[gentoo-amd64] 3200+ -> X2 4000+

[gentoo-amd64] 3200+ -> X2 4000+

[Fernando Boaglio]

Hi guys,

I bought a new X2 4000 processor and I have 2 questions:

-- can I use the same cooler (my processor uses the deprecated Socket 939) ?

-- should I change a lot of attributes in my kernel .config file?

TIA

[]'s
Fernando Boaglio
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Nuitari]

> I bought a new X2 4000 processor and I have 2 questions:
>
> -- can I use the same cooler (my processor uses the deprecated Socket 939) ?

Maybe, depends on how good your cooler is.

> -- should I change a lot of attributes in my kernel .config file?

You need to enable SMP and the multi-core scheduler option
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Tim Allingham]

[-- Attachment #1: Type: text/plain, Size: 589 bytes --]

If its the cooler that came with the 3200+ then no, don't use it for a
dual-core chip. even though most of the times temps will be fine, when
loading both cores heavily it would really strain.

On Wed, 2007-06-27 at 23:00 -0400, Nuitari wrote:
> > I bought a new X2 4000 processor and I have 2 questions:
> >
> > -- can I use the same cooler (my processor uses the deprecated Socket 939) ?
> 
> Maybe, depends on how good your cooler is.
> 
> > -- should I change a lot of attributes in my kernel .config file?
> 
> You need to enable SMP and the multi-core scheduler option

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Donnerstag, 28. Juni 2007, Fernando Boaglio wrote:
> Hi guys,
>
> I bought a new X2 4000 processor and I have 2 questions:
>
> -- can I use the same cooler (my processor uses the deprecated Socket 939)
> ?

no. Btw, the if you buy 'the real thing' the dual cores come with a nice 
self-regulating cooler. And this boxed cpus are not much more expensive than 
a single cpu.

>
> -- should I change a lot of attributes in my kernel .config file?

enable smp. The rest should be fine. You can but don't have to enable 
multicore support. Don't enable numa. Have acpi enabled.
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Florian Philipp]

[-- Attachment #1: Type: text/plain, Size: 873 bytes --]

Am Donnerstag 28 Juni 2007 08:05 schrieb Hemmann, Volker Armin:
> On Donnerstag, 28. Juni 2007, Fernando Boaglio wrote:
> > Hi guys,
> >
> > I bought a new X2 4000 processor and I have 2 questions:
> >
> > -- can I use the same cooler (my processor uses the deprecated Socket
> > 939) ?
>
> no. Btw, the if you buy 'the real thing' the dual cores come with a nice
> self-regulating cooler. And this boxed cpus are not much more expensive
> than a single cpu.
>
> > -- should I change a lot of attributes in my kernel .config file?
>
> enable smp. The rest should be fine. You can but don't have to enable
> multicore support. Don't enable numa. Have acpi enabled.

Shouldn't he set the timer frequency to 250 or 300 as well?

BTW: I'm still not sure if he talks about the boxed cooler or something he 
bought separately.  The latter should do fine...

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Daniel Gryniewicz]

On Thu, 2007-06-28 at 12:16 +0200, Florian Philipp wrote:
> Am Donnerstag 28 Juni 2007 08:05 schrieb Hemmann, Volker Armin:
> > On Donnerstag, 28. Juni 2007, Fernando Boaglio wrote:
> > > Hi guys,
> > >
> > > I bought a new X2 4000 processor and I have 2 questions:
> > >
> > > -- can I use the same cooler (my processor uses the deprecated Socket
> > > 939) ?
> >
> > no. Btw, the if you buy 'the real thing' the dual cores come with a nice
> > self-regulating cooler. And this boxed cpus are not much more expensive
> > than a single cpu.
> >
> > > -- should I change a lot of attributes in my kernel .config file?
> >
> > enable smp. The rest should be fine. You can but don't have to enable
> > multicore support. Don't enable numa. Have acpi enabled.
> 
> Shouldn't he set the timer frequency to 250 or 300 as well?

Not if it's going to be a desktop.  Hires timers are not common in
x86_64 in the kernel yet (no tickless support) so for now 1000 is still
best for desktops.

Daniel

-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Donnerstag, 28. Juni 2007, Daniel Gryniewicz wrote:

> >
> > Shouldn't he set the timer frequency to 250 or 300 as well?
>
> Not if it's going to be a desktop.  Hires timers are not common in
> x86_64 in the kernel yet (no tickless support) so for now 1000 is still
> best for desktops.

in my experience 300 is 'best' for my desktop.
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Beso]

[-- Attachment #1: Type: text/plain, Size: 1636 bytes --]

2007/6/28, Daniel Gryniewicz <dang@gentoo.org>:
>
> On Thu, 2007-06-28 at 12:16 +0200, Florian Philipp wrote:
> > Am Donnerstag 28 Juni 2007 08:05 schrieb Hemmann, Volker Armin:
> > > On Donnerstag, 28. Juni 2007, Fernando Boaglio wrote:
> > > > Hi guys,
> > > >
> > > > I bought a new X2 4000 processor and I have 2 questions:
> > > >
> > > > -- can I use the same cooler (my processor uses the deprecated
> Socket
> > > > 939) ?
> > >
> > > no. Btw, the if you buy 'the real thing' the dual cores come with a
> nice
> > > self-regulating cooler. And this boxed cpus are not much more
> expensive
> > > than a single cpu.
> > >
> > > > -- should I change a lot of attributes in my kernel .config file?
> > >
> > > enable smp. The rest should be fine. You can but don't have to enable
> > > multicore support. Don't enable numa. Have acpi enabled.
> >
> > Shouldn't he set the timer frequency to 250 or 300 as well?
>
> Not if it's going to be a desktop.  Hires timers are not common in
> x86_64 in the kernel yet (no tickless support) so for now 1000 is still
> best for desktops.
>
> Daniel



for what i know on dual core 1000 isn't suited cause it makes the system
unstable.... the suggested is 100 or max 250 if you want a low latency
desktop.... also, the 1000 on single core 64bit is very unsable and the
processore state has a very great shift around.... i experimented that
single core amd64 should have the best timer freq or 250 for a normal
desktop and a dual core amd64 should have 100.... but anyone is free to try
and use other freqs as they like most....

