Dual channel or lower CAS?

Tazz

Distinguished
Jun 3, 2004
14
0
18,510
Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Which RAM setup would yield faster performance; 1 Gig at 2.5 4 4 8 Dual
Channel, or 512 Meg at 2 3 3 6?

I have one 512 MB PC3200 stick of Samsung RAM and one 512 MB PC3200 OCZ
RAM. To get them to run dual channel I have to set the timings at the
max for the Samsung RAM, but the OCZ RAM is rated for 2 3 3 6.

Is going to dual channel supposed to be faster than going from CAS 2.5
to CAS 2?


It seems to me that the OCZ RAM by itself is faster most times; using
SETI as a benchmark. I also would like to know for gaming and video
encoding.

TIA
 
G

Guest

Guest
Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Tazz wrote:
> Which RAM setup would yield faster performance; 1 Gig at 2.5 4 4 8
> Dual Channel, or 512 Meg at 2 3 3 6?

What is the rest of your system, what's your typical memory usage level, and
what programs will you be running? Assuming you don't usually go above
512mb, then for most common apps a resonable-powered S939 system would do
better with dual-channel and a low-powered socket-A system would do better
with the lower latency. But it really depends a lot on what you have in your
system and what you use your system for.

Since you have both sticks already, try each combination with whatever apps
you use that need the most grunt and choose accordingly. And then report
back, because it would be interesting to know exactly how different they
performed :)

[...]

--
Michael Brown
www.emboss.co.nz : OOS/RSI software and more :)
Add michael@ to emboss.co.nz - My inbox is always open


------------ And now a word from our sponsor ------------------
Do your users want the best web-email gateway? Don't let your
customers drift off to free webmail services install your own
web gateway!
-- See http://netwinsite.com/sponsor/sponsor_webmail.htm ----
 

Tazz

Distinguished
Jun 3, 2004
14
0
18,510
Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Michael Brown wrote:
>
> Tazz wrote:
> > Which RAM setup would yield faster performance; 1 Gig at 2.5 4 4 8
> > Dual Channel, or 512 Meg at 2 3 3 6?
>
> What is the rest of your system, what's your typical memory usage level, and
> what programs will you be running? Assuming you don't usually go above
> 512mb, then for most common apps a resonable-powered S939 system would do
> better with dual-channel and a low-powered socket-A system would do better
> with the lower latency. But it really depends a lot on what you have in your
> system and what you use your system for.
>
> Since you have both sticks already, try each combination with whatever apps
> you use that need the most grunt and choose accordingly. And then report
> back, because it would be interesting to know exactly how different they
> performed :)

OK, my CPU is a 2500+@3000+. 10.5x200 seems to be the sweet spot because
a higher multiplier or FSB results in Windows not booting.

I know that going to dual channel AND 2 3 3 6 resulted in a 4.5% speed
increase, but the

system wasn't stable and would reboot itself randomly.

I'll pull out the Samsung RAM and change the timings and see what
happens.


Here's what I found out:
SETI only uses about 16 MB of RAM, but the speed it can move information
to and from the RAM is what counts. I completed six work units (WUs).
Three with dual channel and relaxed timings and three with the faster
timings. Three out of six WUs were completed approx. 2 min. faster with
the dual channel. The other three WUs done with the tighter timings were
approx. 2 min. faster than the dual channel. Go figure.

Encoding a 98 MB MP3 to MP4 took 6:10 (min:sec) with 1 Gig RAM in dual
channel mode and timings at 2.5 4 4 8.
It took 6:11 with 512 MB and timings at 2 3 3 6.

Using DVDShrink I re-encoded The Empire Strikes Back from files on the
hard drive. I set it to compress the main movie as much as it could and
to perform a deep analysis before encoding. It took 58:14 with 512 MB
and fast timings. The dual channel and slower timings took 17 sec.
longer; 58:31

Each test was done after a fresh reboot.

My conclusion; there's not enough of a difference to make me crawl back
under my desk, open my PC case and pull out a stick of RAM. So it'll
stay in the dual channel mode.

This was in no way a scientific test, others will probably have
different results.

</Tazz>