is it easy to make a fabless company !!

[DZC]

Absolutely not!

 

[okguy]

Everything depends on your skills and network !

Network is to find customers : no customer = you are dead
Skills are needed to be different :idea:

Power management is an easy entry, so .... many people have used that idea! good luck to find reliable and huge customers!
If you think for digital, 45nm mask need 2millions$ !!!

Have a good idea and some good relationship then you have it :!:

 

[Pat_Mustard]

In a word, No. Electronics is a very tricky business to get establshed in due to high initial costs and relative lack of good people to go round. Milk delivery is a lot easier.

 

[shashikumar.22]

its will be easy only if you have enough money to sustain the loss if it sinks :|

 

[leo_o2]

What's the best market for IC fabless design house currently?

 

[Teddy]

I would say it is easy. Many companies are fabless because they can not afford the high cost of running the fab.
Based on on the question I expect 1 perhaps 2,3 products. That would never fill up the fab - it would be cheaper to be fabless and run in foundry.
What the question totaly skips are the steps after design and fabrication of IC.
Where will you do packaging? As fabless - outsource.
Qualification? Final Test? Burn in, ESD ??? It all could be done by 3rd party. But if you have fab you do all that inhouse.

So as a startup which wants to make product and bebought by some bigger company - PC and soft would cut it.
To market the product, run volume manufacturing, support, applications etc - good luck with pc and 10 people. It would be tough - not impossible.

Offer me a job and we'll do it.

 

[llbaobao]

Colbhaidh just said excellent words! I think what he said is right! In a common sense!

The following is my own words:
I see some fabless comanies with about less than 5 persons! If there are about 10 persons, then it consumes too much money a year!
Some fabless companies just begun with 1 projects, if they can not finish the project good enough in a short period(for example, 1 year), then without enough money, most of them will bankrupt. Little persons can stick on when the project proceeds not soothly and at the same time there is not enough money as support.

 

[lchen]

i am also wondering, $15K, so cheap...

Well .. he mentioned 0.5u .. I assume it's obsolete .. so, wherever you find some fab doing that, it might be enough to ask for this amount of money ..

we're from chinese ,we just start with 1000k RMB at 2007 ,now we operate this corp for three years ,now have 7 chips on the line ,0.35 um CMOS process in SMIC ,and HJTC.

 

[ahmed osama]

i am also wondering, $15K, so cheap...

Well .. he mentioned 0.5u .. I assume it's obsolete .. so, wherever you find some fab doing that, it might be enough to ask for this amount of money ..

we're from chinese ,we just start with 1000k RMB at 2007 ,now we operate this corp for three years ,now have 7 chips on the line ,0.35 um CMOS process in SMIC ,and HJTC.

can we know in what app. such 7 chips??
thx

 

[elecrom]

I suppose it is all relative:

It costs about 2-4 Billion to build a Fab and about 4 million a month to run it
............
............
But life would be so dull if it were easy .................;

Excellent, enjoyed it. First time I read such a short but comprehensive overview of entire process with "real life" difficulties

 

[deepak242003]

well it depends.. rislk is more.. but eventually profit is also more....

remem " NO PAIN NO GAIN"

 

[RFDE]

Dear Mr.Colbhaidh,

you tell the true story.
but for a baby company, dont you think start with a RF product is a high risk choice?
if not enough seed money, find a low risk project for startup, mabye 0.5um process(mask+16wafer first run $15K) is more comfortable.

Hi butterfish
Where did you get this number ($15k) from ? .. I am preparing a business plan and I indeed was searching for similar numbers.

i think that as you don`t target state of the art technologies , then tape out costs will be much lower , AMS and silterra may be a good choice for low cost tape-outs