--
> gentoo-amd64@gentoo.org mailing list
>
>


-- 
beso

d-_-b

[-- Attachment #2: Type: text/html, Size: 2304 bytes --]

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Aleksey V. Kunitskiy]

[-- Attachment #1: Type: text/plain, Size: 363 bytes --]

When using QEMU it recommends to setup timer to 1000Hz for better guest system 
response. So, I followed the recommendation and my system is stable, though I 
use masked kernel(2.6.21-gentoo-r3). Oh,I forgot - my box is driven by AMD 
Athlon X2 4400+ 

-- 
best regards,
Aleksey V. Kunitskiy
my public GPG/PGP key: http://www.alexey-kv.org.ua/pubkey.asc

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Joshua Hoblitt]

[-- Attachment #1: Type: text/plain, Size: 643 bytes --]

I recently attempted to compare 250 vs. 1000 hz for running Eve-online
under wine and non imperically the 1khz setting 'seems' to have fewer
and sorter lurches then 250hz.

-J

--
On Thu, Jun 28, 2007 at 09:15:28PM +0300, Aleksey V. Kunitskiy wrote:
> When using QEMU it recommends to setup timer to 1000Hz for better guest system 
> response. So, I followed the recommendation and my system is stable, though I 
> use masked kernel(2.6.21-gentoo-r3). Oh,I forgot - my box is driven by AMD 
> Athlon X2 4400+ 
> 
> -- 
> best regards,
> Aleksey V. Kunitskiy
> my public GPG/PGP key: http://www.alexey-kv.org.ua/pubkey.asc



[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Daniel Gryniewicz]

On Thu, 2007-06-28 at 19:45 +0200, Beso wrote:

> 
> 
> for what i know on dual core 1000 isn't suited cause it makes the
> system unstable.... the suggested is 100 or max 250 if you want a low
> latency desktop.... also, the 1000 on single core 64bit is very
> unsable and the processore state has a very great shift around.... i
> experimented that single core amd64 should have the best timer freq or
> 250 for a normal desktop and a dual core amd64 should have 100.... but
> anyone is free to try and use other freqs as they like most.... 
> 

FTR, I've run nothing but 1000 (except for testing) for a very long time
now on all my amd64 boxes (single, dual-core, dual CPU and dual-core
dual CPU) with no stability problems. In every test I've done, response
and latency was better for 1000, and throughput was not measurably
worse.  I cannot recommend anything but 1000 for a desktop system
(unless you can go tickless).

Obviously, for something like this, it's largely subjective.  That's why
the option is there.

Daniel


-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Beso]

[-- Attachment #1: Type: text/plain, Size: 1976 bytes --]

2007/6/28, Fernando Boaglio <boaglio@gmail.com>:
>
> Hi guys,
>
> I bought a new X2 4000 processor and I have 2 questions:
>
> -- can I use the same cooler (my processor uses the deprecated Socket 939)
> ?


don't use the same cooler, cause it can break the cpu.... buy a x2
cool&quiet  cooler and update the bios....

-- should I change a lot of attributes in my kernel .config file?


the configuration under processor types and features should be:
Processor family (AMD-Opteron/Athlon64)  --->
< > /dev/cpu/microcode - Intel CPU microcode support
<*> /dev/cpu/*/msr - Model-specific register support
<*> /dev/cpu/*/cpuid - CPU information support
[*] Symmetric multi-processing support
 [ ]   SMT (Hyperthreading) scheduler support (NEW)
[*]   Multi-core scheduler support
(NEW)
  Preemption Model (Voluntary Kernel Preemption (Desktop))  --->
[*] Non Uniform Memory Access (NUMA) Support
 [*]   Old style AMD Opteron NUMA detection (NEW)   -> this one could be
needed for old 939 sockets but i don't know if the x2 works in that socket
[*] ACPI NUMA detection (NEW)
[*] NUMA emulation
 Memory model (Discontiguous Memory)  --->
 [*] Provide RTC interrupt
(2) Maximum number of CPUs (2-256) (NEW)
[*]   Intel MCE features
[*]   AMD MCE features
Timer frequency (100 HZ)  --->
[*] Function reordering

then enable all the cpu scaling governors under power management option, amd
powernow directly in the kernel (not as module, cause the udev would not
load it at boot time) and add cpufrequtils as a boot service with the
conservative module....
obvioulsy you'll need thermal, fan, processor and ac adapter.... i compiled
this in the kernel support so that i wouldn't need an initramdisk.... but
this is your choice on how to compile them....
after that you should look at http://gentoo-wiki.com/ on howtos for the use
flags, cause you could have some new use flags to use....

TIA
>
> []'s
> Fernando Boaglio
> --
> gentoo-amd64@gentoo.org mailing list
>
>


-- 
besod-_-b

[-- Attachment #2: Type: text/html, Size: 3402 bytes --]

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Simon Cooper]

whats with the NUMA stuff?

Beso wrote:
> 2007/6/28, Fernando Boaglio <boaglio@gmail.com>:
>>
>> Hi guys,
>>
>> I bought a new X2 4000 processor and I have 2 questions:
>>
>> -- can I use the same cooler (my processor uses the deprecated Socket 
>> 939)
>> ?
> 
> 
> don't use the same cooler, cause it can break the cpu.... buy a x2
> cool&quiet  cooler and update the bios....
> 
> -- should I change a lot of attributes in my kernel .config file?
> 
> 
> the configuration under processor types and features should be:
> Processor family (AMD-Opteron/Athlon64)  --->
> < > /dev/cpu/microcode - Intel CPU microcode support
> <*> /dev/cpu/*/msr - Model-specific register support
> <*> /dev/cpu/*/cpuid - CPU information support
> [*] Symmetric multi-processing support
> [ ]   SMT (Hyperthreading) scheduler support (NEW)
> [*]   Multi-core scheduler support
> (NEW)
>  Preemption Model (Voluntary Kernel Preemption (Desktop))  --->
> [*] Non Uniform Memory Access (NUMA) Support
> [*]   Old style AMD Opteron NUMA detection (NEW)   -> this one could be
> needed for old 939 sockets but i don't know if the x2 works in that socket
> [*] ACPI NUMA detection (NEW)
> [*] NUMA emulation
> Memory model (Discontiguous Memory)  --->
> [*] Provide RTC interrupt
> (2) Maximum number of CPUs (2-256) (NEW)
> [*]   Intel MCE features
> [*]   AMD MCE features
> Timer frequency (100 HZ)  --->
> [*] Function reordering
> 
> then enable all the cpu scaling governors under power management option, 
> amd
> powernow directly in the kernel (not as module, cause the udev would not
> load it at boot time) and add cpufrequtils as a boot service with the
> conservative module....
> obvioulsy you'll need thermal, fan, processor and ac adapter.... i compiled
> this in the kernel support so that i wouldn't need an initramdisk.... but
> this is your choice on how to compile them....
> after that you should look at http://gentoo-wiki.com/ on howtos for the use
> flags, cause you could have some new use flags to use....
> 
> TIA
>>
>> []'s
>> Fernando Boaglio
>> -- 
>> gentoo-amd64@gentoo.org mailing list
>>
>>
> 
> 

-- 
Change the world - move a rock

GnuPG Key: http://thecoop.me.uk/gpgkey.asc
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Donnerstag, 28. Juni 2007, Simon Cooper wrote:
> whats with the NUMA stuff?

nothing. If you only have once cpu with two cores, don't use it!

