Getting out of EDA

Posted on September 3, 2009 by paulmcl

Over the last year I’ve had lots of meetings with people who used to work in EDA and have lost their jobs, or, in some cases, still have a job but want to make a longer term change. The subject that comes up all the time is “How do I get out of EDA?”

This is not unreasonable. EDA has shrunk its employment over the last couple of years and it is unlikely to come back again to its previous level. So some people will need to find jobs in new industries.

If you are an engineer in EDA then you know how to do very technical programming. You could certainly do other forms of technical programming. But the sweet spot in the job market is in internet companies and there is a lot of specialized stuff there that you probably don’t have experience of. If someone wants to get an internet startup going quickly then you want people who already know Ruby on Rails, mySQL or the iPhone developer kit or whatever. Not someone really smart who could probably learn that stuff eventually. Personally, I think this is silly. A smart programmer can suck up a language in no time and will run rings around someone less good even with a lot of domain experience. Good programmers are not 30% better than average ones, they are 10 times better. But even if you get hired, you don’t get paid for all that deep knowledge of, say, placement that you’ve spent years acquiring.

If you are in marketing or management it is even more difficult. At one level you have experience of running business to business (b2b) marketing for a software company. But you have years of understanding of IC design and none of relational databases or whatever, which makes it hard to make that transition. Furthermore, most internet companies are business to consumer (b2c) or internet-based business to business which is very similar.

I interviewed over a year ago with a b2c company and I was amazed that they seemed interested in me. It was bit like the Groucho Marx joke about not wanting to be a member of any club that would have him. The fact that they seemed interested in hiring me for a job that I was so manifestly unqualified for (although it would have been interesting to learn) made me doubt their competence.

If you are in some other domains you get stuck in those domains. I have a friend who is in finance. That allows you to work in all sorts of different companies, but always in finance where all companies look very similar. She wants to get out of finance, which is a similar problem. She’s smart enough to do all sorts of jobs but the only jobs that will pay anything close to what she is used to are ones that value all that financial experience.

It’s a tough transition to make. Your experience is what makes you valuable in EDA (or finance or whatever). If you go somewhere where that is not valued it is very hard to make anything close to what you made in EDA. After all, EDA pays pretty well so long as you have a job.

Posted in eda industry, silicon valley | Comments Off

Being CEO

Posted on September 2, 2009 by paulmcl

I talked earlier about how you get to be CEO (basically, luck the first time; track record after that). But what does being a CEO entail?

I think all senior management jobs consist of two separate dimensions that have two separate skill sets. I call these management and leadership. Some people are good at both, some are good at only one.

Management is the basic operational management of the company. Presumably you already know how to do this, at least in your own domain (engineering, marketing, sales, etc) or you probably wouldn’t have been promoted. When you get more senior you have a new challenge: you have to manage people not from your own domain. If you are an engineer, it’s like salespeople are from another planet and you don’t understand what makes them tick. If you are a salesperson you may think the same about engineering. If the company is medium sized things are not so bad since you’ll have a sales manager and an engineering manager to insulate you. But if the company is small then you’ll have to manage the aliens directly. My recommendation is to get some advice. If you’ve never set up a sales commission plan before, don’t assume that because you are a smart engineer who knows Excel that you can just wing it. If you don’t know a friendly VP sales who can give you free advice, find a consultant and pay them. It’s a lot cheaper than making major mistakes.

As CEO you may have only an accountant (or maybe nobody) to support you in finance. I think it makes sense to get a “CFO for a day” consultant to help you unless you are very comfortable with all the finance issues and already have a good feel for how to put together a business plan, how to turn a sales plan into a cash-flow forecast and so on. If your eyes glaze over when you read my blog postings on finance, you need someone to help you. Whatever you do in finance, don’t treat it as a problem that will go away if you ignore it. You’ll need to get a financial audit done at some point, sooner than you expect, and cutting corners will then come to light.

If you are not an engineer by background, you can’t manage engineering. That’s not to say that you aren’t capable of managing engineering but just like salespeople won’t respect you unless you’ve carried a bag, engineering people want to be managed by someone who understands technology and development and knows what it takes to get a product out. If you don’t have an engineering manager you’ll at least need to trust one of the senior engineers to be feeding you the unvarnished truth.

