In doing my end-of-year stats for DeepChip, I thought it would be interesting
to see how those job ads did.  Only looking at the job ads that had been up
for at least 120 days, here's what I found:

          A     State & Company    Job
        -----   ---------------    ----------------------------
        5,961   CA Magma           EDA R&D developer
        4,249   AZ Intel           senior DA support engineer
        3,599   CA Cadence         Open Access product engineer
        3,364   OR Mentor          Catapult tech mktg engineer
        2,673   CA Lightspeed      senior IC design engineer
        2,505   CA Knowlent        EDA R&D developer
        2,498   CA Berkeley DA     senior applications engineer
        2,340   CA Synplicity      logic synthesis R&D engineer
        2,267   TX Denali          senior verification engineer
        2,248   CA Atheros         digital design engineer
        1,787   CA ClearShape      senior apps engineer
        1,866   CA Athena Design   EDA R&D developer
        1,564   CA True Circuits   circuit design engineer
        1,536   MA Dafca           senior applications engineer
        1,471   CA IC Manage       EDA development engineer
        1,445   CA Zenasis         applications engineer
        1,412   CA Jasper          corporate apps engineer
        1,240   CA Ponte           senior field apps engineer
        1,183   CA Blaze DFM       senior software engineer

               A = Total Pageviews over life of Ad

I was happy to see that an average job ad on DeepChip got 2,380 pageviews.
The typical numbers I had heard for the big job boards like monster.com and
dice.com were anywhere from 150 to 500 clicks per tech job.

But these numbers shocked me.  To get close to 6,000 pageviews for a Magma
job ad to be one of their developers?  What the!?  That doesn't make sense!
I know for a fact that there aren't 6,000 software engineers in the entire
EDA industry.  (I remember Gary Smith once saying the world total ballparked
around 5K.)  So what gives?

I dove into the web stats deeper and found another, bigger surprize.  You'd
think if you ran an ad for zoo keepers, at least 90% of the people clicking
on the job description would be from other world zoos, right?  After all,
who looks at nurse job ads besides nurses, right?  Plumbers look at plumber
job ads.  Lawyers look at lawyer ads.  You get the pattern.

But when I dove into the specific stats for that Magma R&D developer ad, only
19% (1,132) of its viewers were from EDA companies -- a whopping 81% (4,829)
were from EDA *user* companies like Intel, TI, LSI Logic, IBM, Infineon, AMD,
ST, Freescale, TSMC, ARM, Philips, HP, Toshiba, Agere, Chartered, Samsung,
ADI, Nvidia, Renasas, NEC, Marvell, Cisco, Xilinx, Nokia, Conexant, Fujitsu,
Transmeta, IDT, Sun, KLA-Tencor, Motorola, UMC, Nortel, AMCC, Cypress, Starc,
Raza Micro, Cray, Northrop Grumman, On Semi, Rambus, Unisys, Vitesse, NASA,
Raytheon, Honeywell, PMC-Sierra, and Altera to name a few.  The other EDA job
ads had a similar weird break; Cadence had 23% (828) and Mentor had 18% (606)
viewers from EDA companies -- with the bulk being from that long laundry list
of EDA buying companies.  The bulk!

That is, lots of EDA users sniff around in the EDA vendor job ads just to
keep tabs on how a specific EDA vendor is doing!  Those nosey mothers!  :)

But the opposite isn't true.  For any typical link on DeepChip, readership is
on average 11% EDA vendors, with the remaining 89% being mostly EDA users.
(Close to 300 Wall Street weenies also hang around, ESNUG 449 #7, but they're
an insignificant part of the 25,000 total readers.)  Generally, the rule of
thumb is if you're an EDA vendor reading something on ESNUG, there's at
least 9 other EDA users also reading exactly what you're reading on ESNUG.
(This rule only falls apart when I do one of those massive Cenuses that the
world comes looking at.  For example, the 2005 Synopsys Census generated a
whopping 86,277 pageviews; EDA vendors viewing it were far less than 11%.)

Anyway...  I browsed the stats for the EDA user jobs: Intel was 12% (510),
Lightspeed 11% (294), Atheros 11% (247) viewed by EDA vendors; which is right
in line with the standard 11% 1-to-9 rule of thumb.

Simply put, the EDA vendors seem to be fairly neutral about their customers;
the vendors were neither extra nosey nor ignoring them -- just neutral.

In short, Intel may spy on Magma, but Cadence is indifferent to Infineon.