-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Dieter Ries]

Hi,


Hemmann, Volker Armin schrieb:
> On Donnerstag, 28. Juni 2007, Simon Cooper wrote:
>> whats with the NUMA stuff?
> 
> nothing. If you only have once cpu with two cores, don't use it!

What exactly does that mean? None of those :

     [*] Non Uniform Memory Access (NUMA) Support
     [*]   Old style AMD Opteron NUMA detection
     [*] ACPI NUMA detection
     [*] NUMA emulation

Or should one use the emulation.

And is it suggested not to use NUMA only on amd cpu's, or is it just the 
same on core2duo processors?

BTW, what is NUMA exactly? And when should one use it?

cu
Dieter


-- 
3rd Law of Computing:
         Anything that can go wr
fortune: Segmentation violation -- Core dumped
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Daniel Gryniewicz]

On Thu, 2007-06-28 at 18:45 +0200, Dieter Ries wrote:
> Hi,
> 
> 
> Hemmann, Volker Armin schrieb:
> > On Donnerstag, 28. Juni 2007, Simon Cooper wrote:
> >> whats with the NUMA stuff?
> > 
> > nothing. If you only have once cpu with two cores, don't use it!
> 
> What exactly does that mean? None of those :
> 
>      [*] Non Uniform Memory Access (NUMA) Support
>      [*]   Old style AMD Opteron NUMA detection
>      [*] ACPI NUMA detection
>      [*] NUMA emulation
> 
> Or should one use the emulation.
> 
> And is it suggested not to use NUMA only on amd cpu's, or is it just the 
> same on core2duo processors?
> 
> BTW, what is NUMA exactly? And when should one use it?
> 

Intel CPUs don't do NUMA, only AMD cpus do.  However, you need at least
2 physical CPUs to actually do NUMA, and you also need memory attached
to each one.  Most cheap dual CPU AMD boards don't actually have memory
hooked to both CPUs, only to one of them.  If your board has banks of
ram (generally on either side of the CPUs) it probably does NUMA.
Otherwise, it doesn't.

That said, enabling NUMA on a non-NUMA box doesn't really hurt, same as
enabling SMP on a non-SMP box.

Daniel

-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Beso]

[-- Attachment #1: Type: text/plain, Size: 2075 bytes --]

2007/6/28, Daniel Gryniewicz <dang@gentoo.org>:
>
> On Thu, 2007-06-28 at 18:45 +0200, Dieter Ries wrote:
> > Hi,
> >
> >
> > Hemmann, Volker Armin schrieb:
> > > On Donnerstag, 28. Juni 2007, Simon Cooper wrote:
> > >> whats with the NUMA stuff?
> > >
> > > nothing. If you only have once cpu with two cores, don't use it!
> >
> > What exactly does that mean? None of those :
> >
> >      [*] Non Uniform Memory Access (NUMA) Support
> >      [*]   Old style AMD Opteron NUMA detection
> >      [*] ACPI NUMA detection
> >      [*] NUMA emulation
> >
> > Or should one use the emulation.
> >
> > And is it suggested not to use NUMA only on amd cpu's, or is it just the
> > same on core2duo processors?
> >


numa is a way of using the integrated dma ram controllers of amd64 cpus....
on intel systems they are useless if you don't have a lot of cpus....

> BTW, what is NUMA exactly? And when should one use it?
> >
>
> Intel CPUs don't do NUMA, only AMD cpus do.  However, you need at least
> 2 physical CPUs to actually do NUMA, and you also need memory attached
> to each one.  Most cheap dual CPU AMD boards don't actually have memory
> hooked to both CPUs, only to one of them.  If your board has banks of
> ram (generally on either side of the CPUs) it probably does NUMA.
> Otherwise, it doesn't.
>
> That said, enabling NUMA on a non-NUMA box doesn't really hurt, same as
> enabling SMP on a non-SMP box.


of what i have experienced on my amd64x2 4800 i had little performance boost
when i used numa compiled.... for what i know on the amd x2 cpus  every core
has its own l2 chache, its own 64bit ddr mem controller and a unique l3
cache.... i have learned that the numa helps the overall system performance
by using in a better way the integrated mem controller and not only by
working as you said, but i may be wrong.... the only fact that i've
experienced on my desktop is that i have a little perfomance boost.... it
may be due to the mb, an Asus M2N-SLI Deluxe AiLifestyle Series....

Daniel
>
> --
> gentoo-amd64@gentoo.org mailing list
>
>


-- 
beso

d-_-b

[-- Attachment #2: Type: text/html, Size: 3062 bytes --]

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Donnerstag, 28. Juni 2007, Beso wrote:

>
> of what i have experienced on my amd64x2 4800 i had little performance
> boost when i used numa compiled.... for what i know on the amd x2 cpus 
> every core has its own l2 chache, its own 64bit ddr mem controller and a
> unique l3 cache....

AMD64 cpus don't have a level3 cache.
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Christoph Mende]

[-- Attachment #1: Type: text/plain, Size: 193 bytes --]

On Thu, 28 Jun 2007 23:31:21 +0200
"Hemmann, Volker Armin" <volker.armin.hemmann@tu-clausthal.de> wrote:
> AMD64 cpus don't have a level3 cache.

... yet. The Phenom will have L3 cache :>

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Beso]

[-- Attachment #1: Type: text/plain, Size: 410 bytes --]

2007/6/28, Christoph Mende <angelos@gentoo.org>:
>
> On Thu, 28 Jun 2007 23:31:21 +0200
> "Hemmann, Volker Armin" <volker.armin.hemmann@tu-clausthal.de> wrote:
> > AMD64 cpus don't have a level3 cache.
>
> ... yet. The Phenom will have L3 cache :>
>
>
yep, i've last read about phenom chips and forgot that the  x2 chips have
only l2 cache, and not also a l3 one.... sorry for the mislead....

