Gigabyte P35-DS3P

On 21 Jan, 06:54, "Fishface" ? wrote:
bornfree wrote:
My RAM says:
DDR2 -533 1GB
pqi 0101-06D1
MEABR422LA

Does that relate to PC2 6400 in any way?

Nope, PC2-4200. PC2-6400 is DDR2-800.


I have a confession to make then. 1GB of memory is in the form of a
DDR2 -533 stick, while the other GB is PC2-6400.

The PC2-6400 I fitted myself, and was £20 inc postage. (equiv to $40
USD). Should I spend that amount again to make them both the same?

On 21 Jan, 08:30, bornfree wrote:
On 21 Jan, 06:54, "Fishface" ? wrote:

bornfree wrote:
My RAM says:
DDR2 -533 1GB
pqi 0101-06D1
MEABR422LA

Does that relate to PC2 6400 in any way?

Nope, PC2-4200. PC2-6400 is DDR2-800.

I have a confession to make then. 1GB of memory is in the form of a
DDR2 -533 stick, while the other GB is PC2-6400.

According to the bios it (my DRAM) was running at 533(something) but
CPU-Z says it is running at 300Mhz.

Also I have a sneaky feeling when I set it higher in the bios it will
stay the same in CPU-Z. I dunno.

Also, when I changed the FSB from 300 to 366, my CPU clock stayed at
2.10Ghz. :confuzzled:

Here's the cheaper ram stick info in cpu z
http://i26.tinypic.com/s16pth.png

Somewhere on teh intarweb "Fishface" typed:
bornfree wrote:

CPUID still reports my clock speed to be 1.86 under "specification"
though .

If SpeedStep is enabled, it will automatically slow down under load.
Some report higher overclocks when this is disabled, as it lowers the
voltage (vCore) also. Check it with a load.

Which CPU do you have? If multi-core, sure it is stressing all cores
by checking the performance tab of Windows Task Manager, assuming you
are running Windows, of course. I just use this:
http://www.techpowerup.com/downloads...rime_2004.html

...which runs two instances of Prime95 torture test. I need to run
two instances of Orthos for the Quad core, though. Shaun posted that
he found a Prime95 version on www.MajorGeeks.com which is ver 25.5.
that would stress all cores. I was a little suspicious that it
wasn't found on the official website, so I didn't download it.
http://www.mersenne.org/freesoft.htm

It's been running fine, far easier than mucking about with two instances of
Orthos and setting affinity etc.. :-)

I see that it has been mentioned as the stress program of choice in a few
mobo reviews I've read recently. (Sorry, I didn't keep URLs)

The page that you linked, http://www.mersenne.org/freesoft.htm hasn't been
updated since July 4, 2007. Odd I know. shrug

Cheers,
--
Shaun.

Somewhere on teh intarweb "RobV" typed:
Fishface wrote:
bornfree wrote:

CPUID still reports my clock speed to be 1.86 under "specification"
though .

If SpeedStep is enabled, it will automatically slow down under load.
Some report higher overclocks when this is disabled, as it lowers the
voltage (vCore) also. Check it with a load.

Which CPU do you have? If multi-core, sure it is stressing all cores
by checking the performance tab of Windows Task Manager, assuming you
are running Windows, of course. I just use this:
http://www.techpowerup.com/downloads...rime_2004.html

Thanks for the link. It does make up for the features Prime95 lacks.

...which runs two instances of Prime95 torture test. I need to run
two instances of Orthos for the Quad core, though. Shaun posted that
he found a Prime95 version on www.MajorGeeks.com which is ver 25.5.
that would stress all cores. I was a little suspicious that it
wasn't found on the official website, so I didn't download it.
http://www.mersenne.org/freesoft.htm

I've been using Prime95 ver 25.5 for a while and it looks as if it's
simply a crude version of what OSP is. It has a main window, two
windows within, one for each core, and it runs the same tests as
regular Prime95 and OSP. I'll certainly use OSP from now on because
it does have all the features that Prime95 lack, all versions
included.

Not true, which you'd know if you had a four-core CPU.

What features does Orthos have that Prime95 ver.25.5 lacks? I'm curious. I
took Orthos off the systems that I'm responsible for in favour of P95/25.5
as it handles as many cores as you have, without you having to start
multiple instances. Also, it gives you the options of test-types, the same
as Orthos does.
--
Shaun.

~misfit~ wrote:
Somewhere on teh intarweb "RobV" typed:
Fishface wrote:
bornfree wrote:

CPUID still reports my clock speed to be 1.86 under "specification"
though .
If SpeedStep is enabled, it will automatically slow down under load.
Some report higher overclocks when this is disabled, as it lowers the
voltage (vCore) also. Check it with a load.

