patrick.net

allow me to explain what it is like to work in a "fabless semiconductor company"; or, why i would go postal and kill everyone

so this douchebag decides to kill his boss and coworkers on a friday afternoon. i can’t pretend to know exactly why this guy offed some other guys but given that i, too, have worked in a four years old and funded fabless semiconductor design startup maybe our experiences are in parallel.

fabless is codeword for lame

in silicon valley there are literally thousands upon thousands of these fabless shops. companies, usually led by an engineer scion from a bigger firm or academic background, that begs for cash from ventures and angels with a pitch to develop either a cheaper copy of an existing integrated circuit design or an amalgamation of multiple designs onto systems-on-a-chip (SoC). the one i worked at had the pipe dream of trying to integrate wifi and wimax standards onto a dual-die SoC while siport is trying to develop a mp3/media radio transmission chip for mobile platforms. the shops are called fabless because they don’t actually own a fab — a foundry to physically produce chips — but rather either sell the designs to a bigger corporation, back then to freescale, or to original equipment manufacturers (OEM) in asia who makes all the electronic shit that you so love.

fabless is codeword for hardcore

startup ceos with a chip on their shoulder or too big of an ego might even try package up their designs and hire a foundry, such as tsmc, to make the chips for sale. the risk for this is crazy, especially considering the extensive time and staff the sales process requires. although if your business were 85% all one company, such as portalplayer with apple, then by all means you’re going to be flush with cash — up to until apple decides to make everything inhouse, then you’re just fucked.

but even then, with faster development, no need for chip production investment, and potentially lucrative market of oem electronics in a global market, the risks are pretty good in comparison to the upside and these shops run all across silicon valley. hell, it’s why we’re called silicon valley. hell, if your company flops there’s bound to another one down the street that’s hiring.

make no mistake, though, when people describe startups before the go-go days of dotcoms (late 80’s and 90’s you punk kids) these chip companies were the shit that they were talking about. people sleeping under their desks. 90 hour workweeks. you eat, sleep, and die by your company.

fabless is codeword for slavery

that then lies the problem. everyone wants in on this game. if you just listen carefully at every chinese restaurant — or indian restaurant — someone is talking loudly about their work in this field. sometimes you might get to evesdrop on their cell phone conversation with their chinese or taiwanese office complaining about the inability to get some shit done.

this crazy competition to produce the designs as fast possible in order to garner sales puts an incredible amount of stress. to who? why, to a nonexistent labor pool. you see, the problem ends up being that there’s just simply not enough labor to go around. the talented electronic engineering or computer science kids homegrown via american education gets snapped up by big names such as intel even before they graduate. the second and third tier engineers lack the experience, you need people with master degrees sprinkled with phds, to actually research and develop an entire chip by themselves. so what’s one to do? import. h1bs.

but without big money the quality of labor you bring in from china or india are still not top-notch. most likely they’re the second-string engineers unable to pick up jobs in the hottest market in the world. but hey, at least they’re cheap. and the cool thing is that once they’re brought here you can basically put them into “corporate housing” which basically amounts to slum apartments that you rent in milpitas or fremont. and since the company has sponsored them, their ability to stay in the usa is completely in your hands.

fabless is codeword for douchebag

that then brings me to this point. why would this douchebag kill his boss? failure. either the entire project is fucked or ran out of money. shit like that happens in the valley. and from there things just go down here.

those with the circumstance to be able to pick up another job will just move on. for them, the lull in between is meaningless. but those that can’t, those without money and no freedom earned, suffered through the insane production schedules. those people will snap. people in their late forties who have immigrated to this country in hopes of making it big will outright fucking snap. and every instance in where they were slighted will quickly turn into a means of justification. the anger you have in thinking you got screwed over will manifest deep in your hearts.

and it’s sad.

i have no snark for the kicker.

http://nevercapitalize.com/post/59796572/allow-me-to-explain-what-it-is-like-to-work-in-a

TOB, I don't even know what my opinion on that is. I'm so numb to this catastrophe that even crazy ideas seem like they should be listened to.