-- 
beso

d-_-b

[-- Attachment #2: Type: text/html, Size: 781 bytes --]

Re: [gentoo-amd64] 3200+ -> X2 4000+

[Richard Freeman]

[-- Attachment #1: Type: text/plain, Size: 1707 bytes --]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Dieter Ries wrote:
> 
> BTW, what is NUMA exactly? And when should one use it?
> 

I don't know the details for linux support for NUMA, but it means
non-uniform memory architecture, as opposed to symmetric
multi-processing (SMP).

In an SMP box all CPUs have equal access to all memory.

In NUMA a particular CPU has faster access to some memory than other memory.

For few CPUs SMP is feasible and simpler to handle - each CPU sits on a
bus with all the RAM, and each CPU accesses any piece of RAM in the same
way.  The problem is that this doesn't scale well - you get bus
contention and maybe even distance/timing issues when you have lots of
CPUs.

The solution to the problem is to give each CPU (or group of CPUs) a
dedicated chunk of RAM, and the ability to access the rest of the RAM by
asking other CPUs/busses/etc for help.  Access to the local memory is
faster than access to non-local memory.  The lowered contention for
local memory makes it possible to scale up the local speed, but you pay
a penalty for accessing non-local memory.

So, an OS needs to know about NUMA to get optimum performance.  For
starters, you'd want to place a processes code and data in the local
memory for the CPU it runs on most of the time.

And I'm sure somebody who knows a lot more about NUMA would be able to
add a lot more to this discussion, and I'm sure I have at least a few
errors in my description....  :)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGg/nZG4/rWKZmVWkRAjIRAJ4ymOD7u/QIBNz2hCtfDQQoZfyPtACgo8OU
hLGhZqMmpK8RxvhhRo8+z98=
=ijjk
-----END PGP SIGNATURE-----

[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/x-pkcs7-signature, Size: 4101 bytes --]

[gentoo-amd64] Re: 3200+ -> X2 4000+

[Regis Decamps]

Fernando Boaglio a écrit :
> Hi guys,
> 
> I bought a new X2 4000 processor and I have 2 questions:

Seems we are having the same choices: I had an AMD 3200 (box version) 
and I bought a AMX x2 4000 (box version too)

The AMD cooler is th same from both CPUs.
-- 
Régis

-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Dieter Ries]

[-- Attachment #1: Type: text/plain, Size: 739 bytes --]

Hi,

Regis Decamps schrieb:
> Fernando Boaglio a écrit :
> The AMD cooler is th same from both CPUs.

I have upgraded from a 3000+ single core 1,8GHz to a X2 4800+ dual core
2*2,4GHz some time ago. I am not using the stock cooler, I have an
Thermaltake cooler. But what I experienced was, that the 4800+ was
running at exactly the same temperature as the 3000+ did, so a dualcore
processor doesn't necessarily have to be warmer than a single core CPU.

I guess, if the airflow your case is OK, you don't _have to_ change the
cooler with the processor. But maybe you are on the safe side if you do.

cu
Dieter

-- 
3rd Law of Computing:
        Anything that can go wr
fortune: Segmentation violation -- Core dumped


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 252 bytes --]

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Samstag, 30. Juni 2007, Regis Decamps wrote:
> Fernando Boaglio a écrit :
> > Hi guys,
> >
> > I bought a new X2 4000 processor and I have 2 questions:
>
> Seems we are having the same choices: I had an AMD 3200 (box version)
> and I bought a AMX x2 4000 (box version too)
>

I bought an 4600+ boxed. And the cooler is a monster compared to the cooler of 
my 3200+ (from artic cooling). With builtin heatpipes and everything. At 
least it regulates himself nicely - as long as my room is cool, the fan does 
not spin very fast (but when it does - and goes beyond 6300rpm, my case 
starts to swing, because it becomes resonant. ugh).
--
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[James Ausmus]

On 6/30/07, Hemmann, Volker Armin <volker.armin.hemmann@tu-clausthal.de> wrote:

<snip>

> I bought an 4600+ boxed. And the cooler is a monster compared to the cooler of
> my 3200+ (from artic cooling). With builtin heatpipes and everything. At
> least it regulates himself nicely - as long as my room is cool, the fan does
> not spin very fast (but when it does - and goes beyond 6300rpm, my case
> starts to swing, because it becomes resonant. ugh).

<snip>

I just put together a new AMD62X2 AM2 system, and have been *quite*
impressed with the cooler that I purchased:

http://www.newegg.com/product/product.asp?item=N82E16835128009

There was a bit of an installation problem on the motherboard that I
have (if you look at the 3rd and 4th pictures on Newegg, you'll see
there are two little metal tabs sticking out on each side of the
bracket at the top of the picture - I had to actually grind those off
to get the cooler to install into the MSI K9N6PGM-F motherboard), but
have been very happy - the fan's maximum speed is 2000 rpm, it is very
quiet (very bright LEDs, but if you can live with that), and does an
excellent job cooling the CPU - I've not seen the CPU rise more than 1
degree above room temperature yet, even under heavy load....

HTH-

James
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Montag, 2. Juli 2007, James Ausmus wrote:

> to get the cooler to install into the MSI K9N6PGM-F motherboard), but
> have been very happy - the fan's maximum speed is 2000 rpm, it is very
> quiet (very bright LEDs, but if you can live with that), and does an
> excellent job cooling the CPU - I've not seen the CPU rise more than 1
> degree above room temperature yet, even under heavy load....

and you checked that with....? k8temp?
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Harry Holt]

[-- Attachment #1: Type: text/plain, Size: 359 bytes --]

On 7/2/07, James Ausmus <james.ausmus@gmail.com> wrote:
>
>
> <snip>
>
> <snip>
>
> quiet (very bright LEDs, but if you can live with that), and does an
> excellent job cooling the CPU - I've not seen the CPU rise more than 1
> degree above room temperature yet, even under heavy load....
>
>
Dude - those CPU temps are in *Celsius* ...


-- 
Harry Holt, PMP

[-- Attachment #2: Type: text/html, Size: 689 bytes --]

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Fernando Boaglio]

Thanks a lot guys, I love this place, I always learn something new =)

Anyway, switching only the processor : 3200+ to X2 4000 and keeping
the same machine (Asus A8V Deluxe + HD IDE 2x80Gb + 2Gb RAM + GeForce
7600GS) how much improvement should I expect ?

