Jump to content
Sign in to follow this  
robbay111

Overclockin' General Question?!

Recommended Posts

Also depends on what hes using to read temps?

And yeah it will only drop 2 to 3C at most.

And 37 is grate on a duel core

My BT is running 32c @2800 idle. Its 26c in the house.

Get core temp and see what it says as that would be standard for all.

My core temp show #0-38c#1 40c

Share this post


Link to post
Share on other sites

Ceramic cures much faster if you burn and cool it, in cycles. Run S+M for 15mins and shut down. Power up again when the cpu reaches ambient and start the process over again. A couple of hours of doing that will cure it as good as your going to get.

Share this post


Link to post
Share on other sites
Also depends on what hes using to read temps?

And yeah it will only drop 2 to 3C at most.

And 37 is grate on a duel core

My BT is running 32c @2800 idle. Its 26c in the house.

Get core temp and see what it says as that would be standard for all

 

wil I have my 165 Opteron @2700 and it idles 30-31C diode temp and it is 22C (If that is 72F :)) ambient. Big T. Core Temp is lower.

 

I don't know if you mean to say all dual cores run hotter. Mine does not seem to be much higher than I am used to with single. This is my only dual so I believe you.

Share this post


Link to post
Share on other sites

haha well this is my first also but going though everyone getting duel cores when they came out all seam to say they run hotter. Don't know Mine sure is running hotter then my 3700 at 3 gig

Share this post


Link to post
Share on other sites
Set the Zalman to go full tilt at 30C and see what happens.

 

figured out what you meant..

 

edit:

i set up the cpu fan to go on full blast of the temp was greater then 33. (however i'm not sure it'll do anything since i have my zalman plugged into my 4-fan-controler.)

 

Run S+M for 15mins and shut down.

 

thanks for the tip!

 

edit #1:

alright, i re-did the heatsink, and temps dropped a little bit. not HUGE but better then they were before.

it seemed to stay at 35 when i was idle, and it kind of bumped up to 36 then back down, then down to 34 then back to 35 kinda thing

new idle temps

 

update time :D

 

edit #2:

alright i think those initial new temps were just that, initial... seems like she's running still at 38 idle.

i ran s+m for 15 minutes, as kit suggested and powered down. did that for about an hour and a half then ran the full cpu test at 85% load. It passed.

 

I passed through both the cpu stress and the hard drive stress test w/ s+m at 100% load.....however i got 2 errors with the memory test portion of the stress, so i'm going to re-run memtest and let it FULLY finish this time. ( this memory error, should i be overly worried about it or no? )

 

i'm also reading on why the cores are at different temps.. read 5 or 6 threads so far. still a little confused but i'll keep reading and try to learn more.

i'll obviously run prime (on both processors) and see how that goes after the full memtest is finished.

 

if my temps stay where they are now, i suppose it isn't the best idea to overclock until i can get my idle temp to ~30 degrees or so right? should i consider rigging some more fans in there or what

Share this post


Link to post
Share on other sites

I didn't realize you were using a separate fan controller (not the Mickey Mouse one that comes with Zalman).

 

Not sure about the mem errors in S&M. I've never used that part of S&M. I would stick with memtest98 and Orthos (use S&M if you want to find problem with overall clock quicker maybe).

 

You can try OC with temps where they are. You will drive yourself crazy and temps will not improve much unless you try to change coolers. Can't remember if you checked for true flatness of your CPU/HS and Zalman base. If you ever pull it again check that. If both are not flat you will not get good contact.

 

Just jump in and start Overclocking. During OC testing you will go above "ideal" temps anyway.

 

Your two cores will usually read different temps. If they are extremely different at similar load it may indicate poor contact between the HS on the CPU and the chip. If the difference is not extreme don't worry about it (what is extreme = good question). To be honest these things can run at very high temps for long periods. How long do we really keep these things anyway. (don't tell anyone I said that:)) 3800+ is $65 on Newegg right now, it's not like you have a $400 CPU to worry about. I am too conservative with temps probably.

 

Mostly we want cooler because it performs better, will OC higher. Maybe this one will do just what you want and simply run hotter than you or I think it should.

 

At some point you just have to jump in ... go for it.

Share this post


Link to post
Share on other sites

better trust S & M on memory test over memtest. as if its faling in S&M it will never prime.

I go from S & M to max OccT then prime.

I only use the memory test in S & M

OccT is a quick Hard 30 min test

Share this post


Link to post
Share on other sites
better trust S & M on memory test over memtest.

 

hm thats weird.. i ran memtest last night and let it run for ~16 hours. it completed 19 passes with no errors.

 

and throughout the day today i've ran and completed both the OCCT test+torture and S+M load test at 100%. i just stopped S+M when it started testing my hard drive.

 

here's the cpu graph from OCCT - CPU-Graph

 

tonight i'll leave orthos running while i sleep and see how that goes. if all goes well then i'll debate starting to overclock a bit.

 

one thing i have noticed is that my cores are fairly close in temps under load, about 4/5 degrees off. right now @ idle the 2nd core is 10C hotter then the 1st core. (38C and 28C respectively)

 

after dinner i'll try reading some more threads.

