Re: NUMERIC type benchmarks - CORRECTED - Mailing list pgsql-hackers

From Tom Lane
Subject Re: NUMERIC type benchmarks - CORRECTED
Date
Msg-id 29736.987384489@sss.pgh.pa.us
Whole thread Raw
In response to Re: NUMERIC type benchmarks - CORRECTED  (Mark Butler <butlerm@middle.net>)
List pgsql-hackers
Mark Butler <butlerm@middle.net> writes:
> ... The correct numbers should be:

> Postgres PL/PGSQL original numeric:    14.8 seconds
> Postgres PL/PGSQL modified numeric:    14.0 seconds
> Postgres PL/PGSQL float8:              10.7 seconds
> GNU AWK:                                2.5 seconds
> Oracle PL/SQL number:                   2.0 seconds

> This means that Tom Lane was absolutely right - for the current numeric type
> implementation, palloc() overhead is not a dominant concern.  A serious
> solution needs to change the internal format to use a larger base, as Tom
> suggested.

What do you get if you use int4 in PL/PGSQL?  The above numbers look to
me like the real problem may be PL/PGSQL interpretation overhead, and
not the datatype at all...
        regards, tom lane


pgsql-hackers by date:

Previous
From: Mark Butler
Date:
Subject: Int64 (long long) Supporting Compiler Requirement Status?
Next
From: Lincoln Yeoh
Date:
Subject: Re: Hey guys, check this out.