis software to support the SystemC language. It allows models and designs written at a high-level, including complex control and datapaths, to be generated into efficient RTL code. As a SystemC-based synthesis tool, it unifies architecture modeling, software prototyping and implementation at a high-level of abstraction, and is the only general purpose
SystemC synthesis solution bridging ESL and RTL.


There is a free download of the ESL Synthesis Extensions (ESE, pronounced csc) to SystemC available through Bluespec's website (http://www.bluespec.com). This free version supports the ESL Synthesis language extensions for untimed simulations with the standard OSCI SystemC simulator.


ESEComp supports Red Hat Linux operating systems.


 
EVE Survey of DAC Attendees Finds Most Satisfied with Verification Environment  In a survey DAC attendees taken during the conference in July, EVE found that 60 percent of respondents were satisfied with their current verification environment. Of 617 surveys collected during the four-day conference, 477 were used to compile the survey.


Among the findings, close to 60 percent of respondents said that their design team performed hardware/software co-verification, and 55 percent use or plan to use hardware-assisted verification. The survey showed trends toward bigger designs -- 72 percent noted that their designers were larger than two million gates -- and a growing need for better performance or software development, leading to a strong desire for hardware-assisted verification solutions.


 
Cadence Pharmaceuticals Initiates Pivotal Phase III Clinical Trial Evaluating Intravenous Acetaminophen for Post-Operative Acute Pain 
Cadence Pharmaceuticals, Inc. (NASDAQ: CADX), a biopharmaceutical company focused on in-licensing, developing and commercializing proprietary product candidates principally for use in the hospital setting, today announced it has begun dosing subjects in a Phase III clinical
trial to evaluate intravenous acetaminophen (IV APAP) for the treatment of post-operative acute pain. The pivotal trial, referred to as CPI-APAP-301, is the second trial initiated as part of Cadence's clinical program to evaluate the efficacy and safety of IV APAP for pain and fever in adults and children. IV APAP has been marketed by Bristol-Myers Squibb Company in Europe since mid-2002, where it is now the market leader among injectable analgesics.


 
Cadence Design Systems Announces Intention to Offer $500 Million Senior Convertible Notes
Cadence announced its intention to commence an offering, subject to market and other conditions, of $250 million principal amount of convertible senior notes due 2011 and $250 million principal amount of convertible senior notes due 2013.  Cadence intends to use up to approximately $200 million of the net
proceeds of this offering to repurchase a portion of its zero coupon zero yield senior convertible notes due 2023 and a portion of the net proceeds to fund the cost of the convertible note hedge transactions described below. Cadence intends to use the remainder of the proceeds from the sale of the 2011 and 2013 notes to purchase shares of its common stock.


 


Other EDA News


  • The MathWorks Adds Support for Signal Integrity Engineering to RF Toolbox
      Other IP & SoC News


  • You can find the full EDACafe event calendar here.


    To read more news, click here.



    -- Jack Horgan, EDACafe.com Contributing Editor.


    Rating: