Corsair 2xCMS1024-3200C2 or TWINX2048-3200C2?

E

einer2060

Limp Gawd
Joined
Aug 9, 2006
Messages
272
Wich ram should i buy for my sistem:
Athlon64 4000+
Asus A8Nsli dual chanel suport
Raptor150 (in future raid 0, next year?)
7800GTX
twinx1024 i think cas3 whit leds corsair (i would sell that)


Aim serching for a ram not for overcloke (maibe next year)

Corsair
2x CMS1024-3200C2(xtreme speed what ever?)
or
TWINX2048-3200C2(paired for dual chanel)

The price here (300$) at me is the same for both (cant reale get here some cheaper ram:/ )
what should i do, and is it a good ram for non and for overkloking?
 
I would advice you the 2gb kit, they are really good and you will see performance increase in games and aplications.

After trying a 2gb kit you will not go back to a one gb kit.
Even if its overclocking it doesnt make that much differnce that would make me go for a 1 gb kit , 2gb is the way forward.

Also the CL2 latencies is very good.
 
on an amd system, you will not be able to run CAS 2... only on an intel system

for an AMD system you'll need to run CAS 2.5.

I even have to run my DIMMS ( the 2x 512 version) at 2T or else the thing is unstable.
 
You set it to SPD Serial -----Detect , i forgot what the p stands for but basically this puts the rtam as default as what normail C2 , C3 , C4 has the speed latencies too.
 
While I only have 2 x 512, it does run at 1T and CAS2 just fine. I would point fingers at your motherboard quicker than I would the ram or processor if you can't hit it.
 
jcurry, what version of those sticks do you have?

I believe the older (1.1 or 1.2) were better.

Unfortunately, I have 1.5, and i have read CAS 2 for intel, and 2.5 for amd.

I think some kits are better than others, and that is the likely cause...

although i do blame my mobo for not being able to do 1T... sigh.
 
jubrany said:
on an amd system, you will not be able to run CAS 2... only on an intel system

for an AMD system you'll need to run CAS 2.5.

I even have to run my DIMMS ( the 2x 512 version) at 2T or else the thing is unstable.
Click to expand...


wrong , i can 2.0 with no problems :p
 
^yeah, it seems that the 2GB kit is completely different chips than the 1GB kit.

I got like the worst possible version of the TwinX C2s.... I feel jipped.

Wonder if I can bitch at Corsair and maybe they'll send me a better version or even send me some XL sticks?
 
jubrany said:
^yeah, it seems that the 2GB kit is completely different chips than the 1GB kit.

I got like the worst possible version of the TwinX C2s.... I feel jipped.

Wonder if I can bitch at Corsair and maybe they'll send me a better version or even send me some XL sticks?
Click to expand...

Probably not now that I saw your post. :)


If the modules don't do what they're advertised to do call our tech support and we'll gladly RMA them for you.


ALSO: The "AMD is only CAS 2.5" thing is only for AthlonXP processors. Athlon64s should work fine. So if you've tried bumping voltage to 2.8 or so, and it still won't run 2-3-3-6 on your Athlon64 system, call us up and give us a shot to make things right.
 
Redbeard said:
Probably not now that I saw your post. :)


If the modules don't do what they're advertised to do call our tech support and we'll gladly RMA them for you.


ALSO: The "AMD is only CAS 2.5" thing is only for AthlonXP processors. Athlon64s should work fine. So if you've tried bumping voltage to 2.8 or so, and it still won't run 2-3-3-6 on your Athlon64 system, call us up and give us a shot to make things right.
Click to expand...
lol
 
Thanks Redbeard!

It's really great having a Corsair rep. onboard here.

I changed my settings from 2.5-3-3-6 @2T with 2.75v to 2-3-3-6 @1T with 2.8v and am running a prime95 blend stress test right now...

at 2-3-3-6 and 2.75 volts it would pass memtest but fail prime in about 30 minutes, so we'll see what the extra .5v does.
 
I don't know whether to be happy or not just yet, but at 2-3-3-6 @ 1T with 2.8 volts these sticks have been running Prime95 blend stress test for about 2 hours or so...

could it have been something so simple? if this thing keeps prime going for a few days, I'll call it a victory.

BTW, is it bad that I do crap while the stress test is running? I mean, CPU usage is 100%, but while prime was going I've been using my computer to watch videos, play music, surf internet, sisoft sandra, etc.
 
jubrany said:
BTW, is it bad that I do crap while the stress test is running? I mean, CPU usage is 100%, but while prime was going I've been using my computer to watch videos, play music, surf internet, sisoft sandra, etc.
Click to expand...
it's fine. probably actually slightly more stressful, as the memory has to do other things at the same time. you are running the blend test, right? :D
 
(cf)Eclipse said:
it's fine. probably actually slightly more stressful, as the memory has to do other things at the same time. you are running the blend test, right? :D
Click to expand...

