By: Richard Cownie (tich.delete@this.pobox.com), February 7, 2013 6:48 am
Room: Moderated Discussions
geeker (geeker.delete@this.mailinator.com) on February 6, 2013 12:16 pm wrote:
> Another question, does Intel and TSMC (or GF for that matter) depreciate their foundry equipment at the same
> rate? If not, then even the 58 47 difference does not capture the real difference between their respective
> margins. TSMC maybe living with a much lower margin than Intel, leaving more on the table for the ARM camp.
A very good question. "Profit" is a bit hard to judge for a business with very high
capital investments in equipment with a fairly short useful lifetime. The exact rules
for depreciation (and the way the company chooses to interpret those rules) could
make a big difference. It might also be interesting to look at cashflow - which is
unambiguous - over each of the last 5-10 years for each of the businesses.
I haven't gone into those financial figures in detail, but I have a sneaking suspicion
that the semiconductor industry might eventually end up a bit like the airline industry, where
the combination of huge capital investments in airplanes, competition, and vulnerability
to slumps in demand and fluctuations in expenses, leave the airline industry in deficit
over its whole existence. But people (and to some extent governments) are still willing to
run airlines, and invest in airlines, because it is (or used to be) a glamorous high-tech
business, and the existence of air travel has indirect economic benefits.
> Another question, does Intel and TSMC (or GF for that matter) depreciate their foundry equipment at the same
> rate? If not, then even the 58 47 difference does not capture the real difference between their respective
> margins. TSMC maybe living with a much lower margin than Intel, leaving more on the table for the ARM camp.
A very good question. "Profit" is a bit hard to judge for a business with very high
capital investments in equipment with a fairly short useful lifetime. The exact rules
for depreciation (and the way the company chooses to interpret those rules) could
make a big difference. It might also be interesting to look at cashflow - which is
unambiguous - over each of the last 5-10 years for each of the businesses.
I haven't gone into those financial figures in detail, but I have a sneaking suspicion
that the semiconductor industry might eventually end up a bit like the airline industry, where
the combination of huge capital investments in airplanes, competition, and vulnerability
to slumps in demand and fluctuations in expenses, leave the airline industry in deficit
over its whole existence. But people (and to some extent governments) are still willing to
run airlines, and invest in airlines, because it is (or used to be) a glamorous high-tech
business, and the existence of air travel has indirect economic benefits.