The second leg of being a CEO or a senior manager is leadership. The most important aspect of this is to get everyone in the company committed to moving the company in the same direction. Unless you make truly stupid decisions, it is more important that everyone is aligned than that the decision is ideal. As General Patton famously said, “A good plan executed violently now is better than the perfect plan next week.” In business, “violently” is the wrong adverb but the sentiment is the same.

Having said that, it is also important that the overall strategy of the company is good and represents the best that the management team can come up with. It is also important to be flexible. If something isn’t working then you’ll need to try something else and preferably while you still have enough money in the bank to find out whether that new approach is good. Remember, most successful startups end up doing something somewhat or completely different from what they set out to do initially.

A general rule about management and especially being a CEO: if something good happens in the company, everyone will tell you about it. If something bad happens, nobody will tell you. Despite the proverb that "bad news travels fast," inside a company bad news travels really slowly so you need to make a special effort to discover it. In the early stages it is good to have someone in engineering who is a personal friend who will not hide bad news. Later on, you need someone in sales like that who’ll tell you what is really happening when the company tries to sell the product. You can’t sit in your CEO office and believe everything that you are told. You have to get out and dig.

Posted in management | Comments Off

Potpourri

Posted on September 1, 2009 by paulmcl

Today’s blog is a potpourri of various small things, some of which refer back to earlier blogs.

Firstly, you probably saw that Cadence appointed John Bruggeman as CMO. He was previously CMO at Wind River (recently acquired by Intel). I met him several times when I was at Virtutech since we were both in the space of embedded software development. We even persuaded him to come and give an opening speech at our sales kickoff one year.

Prior to Wind River, John was at Mercury Interactive (along with Ken Klein the CEO of Wind River). There he came up with the concept of Business Technology Optimization, or BTO, as an umbrella theme for what they did. At Wind River, he tried to reposition the company as more than just embedded with the mantra of Device Software Optimization or DSO. If you were at the Embedded System Conference any time around then, you couldn’t but notice the ten foot high white “D,” “S” and “O.” So what will it be this time? EDA, boring. Electronic System Optimization, that’s hot. ESO, you read it here first!

**

A couple of days after I wrote about mobile payments, Nokia announced Nokia Money targeted at just that space, along with a company Obopay (which is partially owned by Nokia).

And by the way, a pet peeve, Nokia is pronounced with slight emphasis on the first syllable and not with emphasis on the “i” as many people say it. While on that topic, going back to talking about learning Chinese, Beijing (bei=north, jing=capital) is pronounced with a “j” as in “jug” not a sort of harsh “sh”.

**

Going back to Japan, I accused the handset makers of being too inwardly focused and competing only with each other. It looks like they are going to solve that problem by merging them. NEC, Hitachi and Casio are looking to merge their handset operations to create the second largest handset vendor in Japan (Sharp, surprisingly, is number one).

**

I talked earlier about Apple’s strength with iPhone, not so much in volume but in profit. Apparently Apple had (in 1H09) 1.9% of handset unit volume, 8% of revenue and a whopping 32% (almost a third) of all the profit. The numbers are a little distorted since both Sony-Ericsson and Motorola lost money (so reducing the denominator). If we take them out of the equation, they still have 25% of the profits of the profitable handset vendors.

**

Magma came out with surprisingly good results, beating street estimates. Revenue was $28.8M and cash-flow was positive. I recently talked with someone who told me that their synthesis was now better than Cadence or Synopsys but place and route was not (yes, I know that is anecdote and not data). Of course SP&R is a horse race, and there are still too many horses with Synopsys, Cadence, Magma, Mentor/Sierra, Atoptech and others. It is a marathon not a sprint.

Posted in admin | Comments Off

Jay Singh’s view

Posted on August 31, 2009 by paulmcl

I met last week with Jay Singh of Plato Networks. His official title is “Manager CAE” which makes it sound like he’s in charge of schematic capture. Isn’t CAE what we used to call what Daisy, Mentor, Valid did in the early 1980s? Anyway, what he really is responsible for is all aspects of their design methodology.

Unlike many people with such job functions, he takes his responsibility to both assess and nurture new technologies seriously, making his state-of-the-art design data available to both established companies and startups, provided they are prepared to come and camp out in his office since data that valuable isn’t going to leave the building.

This is incredibly valuable for companies that get involved. In my experience, one of the biggest headaches with a new EDA product (whether from a startup or an established company) is getting your hands on designs. If you are creating a tool for the 32nm process node, it just doesn’t do to only test it on some old 90nm designs that you happen to have lying around. You have to find a “teaching customer” who is prepared to let you get your hands on their data, which means that they will also need to give you some manpower since you also need someone who understands the design.