Share this post


Link to post
Share on other sites

only time i use memtest is when changeing timings in bios. i run ONE pass of test 5 and ONE pass of test 8 if they pass its good to get in windows safe then i use S & M for all other memory test.

 

yeah thats normal the secont core is always show higher

Share this post


Link to post
Share on other sites

thanks willy

 

i'll keep that in mind when i start oclockin a bit.

 

i'm still tryin to read more about the core temp variance to understand it better.

 

edit:

 

on the plus side, i ran orthos last night until this morning for ~14 hours and she ran no problem.. temps weren't amazing but they hovered around 46C load which is still not that bad. the temp variance was between 3+6 degrees so i should be fairly stable... i hope! lol

Share this post


Link to post
Share on other sites

alright i've been gradually crankin up my bus frequency and testing via orthos.

 

as of right now i'm at 2.45 stable with ~19 hours of stresssssss.

ortho's + temps

 

however before i ran the full orthos last night, i attempted to clock up to 2.5

i did not receive a boot error, however after about 5 minutes in orthos i recieved an error...

 

now i'm not quite sure what this error message means.

this is what popped up in the log...

1:5 minutes

1:Temperature: CPU Temp: 45.00

1:Temperature: PWMIC: 45.00

1:Temperature: NF4 Chipset: 45.00

1:Voltage: +3.3 Rail: 3.18

1:====================

1:FATAL ERROR: Rounding was 0.5, expected less than 0.4

1:Hardware failure detected, consult stress.txt file.

1:Torture Test ran 5 minutes 24 seconds - 1 errors, 0 warnings.

1:Execution halted.

 

 

i know you'll need to know what my stock settings were and what im running right now so i'll post them underneath.

If they aren't the 'true' defaults they are my previous settings before any update / overclock.

 

Note:

The Column on the LEFT side is more or less the defauls (minus a few) and the RIGHT SIDE are my current settings (minus 250 bus freq.)

 

Genie BIOS Settings:

FSB Bus Frequency............................. - 200					250
LDT/FSB Frequency Ratio....................... - Auto					Auto
CPU/FSB Frequency Ratio....................... - Auto					10
PCI eXpress Frequency......................... - 100Mhz					105Mhz
CPU VID StartUp Value......................... - Startup			  Startup

CPU VID Control............................... - 1.3V					1.3V
CPU VID Special Control....................... - Auto					Auto
LDT Voltage Control........................... - 1.2V					1.20v
Chip Set Voltage Control...................... - 1.5V					1.50v
DRAM Voltage Control.......................... - 2.6V					2.80v 

DRAM Configuration Settings:

DRAM Frequency Set............................ - 166=RAM/FSB:05/06 (Forget Default)	166=RAM/FSB:05/06
Command Per Clock (CPC)....................... - Disabled				Disable
CAS Latency Control (Tcl)..................... - Auto					2.5
RAS# to CAS# delay (Trcd)..................... - Auto					3
Min RAS# active time (Tras)................... - Auto					6
Row precharge time (Trp)...................... - Auto					3

Row Cycle time (Trc).......................... - 7 Bus Clocks				8 Bus Clocks
Row refresh cyc time (Trfc)................... - 12 Bus	Clocks	       12 Bus Clocks
Row to Row delay (Trrd)....................... - 2 Bus Clocks				03 Bus Clocks
Write recovery time (Twr)..................... - 2 Bus Clocks				03 Bus Clocks
Write to Read delay (Twtr).................... - 2 Bus Clocks				02 Bus Clocks
Read to Write delay (Trwt).................... - 3 Bus Clocks				02 Bus Clocks
Refresh Period (Tref)......................... - 3120 Cycles				2560 Cycles
DRAM Bank Interleave.......................... - Emabled				Enabled

DQS Skew Control.............................. - Auto					Auto
DQS Skew Value................................ - 0					0
DRAM Drive Strength........................... - Auto					Auto
DRAM Data Drive Strength...................... - Auto					Auto
Max Async Latency............................. - Auto					9.0
DRAM Response Time............................ - Fast					Fast
Read Preamble Time............................ - Auto					6.0
IdleCycle Limit............................... - 256 Cycles				8 Cycles
Dynamic Counter............................... - Disabled				Disable
R/W Queue Bypass.............................. - 16X					16 x
Bypass Max.................................... - 4X					07 x
32 Byte Granularity........................... - Disabled (4 Bursts)			Disable(4 Bursts)

 

 

I was just wondering if anyone knew what could have caused that error and or what i should try doing to attempt to overcome such error.

 

Note:

I know that i'm attempting to clock with all 4 of my sticks of ram in, if that is contributing to my problem i will remove 2

 

Thanks!

Share this post


Link to post
Share on other sites

I would start adding some vcore 1.325, 1.35, 1.375, 1.4 and test on the way up. You will start to see somewhere it will start to test longer without error.

 

Nice job posting. Do an A64 info and post it. We will see what may be under some of the auto settings.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×