( ESNUG 453 Item 4 ) --------------------------------------------- [03/01/06]
Subject: Luminescent, Takumi, Blaze DFM, Nannor, Sigma-C
Q: As a competitor, what do you think about these specific DFM companies?
In plain English, what exactly do they do? And how do you view them?
Luminescent
Mike -- Inverse lithography concepts have been around for a while. Can we
really boil the ocean? Can Luminescent really get it right and produce a
scalable product? Surmounting that formidable technical challenge, along
with getting a risk- and change-averse manufacturing community to change
everything tells me it will be long road for these guys. If they can keep
money in the bank, there will be a steady stream of tire kickers at least.
They will be around because of the tire kickers.
Atul -- One of Clear Shape's founders, Yao-Ting Wang, did his PhD thesis
at Stanford on inverse lithography, so yes it's an old idea. It's very
different than what the other DFM companies are working on. I think it's
probably a bridge too far. It depends on substantial changes in mask
manufacturing and inspection. There's less burden on OPC, but the mask
becomes more challenging to make. I'm not sure if the mask makers will
sign up for that. But anyway I'll keep watching Luminescent because
they're such a different approach.
Takumi
Mike -- It's a flow/integration play. Some good ideas regarding hot spot
identification in your layout and how to use the information in Proteus and
Calibre OPC. But flow, framework and integration plays don't have a good
track record in this market. They'll either get squashed by the Big 4 or
go irrelevant as users figure out how to do this on their own.
Atul -- I think that Mike is giving them too much credit. As far as I know
they're only providing OPC services in Japan.
Blaze DFM
Mike -- They're trying to optimize individual transistor performance by
modifying they layout through interaction with OPC. They claim they can
reduce leakage; which is a huge benefit. The concern that I have is the
breath of their tool. I think Blaze is stuck until they get integrated
into a broader design flow. One of the Big 4 will probably buy them.
They're not on my list; they'll be part of someone else in 3 years.
Atul -- They create a lot of standard cell variation, which means you have
unworkable libraries. People don't like this. Mike is right, trying to
do this in isolation won't cut it. They're not on my list to survive.
Nannor
Mike -- They modify pre-OPC layouts to improve manfacturability. It's an
important function, but it's a point tool. It needs to be teamed up with
a bigger flow. I think the Big 4 will roll their own Nannor equivalent
so they're in a tough spot.
Atul -- I have to be honest; Nannor is not on my radar.
Sigma-C
Mike -- They have a very accurate litho simulator that will predict what
the circuit will look like in silicon. I believe they claim they can do
65 nm, but they're only in my peripheral vision. KLA-Tencor has a product
called "Prolith" which does the same thing as Sigma-C and my money is on
KLA here. Sigma-C is strong nichey product with strong competition.
Atul -- This is the last one of the mid 1990's litho companies that is
still single (i.e. all the others either got acquired, IPOed, or failed.)
They have a very good niche in process R&D. They're very well regarded.
In the mid-90's they were on my list to watch; not now. They're post-GDSII
litho simulation; it's not true DFM.
Index
Next->Item
|
|