I'm a developer, I keep running databases, IDEs like Eclipse, P2P
tools (Azureus, amule) , aMSN and watching my favorite series with
kplayer, like Lost, Supernatural,The 4400 :P

[]'s
Fernando Boaglio

On 7/2/07, Harry Holt <harryholt@gmail.com> wrote:
>
>
>
> On 7/2/07, James Ausmus <james.ausmus@gmail.com> wrote:
> >
> > <snip>
> >
> > <snip>
> >
> > quiet (very bright LEDs, but if you can live with that), and does an
> > excellent job cooling the CPU - I've not seen the CPU rise more than 1
> > degree above room temperature yet, even under heavy load....
> >
> >
>
> Dude - those CPU temps are in *Celsius* ...
>
>
> --
> Harry Holt, PMP
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Beso]

[-- Attachment #1: Type: text/plain, Size: 1525 bytes --]

2007/7/5, Fernando Boaglio <boaglio@gmail.com>:
>
> Thanks a lot guys, I love this place, I always learn something new =)
>
> Anyway, switching only the processor : 3200+ to X2 4000 and keeping
> the same machine (Asus A8V Deluxe + HD IDE 2x80Gb + 2Gb RAM + GeForce
> 7600GS) how much improvement should I expect ?
>
> I'm a developer, I keep running databases, IDEs like Eclipse, P2P
> tools (Azureus, amule) , aMSN and watching my favorite series with
> kplayer, like Lost, Supernatural,The 4400 :P


for what i can tell of these thigs,  you should have about 5-10 watts of
power consumption save, and about 20% of prestation gain.... the great
prestation gain you'll have when you compile, as that is the most killing
process.... th amsn, p2p, kplayer and working eclipse (not compiling one)
are not processor killers.... the db is not a great processor killer, but
having a better processor should surely help.... i think that you should
have thinked of upgrading the disks also and to put a sata-II/raid or a
scsi/raid disk for dbs and the system and to have the other 2 ata disks on
lvm/raid for the other thigs you're doing.... for someone like you the raid
is a must, or you'll risk losing important stuff.... also i'd advice to have
an external disk to backup the very important material just in case.... the
external disk now is very cheap, and i see that your mb has the
sata-i/raid-1 support so you'll only need to buy one sata disk, that is
hasn't a great cost....

[]'s
> Fernando Boaglio
>
>
>
-- 
beso

d-_-b

[-- Attachment #2: Type: text/html, Size: 1997 bytes --]

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Donnerstag, 5. Juli 2007, Fernando Boaglio wrote:
> Thanks a lot guys, I love this place, I always learn something new =)
>
> Anyway, switching only the processor : 3200+ to X2 4000 and keeping
> the same machine (Asus A8V Deluxe + HD IDE 2x80Gb + 2Gb RAM + GeForce
> 7600GS) how much improvement should I expect ?

the biggest improvement will be 'snappiness' of the desktop. If you ever had a 
multicore/multicpu desktop you never want to go back. It is just amazing.

Everything multi-threaded will be faster too - but it completly depends on the 
app. Some almost double their speed, while others show as much as no 
improvement.

>
> I'm a developer, I keep running databases, IDEs like Eclipse, P2P
> tools (Azureus, amule) , aMSN and watching my favorite series with
> kplayer, like Lost, Supernatural,The 4400 :P

watching a video while compiling something will be much more pleasant ;)
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Fernando Boaglio]

Ok folks, I'm running my new linux box right now: =D

cascao ~ # uname -a
Linux cascao 2.6.21-ck2-r1 #2 SMP PREEMPT Tue Jul 10 21:03:33 BRT 2007
x86_64 AMD Athlon(tm) 64 X2 Dual Core Processor 4200+ AuthenticAMD
GNU/Linux

I've changed this option in my [make.conf]:

MAKEOPTS="-j4"

Should I change anything else?

TIA
[]'s
Fernando Boaglio

On 7/5/07, Hemmann, Volker Armin <volker.armin.hemmann@tu-clausthal.de> wrote:
> On Donnerstag, 5. Juli 2007, Fernando Boaglio wrote:
> > Thanks a lot guys, I love this place, I always learn something new =)
> >
> > Anyway, switching only the processor : 3200+ to X2 4000 and keeping
> > the same machine (Asus A8V Deluxe + HD IDE 2x80Gb + 2Gb RAM + GeForce
> > 7600GS) how much improvement should I expect ?
>
> the biggest improvement will be 'snappiness' of the desktop. If you ever had a
> multicore/multicpu desktop you never want to go back. It is just amazing.
>
> Everything multi-threaded will be faster too - but it completly depends on the
> app. Some almost double their speed, while others show as much as no
> improvement.
>
> >
> > I'm a developer, I keep running databases, IDEs like Eclipse, P2P
> > tools (Azureus, amule) , aMSN and watching my favorite series with
> > kplayer, like Lost, Supernatural,The 4400 :P
>
> watching a video while compiling something will be much more pleasant ;)
> --
> gentoo-amd64@gentoo.org mailing list
>
>
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Florian Philipp]

[-- Attachment #1: Type: text/plain, Size: 1867 bytes --]

Am Mittwoch 11 Juli 2007 14:15 schrieb Fernando Boaglio:

>
> On 7/5/07, Hemmann, Volker Armin <volker.armin.hemmann@tu-clausthal.de> 
wrote:
> > On Donnerstag, 5. Juli 2007, Fernando Boaglio wrote:
> > > Thanks a lot guys, I love this place, I always learn something new =)
> > >
> > > Anyway, switching only the processor : 3200+ to X2 4000 and keeping
> > > the same machine (Asus A8V Deluxe + HD IDE 2x80Gb + 2Gb RAM + GeForce
> > > 7600GS) how much improvement should I expect ?
> >
> > the biggest improvement will be 'snappiness' of the desktop. If you ever
> > had a multicore/multicpu desktop you never want to go back. It is just
> > amazing.
> >
> > Everything multi-threaded will be faster too - but it completly depends
> > on the app. Some almost double their speed, while others show as much as
> > no improvement.
> >
> > > I'm a developer, I keep running databases, IDEs like Eclipse, P2P
> > > tools (Azureus, amule) , aMSN and watching my favorite series with
> > > kplayer, like Lost, Supernatural,The 4400 :P
> >
> > watching a video while compiling something will be much more pleasant ;)
> > --
> > gentoo-amd64@gentoo.org mailing list

> Ok folks, I'm running my new linux box right now: =D
>
> cascao ~ # uname -a
> Linux cascao 2.6.21-ck2-r1 #2 SMP PREEMPT Tue Jul 10 21:03:33 BRT 2007
> x86_64 AMD Athlon(tm) 64 X2 Dual Core Processor 4200+ AuthenticAMD
> GNU/Linux
>
> I've changed this option in my [make.conf]:
>
> MAKEOPTS="-j4"
>
> Should I change anything else?
>
> TIA
> []'s
> Fernando Boaglio

CFLAGS="-march=amd64 -msse3"

Does your kernel have "Symmetric multi-processing support" and "Multi-core 
scheduler support"? Do not activate "SMT (Hyperthreading) scheduler support" 
and "Non Uniform Memory Access (NUMA) Support".
 