I do find it ironic that social security, which my understanding is only returns 1%, now outperformed other private retirement plans. Many of these have a negative return.

PermaRenter, I just had a chance to read your FAB commentary. It opened a window to a world I know nothing about. I could relate to companies owned and run by scientists who think they're business people but the rest of it reads like a depressing story of servitude. That industry sounds like it sucks big time.

Malocm,

The fabless story is written by Alaska Miller. Contact him at
alaska.miller@gmail.com
http://alaskamiller.com/

Best regards,
PermaRenter

Malcolm,

>>I could relate to companies owned and run by scientists who think they’re business people

What about the companies run by "business people" who think they are business people? The kind that thinks Activity Based Costing is da sh*t? I'm serious, this is even bigger a problem.

PermaRenter,

I don't agree with you on your blanket assessment of fabless semiconductor companies. Almost any semiconductor company these days is fabless. Look at Broadcom, Marvell, Atheros and Cavium for some fairly recent successful examples.

Xilinx, anyone? Altera? Sandisk? Or Qualcomm? Almost any semiconductor company started since 1990 has been fabless. There are good economic reason for this to be the case.

That being said, I think you are talking about a certain type of fabless company started by some me-too-Joe-wanna-be-CEO that has neither the business nor technical acumen required to build a success. These kind of companies have been the walking dead, and they hire a lot of inexperienced H1B and less-than-stellar engineers because that is all they can get. They try to make up for it by working 24*7, by sometimes even that is not enough. I'm distinctly not talking about any company in particular here, but that class exists, and there is quite a few of them.

Another observation: If some big time CEO complains endlessly about a shortage of engineers and lobbies for increased H1B quotas, it is often a sign that
the company or the CEO (or both) are crap and is to be avoided.

PermaRenter,

Even with the link at the end, that Cut-Paste job of yours was so long that I mistook the writing for being yours. See what happens when you cut/paste as much as you do, with no introduction. Please write some intro or attribution on top when you are going to paste ANY material.

Correction: My comment is hereby directed to Alaska Miller, not PermaRenter.

Justme, point taken. I've been around long enough to have heard all the buzzwords that everyone in my field then jumps on.
You know them all as well:
6 sigma
activity based costing
JIT (I am a big believer of this one, and it is not just having no invenotry)
Crossing the Chasm

All that shit that all of it has some merit in the right application, but these clowns jump all over the latest and greatest and sacrafice just using common sense. This old buddy system where like I've said before, no one has the guts to point across the conference table as a way of reducing waste in my opinion is the underlying factor. I further agree with you because, like in my field of the study of entrepreneurship a vast number, and I mean vast, of successful entrepreneurs are tough street smart people with a real caring of creating something of value and helping society.

permarenter's excerpt:
fabless shops. companies, usually led by an engineer scion from a bigger firm or academic background, that begs for cash from ventures and angels with a pitch to develop either a cheaper copy of an existing integrated circuit design or an amalgamation of multiple designs onto systems-on-a-chip (SoC).

That is TOO KIND a way to put it. In most cases, the founder(s) are engineers who start the basic work on a parent company's dime (Cisco, Intel, Cadence, Apple, etc.). If the prospect looks like it is commercially viable, they withhold the information and STEAL it from the parent company by walking out and "starting up" a fabless shop with the help of a VC. Hire a few slaves to do the fabless grunt work in an old, damp office building on Scott or Kifer. Once engineering is done, they co-opt a senior sales-exec from the big-co, who helps the startup poach from the big company's customer list.

Their exit strategy is then to sell the company for fat-stacks to one of these same big companies, and "work" there for a couple more years before doing the same thing again.

I could post a dozen names from my own linkedIn list that are repeat offenders in this space, but the risk of a libel suit is too high. These bastards are a cut-throat bunch.

So why people join these kind of sweat shop startups ... they will not get admission to stable big companies? On top of this I think they take huge debt for homebuying. For example this lead test engineer owned a home.
The mix of startup and debt stress becomes too much to handle I guess .... unfortunate!

@SP - LOL. Did you work at Greenfield Networks?

Two words....stock options.