The biggest complaint Jay has is that after he’s done all this for the EDA company that their salespeople then completely ignore it and expect him to do an arms-length deal as if he’d never helped them mature the tool in the first place. He’s thrown out several companies (that I’ll leave nameless) both big and small despite having promising technology because their business people have been too arrogant once their engineering team has left the building having taken advantage of his data and resources to mature their product.

So which companies does he think have interesting technology? By “interesting” he doesn’t mean that they have incremental improvement to legacy tools but rather they are taking a new approach to how to address an aspect of design.

Jay has four companies on his list of potentially disruptive technology.

Firstly Oasys Design Systems. I was originally meeting Jay because he was one of the earliest users of Oasys RealTime Designer and I was going to write up his experiences for the Oasys blog (which I have done, it’s here). But somehow the 30 minute meeting we’d scheduled to talk about this turned into a 2 hour meeting talking about lots of other aspects of EDA too.

The next company is Atoptech, which is taking a new approach to place and route and in Jay’s opinion outperforms the usual suspects in the place and route space with their Aprisa product.

Berkeley Design Automation have a new approach to circuit simulation which is ten or twenty times faster than the traditional circuit simulators. They’ve also managed to do it while preserving complete compatibility with those circuit simulators: the input files are the same, the output files are the same. It’s easy to check if you are getting the same results and you don’t have to do anything extraordinary to your input to gain the speedup.

His fourth company is Analog Rails. The usual way analog has been designed is that a very experienced engineer designs the transistor level, estimating parasitics conservatively. Then, when this is complete, it is handed off to a layout technician who does the layout without really understanding any of the implications of layout decisions on parasitics. This just doesn’t work for 65nm and below. Analog Rails has automatic analog layout with on-the-fly extraction so you end up with smaller implementations faster. They also have some hilarious videos.

One interesting aspect of these four companies is that they don’t overlap at all and between them they cover pretty much the entire design creation process (except for physical verification but everyone uses Calibre anyway). If they were all in one company it would be a complete next-generation design flow.

I should point out that Plato doesn’t necessarily use all of these products in their production flow today. Like most companies, they are conservative about adopting new technology and do so in a very controlled way. But Jay is very aggressive about both evaluating and driving new technology so that it is there when he needs it.

 

Posted in eda industry, methodology | Comments Off

Friday puzzle: Monty Hall

Posted on August 28, 2009 by paulmcl

Last week’s puzzle was the camel and the bananas. Firstly, a moment’s thought shows that the camel can’t get any bananas to market in one go. It is 1000 km away and so the camel will eat all of its maximum load of 1000 bananas and then be stuck at the market anyway. The least number of loads the camel needs to take from the plantation is 3 (3000 bananas, maximum load of 1000). So the camel will have to make 3 trips from the plantation to some intermediate point to move all 3000 bananas (less the ones eaten), meaning 5 trips in all across that distance (3 outgoing, 2 returning). If we arrange that after those 5 trips precisely 2000 bananas are left then it can make two trips from that intermediate point to a second intermediate point, meaning 3 trips across that distance (2 outgoing, 1 returning). At this point we want there to be precisely 1000 bananas left with which the camel sets off to the market. The first point needs to be 200 kilometers away (so that 1000 bananas are consumed on the 5 trips leaving 2000). The second point needs to be 333 1/3 miles further on (533 1/3 kilometers from the plantation) so that 1000 bananas are consumed on those 3 trips.

So it plays out like this. The camel takes 1000 bananas to the first intermediate point 200 kilometers away. There it drops 600 bananas and returns with 200 (and it already ate 200). Again it takes 1000 bananas, drops 600 and returns with 200. Then it takes the last 1000 bananas, gets to the intermediate point, picks up another 200 (to replace the 200 it already ate) and takes them to the second intermediate point a further 333 1/3 kilometers away. There it drops 333 1/3 bananas and returns with 333 1/3 bananas. Then it takes a the remaining 1000 bananas from the first intermediate point to the second intermediate point, arriving with 666 2/3 bananas. It picks up the 333 1/3 bananas already at the second intermediate point (making 1000) and sets off for market 466 2/3 kilimeters away where it arrives with 533 1/3 bananas out of the original 3000.