You might want to play with the "Timer frequency".

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Mittwoch, 11. Juli 2007, Florian Philipp wrote:

>
> CFLAGS="-march=amd64 -msse3"

and O2 ;)
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Thanasis]

on 07/11/2007 05:16 PM Hemmann, Volker Armin wrote the following:
> On Mittwoch, 11. Juli 2007, Florian Philipp wrote:
> 
>> CFLAGS="-march=amd64 -msse3"
> 
> and O2 ;)

I don't see _any_such_flags_ in /etc/make.conf.example.
I use:
CFLAGS="-march=k8 -fomit-frame-pointer -pipe -O2"
CXXFLAGS="${CFLAGS}"

System: notebook AMILO Xa 1526 with AMD Turion(tm) 64 X2 Mobile
Technology TL-56

Am I missing something?


-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Mittwoch, 11. Juli 2007, Thanasis wrote:
> on 07/11/2007 05:16 PM Hemmann, Volker Armin wrote the following:
> > On Mittwoch, 11. Juli 2007, Florian Philipp wrote:
> >> CFLAGS="-march=amd64 -msse3"
> >
> > and O2 ;)
>
> I don't see _any_such_flags_ in /etc/make.conf.example.
> I use:
> CFLAGS="-march=k8 -fomit-frame-pointer -pipe -O2"
> CXXFLAGS="${CFLAGS}"
>
> System: notebook AMILO Xa 1526 with AMD Turion(tm) 64 X2 Mobile
> Technology TL-56
>
> Am I missing something?

if you have 'pni' in /proc/cpuinfo you should add -msse3.
-fomit-frame-pointer is useless/not needed on amd64-.
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Thanasis]

on 07/11/2007 08:20 PM Hemmann, Volker Armin wrote the following:
>> I don't see _any_such_flags_ in /etc/make.conf.example.
>> I use:
>> CFLAGS="-march=k8 -fomit-frame-pointer -pipe -O2"
>> CXXFLAGS="${CFLAGS}"
>>
>> System: notebook AMILO Xa 1526 with AMD Turion(tm) 64 X2 Mobile
>> Technology TL-56
>>
>> Am I missing something?
>
> if you have 'pni' in /proc/cpuinfo you should add -msse3.
> -fomit-frame-pointer is useless/not needed on amd64-.
I do.

# cat /proc/cpuinfo
processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 72
model name      : AMD Turion(tm) 64 X2 Mobile Technology TL-56
stepping        : 2
cpu MHz         : 800.000
cache size      : 512 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 2
fpu             : yes
fpu_exception   : yes
cpuid level     : 1
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext
fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm cr8_legacy
bogomips        : 3611.22
TLB size        : 1024 4K pages
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp tm stc

processor       : 1
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 72
model name      : AMD Turion(tm) 64 X2 Mobile Technology TL-56
stepping        : 2
cpu MHz         : 800.000
cache size      : 512 KB
physical id     : 0
siblings        : 2
core id         : 1
cpu cores       : 2
fpu             : yes
fpu_exception   : yes
cpuid level     : 1
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext
fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm cr8_legacy
bogomips        : 1603.39
TLB size        : 1024 4K pages
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp tm stc

*PS:
1) Should I add any other flags also, besides adding -msse3
(CFLAGS="-march=k8 -msse3 -pipe -O2")
2) Just curious: Why is there such a big difference between the 2 CPUs'
bogomips values reported?


-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Mittwoch, 11. Juli 2007, Thanasis wrote:
> on 07/11/2007 08:20 PM Hemmann, Volker Armin wrote the following:
> >> I don't see _any_such_flags_ in /etc/make.conf.example.
> >> I use:
> >> CFLAGS="-march=k8 -fomit-frame-pointer -pipe -O2"
> >> CXXFLAGS="${CFLAGS}"
> >>
> >> System: notebook AMILO Xa 1526 with AMD Turion(tm) 64 X2 Mobile
> >> Technology TL-56
> >>
> >> Am I missing something?
> >
> > if you have 'pni' in /proc/cpuinfo you should add -msse3.
> > -fomit-frame-pointer is useless/not needed on amd64-.
>
> I do.
>
> # cat /proc/cpuinfo
> processor       : 0
> vendor_id       : AuthenticAMD
> cpu family      : 15
> model           : 72
> model name      : AMD Turion(tm) 64 X2 Mobile Technology TL-56
> stepping        : 2
> cpu MHz         : 800.000
> cache size      : 512 KB
> physical id     : 0
> siblings        : 2
> core id         : 0
> cpu cores       : 2
> fpu             : yes
> fpu_exception   : yes
> cpuid level     : 1
> wp              : yes
> flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
> mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext
> fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm
> cr8_legacy bogomips        : 3611.22
> TLB size        : 1024 4K pages
> clflush size    : 64
> cache_alignment : 64
> address sizes   : 40 bits physical, 48 bits virtual
> power management: ts fid vid ttp tm stc
>
> processor       : 1
> vendor_id       : AuthenticAMD
> cpu family      : 15
> model           : 72
> model name      : AMD Turion(tm) 64 X2 Mobile Technology TL-56
> stepping        : 2
> cpu MHz         : 800.000
> cache size      : 512 KB
> physical id     : 0
> siblings        : 2
> core id         : 1
> cpu cores       : 2
> fpu             : yes
> fpu_exception   : yes
> cpuid level     : 1
> wp              : yes
> flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
> mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext
> fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm
> cr8_legacy bogomips        : 1603.39
> TLB size        : 1024 4K pages
> clflush size    : 64
> cache_alignment : 64
> address sizes   : 40 bits physical, 48 bits virtual
> power management: ts fid vid ttp tm stc
>
> *PS:
> 1) Should I add any other flags also, besides adding -msse3
> (CFLAGS="-march=k8 -msse3 -pipe -O2")

no.  I experimented a lot with different flags in the past - and if there are 
differences they are so small that I missed them. When chasing bugs I had to 
recompile everything with safe flags, and nothing changed. Maybe the system 
is a little bit more stable now...


> 2) Just curious: Why is there such a big difference between the 2 CPUs'
> bogomips values reported?