Cisco Completes Acquisition of Greenfield Networks
SAN JOSE, Calif., December 7, 2006 - Cisco (NASDAQ: CSCO) - Cisco Systems® today announced it has completed the acquisition of privately-held Greenfield Networks Inc. of Sunnyvale, California. Greenfield Networks provides integrated circuits, hardware and software optimized for Ethernet packet processing that enables next-generation Metro Ethernet services

Greenfield has a proven track record in developing and deploying semiconductors for the Metro Ethernet market. This technology is highly complementary to Cisco's existing line of Metro Ethernet products.

With the close of the transaction, the Greenfield team and product portfolio are now integrated into Cisco's Ethernet and Wireless Technology Group (EWTG) led by senior vice president, Kathy Hill.

Qualcomm's success has really done a disservice to how tech is financed. I know I'll catch flack, and those who know me know that I never mean to be disrespectful of anyone but tech guys like engineers and writers are so mesmerized by 'get super rich through a short amount of hard work' schemes that they only need one Qualcomm success story and then the next thing you know you have incredibly skilled people working for $50,000 a year and a pile of stock options that way more often than not have more value as toilet paper rather than as financial instruments.

>> justme, Permarenter is hyperlink challenged.

That is correct.

>> a year and a pile of stock options that way more often than not have more value as toilet paper rather than as financial instruments.

I have had my share of toilet paper with Atoga Systems .....

At Metallic Power I received a promotion and a bonus, which was stock options. I'm like, "Wow, thanks a lot you guys." :)

In all fairness, at least I saw the writing on the wall. I actually saw some people write checks to exercise them. Talk about getting kicked in the nuts twice. They opened their legs to take it a second time.

wow three people die and one gets put away for life…

Hopefully, the DA will push for the death penalty.

Malcolm,

I think we're on the same page. There are clowns both in business and in science+engineering.

On the business side of things, I am especially weary of those that use the expression "on the business side of things".

:-)

>> justme, Permarenter is hyperlink challenged.

>That is correct.

But perhaps we can start the post with a simple attribution, even if the hyperlink does not quite make it?

I call "stock options" as "options stock". You exercise your option to stock them at your bathroom (and write the check) :-)

Malcolm,

Also good points about the various fads that are out there. Allow me some translation:

6-sigma quality: We, the manufacturing people, are the most important ones in the company.

Market Driven Quality: Hey, what about us in the marketing department, WE are the important ones. Everyone knows that.

Activities Based Costing: Hey, what about us in finance and data-mining? We are important, too, and we will save you BIG money.

....and so on and so forth. Scott Adams, if you ever read this, you have my permission to use.

SP,

Good points, but many times startups also happen because certain employees absolutely cannot get Big Company X to support their idea for a new product.

No swindle involved, necessarily.

TOB,

>>justme, you forgot the lawyers! dont ferget about them!

Do they have a trade-marked program? Or does Lawsuit-driven Quality (TM) just come naturally? :-)

TOB, right on. Wow, you guys have a good memory. I remember something along those lines. I had started at the company in 2000. I vaguely remember something along the lines of "If you do it this way you have to recognize the gains right away. If you do it this other way you recognize the gains at the time you exercise or sell them." My memory is a little fuzzy on that but I think you are talking about that same tax ramification. It might also explain how they did all of a sudden start treating them like Monopoly money.

Judicial Activisim? The thought that sticking it to the man somehow makes the world a better place.

Again, in the right circumstances, a perfectly valid point of view.

Malcolm,

Perhaps you are thinking about what is known as an "83(b) election" in tax law parlance.

http://financial-dictionary.thefreedictionary.com/Section+83(b)+Election

Justme, that's the one.

TOB, ain't that the truth. I have very specific accurate memories. I'm amazed at the detail of things I remember from so long ago. I remember incredible detail having to deal with these types of prick-fucks that we all seem to have had to endure in our careers. Some things I would rather just forget but I can't.

Funny, on the History Channel I'm watching some guy who can draw a city just from seeing it from a good vantage point for a few minutes. Very interesting.