Thanks for your input sir. It's the blend stress test I'm running and my available memory hovers somewhere between 40 and 60 MB free.
 
Sir Eclipse saves the day ...........................again , amazing.
 
It failed!

6 Hours and 22 Minutes into the Blend Stress Test, my TwinX1024-3200C2 kit failed with the DIMMS set to 2-3-3-6 1T and 2.8 volts...

[Thu Aug 17 00:20:00 2006]
Self-test 1024K passed!
[Thu Aug 17 00:37:36 2006]
Self-test 8K passed!
[Thu Aug 17 00:55:36 2006]
Self-test 10K passed!
[Thu Aug 17 01:14:38 2006]
Self-test 896K passed!
[Thu Aug 17 01:34:21 2006]
Self-test 768K passed!
[Thu Aug 17 01:51:44 2006]
Self-test 12K passed!
[Thu Aug 17 02:08:43 2006]
Self-test 14K passed!
[Thu Aug 17 02:26:48 2006]
Self-test 640K passed!
[Thu Aug 17 02:42:22 2006]
Self-test 512K passed!
[Thu Aug 17 02:58:54 2006]
Self-test 16K passed!
[Thu Aug 17 03:14:38 2006]
Self-test 20K passed!
[Thu Aug 17 03:30:29 2006]
Self-test 448K passed!
[Thu Aug 17 03:46:06 2006]
Self-test 384K passed!
[Thu Aug 17 04:03:49 2006]
Self-test 24K passed!
[Thu Aug 17 04:22:13 2006]
Self-test 28K passed!
[Thu Aug 17 04:37:37 2006]
Self-test 320K passed!
[Thu Aug 17 04:56:17 2006]
Self-test 256K passed!
[Thu Aug 17 05:14:21 2006]
Self-test 32K passed!
[Thu Aug 17 05:29:23 2006]
Self-test 40K passed!
[Thu Aug 17 05:48:00 2006]
Self-test 224K passed!
[Thu Aug 17 06:06:04 2006]
Self-test 192K passed!
[Thu Aug 17 06:21:31 2006]
Self-test 48K passed!
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.

I guess 2.85volts would not be encouraged right?
 
Is this the memory test or prime , looks like prime , you could put it up , but i try to get something which passes a wee bit lower than that and then work up with voltages and bus increases.

Sometimes your cpu is stable but your memory seems to be the problem, run mem test too.
 
I have now the 2Gb twinx3200C2 2 3 3 6

How must i set the bios?

Cas Latency (TcL) 2 ? (default motherboard seting on auto 3)
Min Ras# aktive time (tras) 6T ? (default... 8T)
Ras# to Cas# delay (trcd) 2T ? (default... 3T)
Row precharge time (Trp) 2T ? (default... 3T)
Row cycle time (Trc) 11T ? (default)
Read to write time (trwt) 5T ? (default)
Write recovery time Twr 3T (default)
1T/2T memory timing 1T ? (default 2T for unbuffered)


What should i set:
Hyper Transport Frequency:
auto, 1x, 2x, 3x, 4x, 5x?

Sistem: Athlon 64 4000+, corsair twinx 2Gb 3200C2, Gforce 7800GTX 256, asus A8N sli, nothing overclocked.



Are this voltages ok?
1.5 Vcore = 1.40V?
3.3V = 3.23/3.24V?
5V = 5.02V?
12V = 11.96V
Enermax 535W coolergiant.
 
For the TWINX2048-3200C2 modules, this is your settings:

Cas Latency (TcL) 2
Min Ras# aktive time (tras) 6T
Ras# to Cas# delay (trcd) 3T
Row precharge time (Trp) 3T
Row cycle time (Trc) 11T
Read to write time (trwt) 5T
Write recovery time Twr 3T
1T/2T memory timing 2T (might work at 1T but try 2T first)


What should i set:
Hyper Transport Frequency:
5x

CPU Vcore: 1.5-1.55V




Just so you know the CMS1024-3200C2 is the same thing as the TWINX2048-3200C2 except that the TWINX is 2 of the CMS modules that have been tested together.
 