I really don't know ;)
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Fernando Boaglio]

I found this helpful information:

If you have a newer Athlon64("Venice" or "San
Diego")/Athlon64-X2("Manchester" or "Toledo") (check for pni in cat
/proc/cpuinfo) you can also add -msse3 to your CFLAGS to enable SSE3
support. Any chip using the 90nm process except for the "Winchester"
class Athlon64 supports SSE3.

Also bear in mind that all Athlon64 X2 and Opteron 165, 170, 175, 180,
and 185 processors are dual-core CPUs so make sure that you use
MAKEOPTS="-j3" in your make.conf. This doesn't affect the code but
might reduce compile times.

All socket AM2 processors are SSE3 capable as well.

http://gentoo-wiki.com/Safe_Cflags#Athlon_64_X2_.28AMD.29

On 7/11/07, Thanasis <thanasis@asyr.hopto.org> wrote:
> on 07/11/2007 08:20 PM Hemmann, Volker Armin wrote the following:
> >> I don't see _any_such_flags_ in /etc/make.conf.example.
> >> I use:
> >> CFLAGS="-march=k8 -fomit-frame-pointer -pipe -O2"
> >> CXXFLAGS="${CFLAGS}"
> >>
> >> System: notebook AMILO Xa 1526 with AMD Turion(tm) 64 X2 Mobile
> >> Technology TL-56
> >>
> >> Am I missing something?
> >
> > if you have 'pni' in /proc/cpuinfo you should add -msse3.
> > -fomit-frame-pointer is useless/not needed on amd64-.
> I do.
>
> # cat /proc/cpuinfo
> processor       : 0
> vendor_id       : AuthenticAMD
> cpu family      : 15
> model           : 72
> model name      : AMD Turion(tm) 64 X2 Mobile Technology TL-56
> stepping        : 2
> cpu MHz         : 800.000
> cache size      : 512 KB
> physical id     : 0
> siblings        : 2
> core id         : 0
> cpu cores       : 2
> fpu             : yes
> fpu_exception   : yes
> cpuid level     : 1
> wp              : yes
> flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
> mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext
> fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm cr8_legacy
> bogomips        : 3611.22
> TLB size        : 1024 4K pages
> clflush size    : 64
> cache_alignment : 64
> address sizes   : 40 bits physical, 48 bits virtual
> power management: ts fid vid ttp tm stc
>
> processor       : 1
> vendor_id       : AuthenticAMD
> cpu family      : 15
> model           : 72
> model name      : AMD Turion(tm) 64 X2 Mobile Technology TL-56
> stepping        : 2
> cpu MHz         : 800.000
> cache size      : 512 KB
> physical id     : 0
> siblings        : 2
> core id         : 1
> cpu cores       : 2
> fpu             : yes
> fpu_exception   : yes
> cpuid level     : 1
> wp              : yes
> flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
> mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext
> fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm cr8_legacy
> bogomips        : 1603.39
> TLB size        : 1024 4K pages
> clflush size    : 64
> cache_alignment : 64
> address sizes   : 40 bits physical, 48 bits virtual
> power management: ts fid vid ttp tm stc
>
> *PS:
> 1) Should I add any other flags also, besides adding -msse3
> (CFLAGS="-march=k8 -msse3 -pipe -O2")
> 2) Just curious: Why is there such a big difference between the 2 CPUs'
> bogomips values reported?
>
>
> --
> gentoo-amd64@gentoo.org mailing list
>
>



-- 
[]'s
Fernando Boaglio
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Piotr Pruszczak]

My Opteron 165 (dual-core) shows difference of about 100 bogomips 
between processors

So, I think this is GOOD question - why the difference is so high (over 
50%) ?

>> bogomips        : 3611.22
>> TLB size        : 1024 4K pages

>>

>> bogomips        : 1603.39
>> TLB size        : 1024 4K pages


>> 2) Just curious: Why is there such a big difference between the 2 CPUs'
>> bogomips values reported?
>>
>>
>> -- 
>> gentoo-amd64@gentoo.org mailing list
>>
>>
> 
> 
> 


-- 
Piotr
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Bernhard Auzinger]

> >> bogomips        : 3611.22
> >> TLB size        : 1024 4K pages
> >>
> >>
> >>
> >> bogomips        : 1603.39
> >> TLB size        : 1024 4K pages
> >>
> >>
> >> 2) Just curious: Why is there such a big difference between the 2 CPUs'
> >> bogomips values reported?
> >>
> >>
> >> --
> >> gentoo-amd64@gentoo.org mailing list
>
> --
> Piotr

There has to be no difference between the two cores on dual-core processors. 
That's because the bogomips-benchmark is a "calculated" benchmark that 
depends mostly on the clock frequency and since the two cores on dual-core 
athlons/opterons use the same clock, it's not possible to have different 
values on that type of dual-core processor. I've absolutely identical values 
(4825.77 bogomips) for both cores of my Athlon64 X2 4600+. 

If you have two processors (f.ex. two single-cores) on your mainboard, than 
you will have two slightly different values because of the slightly different 
clock of the two processors. In your case it seems that the second processor 
has a decreased clock. I'm sure you have enabled cool'n'quiet.

Rgds
Bernhard
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Thanasis]

on 07/12/2007 01:17 PM Bernhard Auzinger wrote the following:
>>>> bogomips        : 3611.22
>>>> TLB size        : 1024 4K pages
>>>>
>>>> bogomips        : 1603.39
>>>> TLB size        : 1024 4K pages
>>>>
>>>> 2) Just curious: Why is there such a big difference between the 2 CPUs'
>>>> bogomips values reported?
>>>> --
>>>> gentoo-amd64@gentoo.org mailing list
>> --
>> Piotr
>
> There has to be no difference between the two cores on dual-core processors. 
> That's because the bogomips-benchmark is a "calculated" benchmark that 
> depends mostly on the clock frequency and since the two cores on dual-core 
> athlons/opterons use the same clock, it's not possible to have different 
> values on that type of dual-core processor. I've absolutely identical values 
> (4825.77 bogomips) for both cores of my Athlon64 X2 4600+. 
>
> If you have two processors (f.ex. two single-cores) on your mainboard, than 
> you will have two slightly different values because of the slightly different 
> clock of the two processors. In your case it seems that the second processor 
> has a decreased clock. I'm sure you have enabled cool'n'quiet.
>
> Rgds

