( ESNUG 453 Item 3 ) --------------------------------------------- [03/01/06]
Subject: Aprio, Clear Shape, KLA-Tencor, Brion, Invarium
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?
Aprio
Mike -- We make designer-friendly mask data prep tools. We've built an
incremental OPC engine that lets a designer reuse OPC results. Aprio
can generate the OPC results or you can use Proteus or Calibre OPC. It
lets you take divide-and-conquer approach to mask data prep. Halo gives
the user OPC and litho sim. It's fast because its incremental. It's
accurate because it uses the same models the fab uses. It allows the
designer to quickly see what the true final silicon image will be.
Atul -- I would say that Mike has impacted the company positively since
he came on board. There's no question there. Early on they were doing
OPC fixing, now I think they need to move into the pre-GDSII design. It
remains to be seen what Aprio is going to do there.
Clear Shape
Mike -- They have a litho simulator that predicts what a circuit will look
like in silicon. Their model is high level abstraction and it doesn't use
any OPC or litho simulation engines. I agree with the concept. I disagree
on implementation. Can you achieve high accuracy with high abstraction?
I don't know enough about it to determine if it's workable or not; but
this tradeoff is a key question in my mind. Their strategy of providing
upstream data to the designer is right.
Atul -- Mike missed that we use the prediction of what a circuit looks
like in silicon to account for changes in timing, SI, and leakage power.
We integrate the fabs' CMP models to account for thickness variation and
are talking to the Praesagus people about their tools.
KLA-Tencor
Mike -- KLA has already announced the DesignScan product, which inspects
GDSII, and not physical glass. They have deep expertise in modeling and
simulation using special purpose HW. They're very good at that. This is
why I think they'll do well over time.
Atul -- I am looking to KLA-Tencor and Applied Materials to make some
moves in the DFM space. So far they have been relatively quiet but their
deep pockets and fab access puts them in a unique position. They have lots
of yield experience. I have them on my long term watch list.
Brion
Mike -- Hit the market at the right time with the right product. They're
doing mask level verification with a lithography simulator in a HW box.
Working with TSMC to develop their verification tool was also a good move.
But special-purpose hardware shrivels and dies without constant performance
improvement and a steady stream of new applications for it. It a point
tool, not a solution. Can Brion get that done? I don't think they can,
and therefore they're not going to survive.
Atul -- They're doing post-OPC verification which is already entirely owned
by Mentor and Synopsys. They're using FPGA based HW. Brion is hoping to
scale this into OPC, but I personally don't bet on proprietary HW for more
than 1/2 a generation.
Invarium
Mike -- Claims they can do OPC across the process window on off-the-shelf
Linux PCs. Too good to be true? Where I grew up, performance and accuracy
always had tradeoffs. There's no free lunch. There must be a compromise
somewhere, I just don't see it. They're not on my list of top DFM co's.
Atul -- They provide OPC services based on technology that might not be
applicable in the mainstream. They do post-GDSII OPC. Invarium started
marketing some tools and then pulled back to services. They're not on my
list either.
Index
Next->Item
|
|