Today’s puzzle: You are on Let’s Make a Deal. There are three doors. Behind one door is a car. Behind the other two doors are goats. You pick door number 1. Monty Hall, the host, who knows what is behind each door, then opens a different door to show you a goat. He offers you the choice of sticking with your choice of door number 1 or switching to the remaining closed door. Should you switch?

To make the problem unambiguous, assume that after you pick a door, Monty Hall must then open a door and he must pick a door that he knows has a goat behind it and that he must then give you the chance to switch.

Posted in puzzles | Comments Off

Learning French

Posted on August 28, 2009 by paulmcl

I wrote earlier about learning Chinese, but the first foreign language I learned was French. I did it at school and I first went to France on a school trip in 1964. I even have a French O-level (the exams you used to take in Britain when you were 15 or 16). However, although my written French was OK then, my spoken French was terrible. Partially because the way you learn languages in school emphasizes written stuff over spoken too much (probably because it is easier to test and grade).

But mainly because French as it is spoken is not at all the way it is written. For example, correct French for we are going to eat is “nous allons manger” but spoken is much more likely to be “nous, on va manger” which literally translated is “we, one is going to eat.” There is a future tense “j’attendrai” that expresses that you will wait, but spoken it isn’t used and you say it the English way “je vais attendre” or “I am going to wait.” That’s before getting to the fact that the French have lots of slang (argot) that you never get taught in school. Weirdly the French even have some words that are obscene depending on context. For instance, un baiser is a kiss, but baiser as a verb is pretty much the equivalent of the F-word and considered just as obscene.

I really only learned to speak French when I went to live there. The best way to learn any foreign language is to (a) go and live in the country (b) get a native girlfriend or boyfriend. Since I was married when I went to France, I was only able to do the first of these! When you go and immerse yourself in the language like that a couple of things are surprising. First, you will dream in the foreign language. Dreaming seems to be some sort of mental garbage collection, and there’s a lot of foreign garbage to be collected when you are learning so much. The second thing, which occurs when you’ve been there for a longer time, is when you can’t remember the word for something in English but you know it in the foreign language. I remember trying to think of the English for a purchase-order but I’d forgotten what we called it; but I knew it was bon de commande in French.

French has genders, but unlike German at least there are only two. Cute things tend to be feminine: un batiment (masculine, a building) but une maison (feminine, a house). But like most foreigners you perfect the sound halfway between the masculine and feminine articles (le/la or un/une) and assume anyone listening will hear the correct one.

But the biggest problem is that French has lots of silent letters that are written but not pronounced. It is fairly easy when reading to ignore them but it is hard when writing to remember just what to add back in. My written French is terrible even though I now speak it pretty much fluently. The silent letters result in words being spelled totally differently but pronounced the same. For example, all of the following are pronounced identically: “ver” a worm, “vert” green, “vers” towards, “verre” a glass (not to mention “verres” glasses). And just to add to the confusion “vers” doesn’t just mean towards, it’s also a verse of poetry and the plural of the aforementioned worm.

Another good way of learning any language is to listen to the news. In France there is a radio program France-Info that puts out the news every 7 minutes with other stuff in between. Newsreaders speak clearly, fairly slowly, don’t use slang and generally are easy to understand. Since you already sort of know what is in the news much of the time (and if you miss it it’s coming round again in 7 minutes) it helps a lot when you first are learning. Trying to understand a French sitcom is the absolute opposite: spoken fast, slurred words, slang, jokes, a nightmare for the beginner.

Posted in culture | Comments Off

Mobile payments

Posted on August 27, 2009 by paulmcl

I have been doing some work recently with a biometric mobile payment startup. So I went to a VCtaskforce meeting last week about mobile payments. I learned some new stuff to go along with what I already knew. The first interesting thing, which apparently surprised the organizers, was that the room was completely full. There’s a lot of interest in mobile payments.