Redbeard said:
For the TWINX2048-3200C2 modules, this is your settings:

Cas Latency (TcL) 2
Min Ras# aktive time (tras) 6T
Ras# to Cas# delay (trcd) 3T
Row precharge time (Trp) 3T
Row cycle time (Trc) 11T
Read to write time (trwt) 5T
Write recovery time Twr 3T
1T/2T memory timing 2T (might work at 1T but try 2T first)


What should i set:
Hyper Transport Frequency:
5x

CPU Vcore: 1.5-1.55V




Just so you know the CMS1024-3200C2 is the same thing as the TWINX2048-3200C2 except that the TWINX is 2 of the CMS modules that have been tested together.
Click to expand...

I have some questions for you too.
Runnig a Asus P5P800 SE with Intel 865PE chipset and "hyperpath"(What Asus calls PAT on 865 boards and 3 settings I havn't got a clue to set (running them on AUTO now)
(Using the same Corsair RAM modules 2x1 GB in Dualchannel mode)

DRAM Burst Length (4, 8 clocks)
DRAM Idle Timer (0T, 8T, 16T, 64T)
DRAM Refresh Rate(15.6 uSec, 7.8 uSec, 64 uSec, 64T)

Thanks in advance :)
 
They told me to "Disable USB Legacy Support", then run memtest for a few passes with the ram set to 2-3-3-6 and 2.75 v.
So I did that, and I also disabled "remap for 4GB" option, and I set the CMD Rate to 1T. It made 2 passes so i went to running Prime 95.

now Prime has been running for like 10 hours or so, and hasn't failed yet... I'm confused! everything seems good now, but I don't want to keep USB legacy support disabled, because I won't be able to use my G5 mouse in safemode...

Could legacy support for USB have been causing prime to fail, or is it that I disabled 4GB remapping?

in any case, I'm gonna let this go for a WHILE (i.e. 72 hours) and if it works, I'm gonna turn USB legacy support back on in the BIOS and see if it will go another 72... I'm thinking it could be the 4GB remapping option...
 
well, I failed again after 11 hours.

I switched my write to rewrite time from 4T to 5T, and if it fails again, I'm giving up on 1T CMD.

But yeah, if anything, this exercise in frustration has helped me realize that yes, 2-3-3-6 is stable on my machine, but only at 2T.

is this problem because my board doesn't go below 201 FSB (402 DDR)? LOL, maybe that one megahertz is causing instability.
 
I set it 1T, it is working, didn't done any testes jet only windows...

What should i do about the vcore voltage 1.4V that is to low.
 
einer2060 said:
I set it 1T, it is working, didn't done any testes jet only windows...

What should i do about the vcore voltage 1.4V that is to low.
Click to expand...

I believe you have a S939 San Diego Core.... 1.4v is the correct voltage. 1.5 would be a considerable overvolt.
 
jubrany said:
I believe you have a S939 San Diego Core.... 1.4v is the correct voltage. 1.5 would be a considerable overvolt.
Click to expand...

It would be an overvolt, but should still be perfectly safe on even the stock aircooler. It will shorten the lifespan of your CPU every so slightly, but what's the difference between 10 years and 9 years when you're going to get a new one in 3-5 years anyway, right? Heh. Well, maybe that's not how everyone thinks...
 
Redbeard said:
It would be an overvolt, but should still be perfectly safe on even the stock aircooler. It will shorten the lifespan of your CPU every so slightly, but what's the difference between 10 years and 9 years when you're going to get a new one in 3-5 years anyway, right? Heh. Well, maybe that's not how everyone thinks...
Click to expand...
When i buy mine i only think just a year will do.
 
When I asked Corsair why I could only run at 2T timings, they replied:

Corsair Technical Support Responds: Command Rate (1T/2T) is the delay, in clock cycles, between each instruction is executed by the CPU; therefore whether the system is capable of running 1T command rate is depended on the CPU, power supply, and motherboard. I hope this helps you out!
Click to expand...

So basically 1T/2T has nothing to do with the RAM, and everything to do with everything else?

Is this really true???
 
jubrany said:
When I asked Corsair why I could only run at 2T timings, they replied:



So basically 1T/2T has nothing to do with the RAM, and everything to do with everything else?

Is this really true???
Click to expand...
Yep very true, If your ram is capable but your cpu is too slow then it wont be able to hack it , actually I do think its the chipset that really makes the difference , with it controlling the data. I myself tried this on a system and see if it could hack the timings and I got 30% more bandwidth than i started with.
 
damn, because my bandwidth goes from 4800mb to 5500mb when i go from 2T to 1T...

so if I run a divider and OC my CPU to let's say 2400MHz (400 MHz OC) it might become stable at 1T?
 
You must log in or register to reply here.
Back
Top