Which CPU do you have? If multi-core, sure it is stressing all cores
by checking the performance tab of Windows Task Manager, assuming you
are running Windows, of course. I just use this:
http://www.techpowerup.com/downloads...rime_2004.html
Thanks for the link. It does make up for the features Prime95 lacks.

...which runs two instances of Prime95 torture test. I need to run
two instances of Orthos for the Quad core, though. Shaun posted that
he found a Prime95 version on www.MajorGeeks.com which is ver 25.5.
that would stress all cores. I was a little suspicious that it
wasn't found on the official website, so I didn't download it.
http://www.mersenne.org/freesoft.htm
I've been using Prime95 ver 25.5 for a while and it looks as if it's
simply a crude version of what OSP is. It has a main window, two
windows within, one for each core, and it runs the same tests as
regular Prime95 and OSP. I'll certainly use OSP from now on because
it does have all the features that Prime95 lack, all versions
included.

Not true, which you'd know if you had a four-core CPU.

What features does Orthos have that Prime95 ver.25.5 lacks? I'm curious. I
took Orthos off the systems that I'm responsible for in favour of P95/25.5
as it handles as many cores as you have, without you having to start
multiple instances. Also, it gives you the options of test-types, the same
as Orthos does.

Orthos "blinks and beeps" more when there is an error :-)

I've tried Orthos, and 255 and 255A and they all seem to be doing
essentially the same kind of testing. I find Orthos seems to take
a bit more time to get started.

I don't understand the source of the 255 and 255A variants. They were
found in the mersenne.org/gimps directory, and were not indexed in the
"freesoft" section. The reason for that could well be, that someone
contributed a multithread version, and since the "official" development
people didn't produce it, it was hosted on the server but not advertised.
I wouldn't have found that one myself, except to see it mentioned on
an overclocking web site. Otherwise I'd still be using the Orthos
version.

Paul

Somewhere on teh intarweb "Paul" typed:
~misfit~ wrote:
Somewhere on teh intarweb "RobV" typed:
Fishface wrote:
bornfree wrote:

CPUID still reports my clock speed to be 1.86 under
"specification" though .
If SpeedStep is enabled, it will automatically slow down under
load. Some report higher overclocks when this is disabled, as it
lowers the voltage (vCore) also. Check it with a load.

Which CPU do you have? If multi-core, sure it is stressing all
cores by checking the performance tab of Windows Task Manager,
assuming you are running Windows, of course. I just use this:
http://www.techpowerup.com/downloads...rime_2004.html
Thanks for the link. It does make up for the features Prime95
lacks.
...which runs two instances of Prime95 torture test. I need to run
two instances of Orthos for the Quad core, though. Shaun posted
that he found a Prime95 version on www.MajorGeeks.com which is ver
25.5. that would stress all cores. I was a little suspicious that
it wasn't found on the official website, so I didn't download it.
http://www.mersenne.org/freesoft.htm
I've been using Prime95 ver 25.5 for a while and it looks as if it's
simply a crude version of what OSP is. It has a main window, two
windows within, one for each core, and it runs the same tests as
regular Prime95 and OSP. I'll certainly use OSP from now on because
it does have all the features that Prime95 lack, all versions
included.

Not true, which you'd know if you had a four-core CPU.

What features does Orthos have that Prime95 ver.25.5 lacks? I'm
curious. I took Orthos off the systems that I'm responsible for in
favour of P95/25.5 as it handles as many cores as you have, without
you having to start multiple instances. Also, it gives you the
options of test-types, the same as Orthos does.

Orthos "blinks and beeps" more when there is an error :-)

LOL, I thought about that afterwards. The green/red "Paging" thing is a
great feature. ;-)

I've tried Orthos, and 255 and 255A and they all seem to be doing
essentially the same kind of testing. I find Orthos seems to take
a bit more time to get started.

Yep. I agree, they all seem to use the same engine. However, P95 25.5 will
automagically start as many threads as there are cores. That's gotta be a
bonus no?

I don't understand the source of the 255 and 255A variants. They were
found in the mersenne.org/gimps directory, and were not indexed in the
"freesoft" section. The reason for that could well be, that someone
contributed a multithread version, and since the "official"
development people didn't produce it, it was hosted on the server but
not advertised. I wouldn't have found that one myself, except to see
it mentioned on an overclocking web site. Otherwise I'd still be using the
Orthos
version.

That's how I found it too, reading web forums on OCing.

Cheers,
--
Shaun.

~misfit wrote:

It's been running fine, far easier than mucking about with two instances
of
Orthos and setting affinity etc.. :-)

It sure would be!

I see that it has been mentioned as the stress program of choice in a few
mobo reviews I've read recently. (Sorry, I didn't keep URLs)

As have I. It is comforting to know that it is available frome mersenne.org
by
FTP . ftp://mersenne.org/gimps/

I was going to check-out which APIs it uses with QuickView first.
http://personal.sirma.bg/Jogy/tricks.html