The world of mobile payments has a multi-dimensional taxonomy:

  1. Is this a developed world or developing world mobile payment system? In the developed world we have invested a lot in payment infrastructure for credit cards and ATMs. In the developing world this is much less developed and in the rural parts of the developing world not developed at all. Here’s the interesting statistic: there are four billion mobile phones in the world, but only one billion credit cards and two billion bank accounts (so one billion people with a bank account but no credit card). If you were Visa going to a new country to set up a way for people to pay for stuff, you’d do it with phones rather than trying to replicate the type of landline based system of terminals, ATMs etc that we have built. But in the developed world mobile payments have to compete with credit and debit cards, which already work pretty efficiently. Remember, most people in the world have prepaid cell phones, not billed. Even in the US, prepay is the fastest growing part of the market.
  2. Is it digital goods or physical goods that you are paying for? And a secondary question is how big the payment is. Cheap digital goods (music tracks, ringtones, wallpaper, virtual currency in games etc) can easily be handled by network operators who just put it on your bill. If it requires shipping something then the operators are not set up to do that, and if the price becomes too high then their support costs go up (every call to a network operator costs $6 on average).
  3. Is it remote or proximity payment. Proximity payment means that you wave your phone over some sort of sensor in a store. The leading technology to make this work is NFC (near field communications) but it is somewhat on the back-burner now due to the recession. Maybe in 3 or 4 years it will be come important. Remote (meaning that you are not necessarily at a store at all) is the most interesting part of the market anyway. I don’t consider proximity payments in the rest of this.

The players in this are banks, network operators (like Verizon and AT&T) and third party companies trying to build payment infrastructure independent of banks or network operators (or perhaps in partnership with them, depending on business model). The banks have so far not been very successful at setting up mobile payments although in the longer term they might be since the network operators are really only good at a few things: building networks, transmitting data over them, and billing people. So they are set up to sell you a ringtone. But once physical goods start to get involved it is no longer something that the network people know how to do and is more the preserve of people like Visa who already have all the fraud infrastructure, chargeback infrastructure and a business model to support it.

In Kenya (and now some other African countries) Vodaphone and their local partners set up a system called M-pesa. M for mobile, pesa is the Swahili for money (there’s no end to the interesting stuff you learn on this blog). It is basically branchless banking. With your phone you can transfer money from your pre-pay account to anyone else’s. You can go to many stores and pay in cash to top up your pre-pay account. Or you can take money out of your pre-pay account. So a worker in a city can go to the grocery store, add $50 to their account, transfer it to the account of their wife back in a village far from any bank, and they can go to the grocery store and take the $50 out of their account. Apparently people also use it to avoid getting robbed too: pay money in before getting on the bus, take it out at the other end. There are 7M users (in just a couple of years) with 2M transactions daily in just Kenya.

Mobile payments today are $24B annually. It is all either digital goods or cash infrastructure with things like M-pesa, essentially no physical goods. However, there is one big problem: security. The moment Paypal launches in a pre-paid market they see attacks go up. In fact, mobile security is better than regular internet security (the network really does know which phone it is and whose it is) but it is inadequate. If stealing your phone enables someone to empty your bank account you need better protection than just a four digit PIN.

Biometrics offers one of the best solutions to this: put a fingerprint sensor on each phone. That works OK for high-end phones but doesn’t do so well for the lowest end where phones cost only $20 to build and can’t absorb the cost of a sensor. The challenge is to have high security but a seamless customer experience.

Who might you not have thought about who could suddenly become a player in this space who is not a bank nor a network operator? Paypal, obviously. But also there is somebody who has 300M accounts already set up with payment instruments. Somebody who already sells over 25% of all music using this technology, some of it in the mobile market. Apple, with iTunes and the appStore, have moved a significant amount of distribution and billing away from the carriers. They have an infrastructure they could do more with if they decided to.

There’s another gorilla lurking who already has a huge number of accounts (I don’t know how many) already set up with payment instruments and already in the business of shipping a lot of physical goods? Amazon.

Posted in security | Comments Off

What’s Dan Dobberpuhl doing?

Posted on August 26, 2009 by paulmcl

As I’ve pointed out before, most of the differentiation in the iPhone is in the software and the industrial design. Almost none is in the hardware which, especially in the first version, was all off-the-shelf not particularly special standard parts.

So what’s all this chip design that Apple is doing? Apple purchased PA Semiconductor over a year ago. PA Semiconductor was building an ultra-low power version of the PowerPC architecture. Supposedly, however, they warned their customers that the buyout was unrelated to their products and just to their expertise. So I think we can assume Apple isn’t going to go back to using PowerPC in either Macs or iPhones (which have always been ARM based anyway). Their expertise is in building processors with high performance for extremely low power. Sounds like just the sort of thing you’d need in an iPhone.

