In EEdesign.com today, Dylan McGrath covered one of those Wall Street
financial analysts get togethers where they chatted up EDA; in this case it
was analyst Dennis Wassung of Canaccord Adams hosting the kaffee klatsch.
The topic du jour was DFM.  What I got from Dylan's report was that Wassung
claimed the current DFM market was:

    manufacturing-process-aware chip design tools:   $50 million
             RET, OPC, PSM stuff used by the fabs:  $150 million
   process characterization, yield analysis, TCAD:  $160 million

which totals to $360 million.  In direct contrast, Joe Sawicki, the VP/GM of
DFM at Mentor claimed that the total DFM market was actually $700 in 2005.
Sawicki broke down the 2005 DFM market to:

              Mentor:  ############################### 31%
            Synopsys:  ################################# 33%
             Cadence:  ################## 18%
              others:  ################## 18%

The really ballsy thing happened when Sawicki predicted that DFM was going
to grow an additional $800 in 4 years to a total of $1.5 billion.

Why I see it as ballsy is that 4 years isn't that far off.  For the longest
time, the EDA market has hovered around $4 billion total for year -- a fact
that has so pissed off the EDAC people that they've been busting their
collective humps to do anything they could to break past that $4 billion.

So far EDAC has only come up with adding "services" to the EDA pie, but
6 months ago Mike Santarini of EDN gave the EDAC people a raft of shit when
they recently added "IP sales" to that now infamous $4 billion EDA total.

Sawicki is like Paul on the road to Damascus, presenting a vision of $800
million of non-bullshit EDA growth.  That's one ballsy prediction.  And its
in an industry that's brutal to false prophets: remember when Ron Collett
predicted that VHDL was going to overtake Verilog in 1992?  (ESNUG 316 #16)