15 Aug
2005
15 Aug
'05
3:12 p.m.
Hello,
I have a purely academic question stemming from an accidental install on
my part:
Will my AMD64 laptop run PD and other sound apps (using mingo's RT
patched kernels) faster in native 64-bit linux or faster as a
conventional 32-bit linux?
I ask this because I am about to install the "correct" 64-bit linux but
I will then have array problems and other PD errors that have not yet
been solved... So if my machine is faster in 64 -bit linux the
shortcommings of 64-bit PD are outweighed by the speed gain.
Thanks all!
-thewade
16 Aug
16 Aug
5:26 a.m.
On Mon, Aug 15, 2005 at 01:11:59 -0700, thewade wrote:
Hello,
I have a purely academic question stemming from an accidental install on
my part:
Will my AMD64 laptop run PD and other sound apps (using mingo's RT
patched kernels) faster in native 64-bit linux or faster as a
conventional 32-bit linux?
Unless you've got more than 3 gig of ram I doubt it will make much
difference. 32bit binaries are smaller, and hence faster, but there are a
couple of 64bit only instructions that the compiler might use that could
give some speed benefit.
If you try it, why dont you measure the difference and tell us?
Personally I'd be supprised if the 64bit version is measurably faster, but
I'm often supprised :)
- Steve
5:59 a.m.
Hi
I have been running 64 on my Audio pc for a few months now and would say
it seems
to cruise along better re CPU usage but as I dont know how to measure
this I cant give you
a total.
Just seems faster.
Bob
Steve Harris wrote:
On Mon, Aug 15, 2005 at 01:11:59 -0700, thewade wrote:
Hello,
I have a purely academic question stemming from an accidental install on
my part:
Will my AMD64 laptop run PD and other sound apps (using mingo's RT
patched kernels) faster in native 64-bit linux or faster as a
conventional 32-bit linux?
Unless you've got more than 3 gig of ram I doubt it will make much
difference. 32bit binaries are smaller, and hence faster, but there are a
couple of 64bit only instructions that the compiler might use that could
give some speed benefit.
If you try it, why dont you measure the difference and tell us?
Personally I'd be supprised if the 64bit version is measurably faster, but
I'm often supprised :)
- Steve
6:09 a.m.
On Tue, Aug 16, 2005 at 10:58:46AM +0100, studio-64 wrote:
Hi
I have been running 64 on my Audio pc for a few months now and would say
it seems
to cruise along better re CPU usage but as I dont know how to measure
this I cant give you
a total.
Fair enough. The JACK CPU usage %age is probably as good a measure as any
FWIW.
- Steve
6:48 a.m.
Good point I'll check that from now on and if I see a difference I'll
post here, as a real world example.
cheers
Bob
Steve Harris wrote:
On Tue, Aug 16, 2005 at 10:58:46AM +0100, studio-64
wrote:
Hi
I have been running 64 on my Audio pc for a few months now and would say
it seems
to cruise along better re CPU usage but as I dont know how to measure
this I cant give you
a total.
Fair enough. The JACK CPU usage %age is probably as good a measure as any
FWIW.
- Steve
24 Aug
24 Aug
1:53 p.m.
On Tue, Aug 16, 2005 at 10:25:15AM +0100, Steve Harris wrote:
Unless you've got more than 3 gig of ram I doubt
it will make much
difference. 32bit binaries are smaller, and hence faster, but there are a
couple of 64bit only instructions that the compiler might use that could
give some speed benefit.
If you try it, why dont you measure the difference and tell us?
Personally I'd be supprised if the 64bit version is measurably faster, but
I'm often supprised :)
You wouldn't expect the increase in register count to make a fairly
significant difference?
--
Joshua D. Boyd
jdboyd(a)jdboyd.net
http://www.jdboyd.net/
http://www.joshuaboyd.org/
2:40 p.m.
Joshua Boyd <jdboyd(a)jdboyd.net> writes:
On Tue, Aug 16, 2005 at 10:25:15AM +0100, Steve Harris
wrote:
More registers generally help, but so does a smaller cache footprint.
Which matters more depends very much on exactly what you are doing.
--
joq
Unless you've got more than 3 gig of ram I
doubt it will make much
difference. 32bit binaries are smaller, and hence faster, but there are a
couple of 64bit only instructions that the compiler might use that could
give some speed benefit.
If you try it, why dont you measure the difference and tell us?
Personally I'd be supprised if the 64bit version is measurably faster, but
I'm often supprised :)
You wouldn't expect the increase in register count to make a fairly
significant difference?
6853
days inactive
6862
days old
linux-audio-user@lists.linuxaudio.org
6 comments
6 participants
participants (6)
-
Bob -
Jack O'Quin -
Joshua Boyd -
Steve Harris -
studio-64
-
thewade