|[7 earlier articles]|
|Re: Code quality email@example.com (1993-01-07)|
|Re: Code quality firstname.lastname@example.org (1993-01-07)|
|Re: Code quality grover@brahmand.Eng.Sun.COM (1993-01-07)|
|Re: Code quality email@example.com (1993-01-08)|
|Re: Code quality polstra!jdp@uunet.UU.NET (1993-01-12)|
|Re: Code quality firstname.lastname@example.org (1993-01-13)|
|Re: Code quality email@example.com (1993-01-25)|
|Re: Code quality firstname.lastname@example.org (1993-02-01)|
|From:||email@example.com (Andy Glew)|
|Organization:||Intel Corp., Hillsboro, Oregon|
|Date:||Mon, 25 Jan 1993 07:59:06 GMT|
firstname.lastname@example.org (Dale R. Worley) writes:
>How important is generated code quality these days? ...
>Is there much of a market for another 10% in speed of generated code?
>From a hardware company's point of view:
10% is *definitely* interesting.
Even 1% wins are interesting, because if you get enough of them...
Modern microprocessors of comparable price sit in a range of maybe 40%.
10% is a significant chunk of your competitive advantage. It's equivalent
to a few months of hardware development. If you don't go for the
optimizing compiler, your competitor will.
Moreover, compilers have the advantage of being able to be released
*between* silicon releases - giving a midlife kicker to your product.
On the other hand:
I've heard of software companies for which compile time is everything, who
refuse to add optimizations that will make -O more than twice as slow as
The marketplace takes all kinds.
Andy Glew, email@example.com
Intel Corp., M/S JF1-19, 5200 NE Elam Young Pkwy,
Hillsboro, Oregon 97124-6497
Return to the
Search the comp.compilers archives again.