Actually I had never changed the CPUs freq/governor states, and they
were set to 800MHz/userspace.
After I changed their state using gnome applet, and from then on,  the
two CPUs' bogomips values  where equal.
For governor set to performance here they are:
#cat /proc/cpuinfo |grep -E 'proc|MHz|bogomips'
processor       : 0
cpu MHz         : 1800.000
bogomips        : 8125.24
processor       : 1
cpu MHz         : 1800.000
bogomips        : 8125.24

However, if I boot into M$ windblows (dual boot system), then boot into
linux without touching the CPUs' freq/gov state the MHz/bogomips values
are changed back to:
processor       : 0
cpu MHz         : 800.000
bogomips        : 3611.94
processor       : 1
cpu MHz         : 800.000
bogomips        : 1603.39

The 800 MHz is because I have compiled with
CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE=y
CONFIG_CPU_FREQ_GOV_USERSPACE=y


-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Beso]

[-- Attachment #1: Type: text/plain, Size: 2522 bytes --]

try using powerthend as a daemon with userspace governor under linux, and in
winzoz download the cpu program for dual core daemon from amd site....
powethend gives you the power you need when you need it, and for i have
learned it is able to handle both cores....

2007/7/12, Thanasis <thanasis@asyr.hopto.org>:
>
> on 07/12/2007 01:17 PM Bernhard Auzinger wrote the following:
> >>>> bogomips        : 3611.22
> >>>> TLB size        : 1024 4K pages
> >>>>
> >>>> bogomips        : 1603.39
> >>>> TLB size        : 1024 4K pages
> >>>>
> >>>> 2) Just curious: Why is there such a big difference between the 2
> CPUs'
> >>>> bogomips values reported?
> >>>> --
> >>>> gentoo-amd64@gentoo.org mailing list
> >> --
> >> Piotr
> >
> > There has to be no difference between the two cores on dual-core
> processors.
> > That's because the bogomips-benchmark is a "calculated" benchmark that
> > depends mostly on the clock frequency and since the two cores on
> dual-core
> > athlons/opterons use the same clock, it's not possible to have different
> > values on that type of dual-core processor. I've absolutely identical
> values
> > (4825.77 bogomips) for both cores of my Athlon64 X2 4600+.
> >
> > If you have two processors (f.ex. two single-cores) on your mainboard,
> than
> > you will have two slightly different values because of the slightly
> different
> > clock of the two processors. In your case it seems that the second
> processor
> > has a decreased clock. I'm sure you have enabled cool'n'quiet.
> >
> > Rgds
>
> Actually I had never changed the CPUs freq/governor states, and they
> were set to 800MHz/userspace.
> After I changed their state using gnome applet, and from then on,  the
> two CPUs' bogomips values  where equal.
> For governor set to performance here they are:
> #cat /proc/cpuinfo |grep -E 'proc|MHz|bogomips'
> processor       : 0
> cpu MHz         : 1800.000
> bogomips        : 8125.24
> processor       : 1
> cpu MHz         : 1800.000
> bogomips        : 8125.24
>
> However, if I boot into M$ windblows (dual boot system), then boot into
> linux without touching the CPUs' freq/gov state the MHz/bogomips values
> are changed back to:
> processor       : 0
> cpu MHz         : 800.000
> bogomips        : 3611.94
> processor       : 1
> cpu MHz         : 800.000
> bogomips        : 1603.39
>
> The 800 MHz is because I have compiled with
> CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE=y
> CONFIG_CPU_FREQ_GOV_USERSPACE=y
>
>
> --
> gentoo-amd64@gentoo.org mailing list
>
>


-- 
beso

d-_-b

[-- Attachment #2: Type: text/html, Size: 3792 bytes --]

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Bo Ørsted Andresen]

[-- Attachment #1: Type: text/plain, Size: 478 bytes --]

On Wednesday 11 July 2007 18:10:55 Thanasis wrote:
> on 07/11/2007 05:16 PM Hemmann, Volker Armin wrote the following:
> > On Mittwoch, 11. Juli 2007, Florian Philipp wrote:
> >> CFLAGS="-march=amd64 -msse3"
> >
> > and O2 ;)
>
> I don't see _any_such_flags_ in /etc/make.conf.example.
> I use:
[SNIP]
> Am I missing something?

You shouldn't be looking in make.conf.example but rather in `man gcc`. But I 
don't see such march there either...

-- 
Bo Andresen

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Florian Philipp]

[-- Attachment #1: Type: text/plain, Size: 660 bytes --]

Am Mittwoch 11 Juli 2007 19:54 schrieb Bo Ørsted Andresen:
> On Wednesday 11 July 2007 18:10:55 Thanasis wrote:
> > on 07/11/2007 05:16 PM Hemmann, Volker Armin wrote the following:
> > > On Mittwoch, 11. Juli 2007, Florian Philipp wrote:
> > >> CFLAGS="-march=amd64 -msse3"
> > >
> > > and O2 ;)
> >
> > I don't see _any_such_flags_ in /etc/make.conf.example.
> > I use:
>
> [SNIP]
>
> > Am I missing something?
>
> You shouldn't be looking in make.conf.example but rather in `man gcc`. But
> I don't see such march there either...

Hmm, you're right. amd64 is wrong, I use athlon64 which is exactly the same as 
k8, athlon-fx or opteron.

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-amd64] Re: 3200+ -> X2 4000+

[Hemmann, Volker Armin]

On Mittwoch, 11. Juli 2007, Florian Philipp wrote:
> Am Mittwoch 11 Juli 2007 19:54 schrieb Bo Ørsted Andresen:
> > On Wednesday 11 July 2007 18:10:55 Thanasis wrote:
> > > on 07/11/2007 05:16 PM Hemmann, Volker Armin wrote the following:
> > > > On Mittwoch, 11. Juli 2007, Florian Philipp wrote:
> > > >> CFLAGS="-march=amd64 -msse3"
> > > >
> > > > and O2 ;)
> > >
> > > I don't see _any_such_flags_ in /etc/make.conf.example.
> > > I use:
> >
> > [SNIP]
> >
> > > Am I missing something?
> >
> > You shouldn't be looking in make.conf.example but rather in `man gcc`.
> > But I don't see such march there either...
>
> Hmm, you're right. amd64 is wrong, I use athlon64 which is exactly the same
> as k8, athlon-fx or opteron.

my bad. I should not answer emails when tired ....

-- 
Conclusions 
 In a straight-up fight, the Empire squashes the Federation like a bug. Even 
with its numerical advantage removed, the Empire would still squash the 
Federation like a bug. Accept it. -Michael Wong 
--
gentoo-amd64@gentoo.org mailing list