In July, ARM announced that a leading OEM handset maker had taken out a multi-year ARM architecture license. This is widely believed to be Apple based on a lot of anecdotal evidence.

The CEO of PA Semiconductor was Dan Dobberpuhl (who also happened to be on my technology advisory board at Ambit so I know him somewhat). Prior to founding PA he was at SiByte (acquired by Broadcom) [corrected from earlier]. Prior to that he was one of the main architects of both StrongARM, ARM on steroids (which Intel renamed Xscale when they acquired Digital’s semiconductor business) and Alpha, the highest performance microprocessor for its power.

Don’t forget that Apple was one of the original investors in ARM when it was spun out of Acorn/Olivetti. The Newton was ARM based. Their dealings go back a long way.

There are also rumors of Apple producing a tablet computer, or an internet device or something generally between a laptop and an iPhone. It is interesting to speculate whether this would be like a Mac and powered by Intel (presumably Atom, but maybe not) or like an iPhone and powered by an ARM processor. They have some sort of event set up for October and all the rumor mills are working overtime.

So my guess is that Apple has decided that with the team of low power experts headed up by Dobberpuhl, and an ARM architecture license, that they can design a cell-phone with more processing power for the same battery life than the competition. They can then use that processor power with their software expertise to deliver a user-experience that will be very hard to equal for anyone else using a regular ARM to run Symbian, PalmOS or Android. Time will tell.

Posted in semiconductor | Comments Off

Analog mobile TV

Posted on August 25, 2009 by paulmcl

I talked today to a company called Telegent Systems who make chips for adding TV to cell-phones. The bulk of their business is adding ordinary over-the-air analog TV to cell-phones. You’ve probably never seen this since the US is not the primary market for this, not least because we finally just turned off analog TV. But it is a potentially huge market. It is unusual to find an American company completely focused on the non-US market since there really isn’t a domestic market. The US is such a big market and is a leader in many segments so it is natural to think of the rest of the world as where you go to once you’ve nailed the US. Sometimes it’s even more parochial. Nail Silicon Valley, then the rest of the US, then maybe Japan and Western Europe. But some markets are completely backwards from this.

It is always easy here to miss what is happening in wireless because the US is its own market, and a laggard rather than a leader. It has some standards not really used anywhere else. It is dominated by walled-garden operators like Verizon and AT&T who have their own stores, who have 2 year contracts and so on. The rest of the world isn’t like that. The average Chinese cell-phone owner changes their phone every 6 months. Stores in much of the world sell phones to work on all operators, perhaps because I don’t think there is any other country where the radio-interface isn’t standard for the whole country, usually the near ubiquitous GSM. Even someone as smart as Andy Kessler calls for all phones to work on all networks without realizing that this is a technical issue not just a business one. The phone and the network service are really two separate purchases: you can buy a phone and then buy a SIM-card which controls the network service, and just put the SIM-card in the phone. So the US is very non-innovative since it is all controlled by what the network operators like or don’t like, compared to some of the freer markets overseas which are driven by what the consumer likes or doesn’t like. It’s still like the internet was in the dark days when most consumers connected to AOL.

Apple gave Verizon (apparently) and AT&T a take-it-or-leave-it deal with the iPhone which served to break up the walled-garden somewhat. Until then, for instance, if you wanted to watch a video on your AT&T phone you couldn’t to go YouTube or anywhere else, you had to go to AT&T’s video site. Verizon are probably regretting saying “no” to the deal as they hemorrhage high-end customers to AT&T to get iPhones despite their better network quality.

There will be over 54 million phones shipped this year receiving analog TV and all of them are powered by Telegent. But that’s a tiny bit of their potential market since there are something like 5B people in the world who have a cell phone but no access to mobile TV. In fact, in many cases, no access to TV at all. Buying a phone with TV capability may be their only way to get a television. So despite the world going digital, both Instat and Forward Concepts are forecasting that analog TV will dominate the market for the time being.

Telegent are a fabless semi company based partially here in Silicon Valley and partially in Shanghai. Their end customer are the cell-phone manufacturers who see this as the first new feature that they can add to phones, especially low-end phones, since mp3 players many years ago. It gives them differentiation and, since it is picking up the standard terrestrial analog broadcast signal, it doesn’t depend on the wireless network operators’ support or the need to negotiate any special broadcast rights (good luck on doing that in less than a year or two). It is driven by consumer demand, since TV is free over the air. The only cost is the incremental cost of the chip in the phone and it’s an under-$10 part. Telegent solved the old problem of how do you build an analog TV receiver in a new way, namely putting everything into a single cheap chip, making do with a small cell-phone screen, and making do with a tiny antenna nothing like the one you used to see on every house’s roof, which would rather spoil the show. Some interesting technical challenges.