He is a sevant. He is technically autistic.

The Indian CEO at SiPort fired the gunman.

That was a bit too mean, while I am not going to say he deserves what he got, but he obviously didn't have very high EQ. As a small startup, you never fire anyone even if that person has a poor performance. Especially when the economy is teetering on depression, if you are not providing good reference for your employee, he will most likely have a very hard time finding the next job, and on top of this, he is not eligible for unemployment benefits.

Everyone I know who has to "fire" a employee always ends up convincing him to "resign" within a couple of weeks, and promised to provide good job reference, hey no skin off our back.

Well, as a grown up you will just have to bear the consequence of cornering people. If you are mean, you are just subjecting yourself to a higher risk of getting screwed over some time down the road, it just seems the retribution of the Siport CEO is a bit too harsh.

# justme Says:
many times startups also happen because certain employees absolutely cannot get Big Company X to support their idea for a new product. No swindle involved, necessarily.

In those cases, the exit strategy is not simply "we will get bought out by Cadence/Cisco/Intel/etc." The ones I am talking about are where the pattern is very consistent. Here is the business plan in a nutshell:

Step 1. Big Company X identifies a new market, or more often an incremental improvement of existing product.

Step 2. A few people then ditch the big company and get VC funding, leaving a crater behind which Big Company has to fill again.

Step 3. Startup rushes the product to development, working their slaves to the bone.

Step 4. Startup hires a senior manager or exec from parent company, who helps with "business development".

Step 5. Startup gets acquired either by Big Company X or their competitor, Big Company Y. Founders and VC's make out big, a couple of early engineers do okay, the rest of the burnt-out employees are thankful to be now part of the big company.

Step 6. The founders of the startup poke around for the next big opportunity to go back to step 1 and steal stuff.

Somehow, this seems more common in the semiconductor and networking industries. Not so much in application software (like Oracle) or in the internet space (like Google). Acquisitions by ORCL and GOOG have a different sort of history.

I think SP is talking about someone we both know :-)

I am also perplexed by the fact that they could get away with this repetitively (the "serial" entrepreneurs), why didn't their old employer sue them for IP issues?

>> OO Says:

>> The Indian CEO at SiPort fired the gunman.

>> That was a bit too mean, while I am not going to say he deserves what he got, but he obviously didn’t have very high EQ.

OO, you got that right! Now is the time to answers Peter P survey:

====================
I want to know… IF you get laid off, what upsets you the most?

1) your finance is threatened
2) you honor is violated
3) you feel mistreated after putting in so much for the company

(2) and (3) are EQ issues.

I do blame the system for pushing people so hard on people’s finance. People should not feel that they must overpay for homes and other toys. Perhaps this is a flaw of the credit-based society.
====================

Wreckless Eric -- Take the Cash

Wreckless Eric — Take the Cash -- Well, it was a You Tube link...

PermaRenter,

I am not saying the gunman has high EQ, he obviously has some serious psycho problems.

But as a CEO, you should make allowance that among lots of people you interact with, some of them have erratically low EQ and they may go non-linear if you push them a bit hard. Who is to lose more? The gunman who is a loser to begin with, or a serial entrepreneur with a better life to enjoy?

SP,

You got the business plan right. Many 2nd tier Chindian engineers join these kind of startups to gain entry in Cadence/Cisco/Intel.....

Yes I disagree and that is why I say people do not understand ABC and then they blame it. you have given a perfect example how people use it incorrectly.

In youe example you have given how to collect the cost information for the department. First of all once this cost has been calcualted the capacity of the departments as 'total # of cheks writtne' has to be decided. This can be done putting a process on paper and the time perform the activities. Then total capcity is the total time available i.e. total time of all the people who take part in any of the activities mentioned. the supervisor's time is not taken here bu the cost is. This gives two things cost of writing the check and capacity used and unused.

Then the cost of writing chek has to be varified why it is so high or low... The price quoted by the vendor has to be addded with the costs of the activities that are still going to be there or being added due to outsourcing. If this is less than the cost of writing check inhouse then only you can think of outsourcing.

Tell me now.

Rajen