They started 5 years ago with the Chinese market as their first target. They always knew that their customers would primarily be overseas, at least for these initial products. Since then they have expanded into Latin America, South East Asia, Middle East, Russia. In China, since, as I said above, the average cell-phone customer changes their phone every 6 months, this has meant that adoption has been driven very fast without needing to wait years for the installed base of handsets to turn over.

It has also turned out to be a big asset in emergencies. In the recent earthquake in Sichuan the only way people were getting information at all was on their mobile phone TV since so many cell-phone towers had been destroyed, but over-the-air TV could be picked up from as far away as Taiwan. The Telegent mobile TV chip has better sensitivity than most normal TVs plus, for several hours of watching, didn’t require the power grid to be working.

They are moving into digital TV and, as they do, the US market will start to become more important. Telegent want to get TV into every laptop too, since with a bigger battery and a bigger screen it will be a better experience. Their goal in life is to make every LCD flatpanel into a TV: cellphones, laptops, displays, photo-frames, DVD players.

Meanwhile, in Africa next year there is a big event that will potentially drive a lot of interest in watch-anywhere mobile TV, namely the World Cup, which is hosted by South Africa this time. Something else where the US is not the primary market and where you won’t realize its importance without leaving the country.

Posted in semiconductor | Comments Off

Big company guys don’t do small

Posted on August 24, 2009 by paulmcl

Big company guys think that they can run startups because they’ve run small divisions of big companies. So that must be the same, right?

Actually the two things are very different and not many people seem to be good at making the transition once they have got used to how a big company works, with their assistants, and finance organization, and HR department and all the rest.

When I was at VLSI and the fab was not running effectively, the company would hire a VP from TI or Motorola (where the CEO had previously worked and so knew good people he’d worked with before). These guys were used to running a fab that was running smoothly, with a large organization around them. They were not used to sorting out a dysfunctional fab with very few people to support them. When they didn’t work out, they were doubly expensive because they needed big severance packages to get rid of them.

When you become CEO of a startup, you have do everything yourself. Especially if the startup is attempting to run very lean with minimal cash burn, and conserve most of that cash for engineering. You want to put together a business plan? Fire up Excel. There’s at most a part-time accountant in the finance department and you can’t delegate it to them. Even if you have a “CFO for a day” part-time senior finance consultant, they don’t understand the business intimately like you should because that’s bound up with strategy which is not just something financial. They can help review the plan but they can’t do it for you.

If you’ve not got a very good engineering manager then you can’t rely on the current schedules. And you don’t have enough money to do what you would in a big company and hire a good engineering manager, or even a really good product management specialist. You have to do that yourself too. In a typical startup, as CEO, you will probably be the only person who isn’t writing code or designing chips.

Another problem with big companies is that people don’t really know how successful their business really is, since it is often very bound up in company-wide financial measures that are not closely enough tied to reality. So it is easy to look good when you aren’t, or looks undeservingly bad. If you are in a big EDA company, nobody knows how to really allocate revenue from big volume purchases to product lines. If you are in a semiconductor company, the cost model is rarely as accurate as necessary, and fab variances (because the fab is overloaded, or underloaded, or not yielding as expected) distort it again.

If your company has a few hugely profitable product lines (think Intel or Synopsys) then the smaller product lines may look good or not depending on how the overhead of the company is handled, and whether the profitable lines eat a lot of overhead leading to everyone else looking good (margin bleed-through), or the opposite, leading to everyone else looking worse than reality. It is too expensive to do full activity-based costing (ABC) and so overhead is often misleading. If cost of sales is a fixed percentage  of revenue, that assumes all products and all order sizes are equally easy to sell, which is clearly not true. But this may make some product lines look great (hire that manager) and others look poor (and he looked so promising) even though it purely an artifact of the underlying management accounting.

Although it is possible to make the transition from a big company to a startup, but both EDA and fabless semiconductor are littered with people who failed to do that. They were very successful at running a division of a big company, but were unable to translate that skill into success at either founding or coming into a startup and getting it to run well.

Posted in management | Comments Off