AMDŽ Socket-AM2 Overclocking thread (V2), discuss our setup/overclocking issues
|
|
 Jun 18 2007, 03:08 PM
Return to original view | Post
#1
|
 
Elite
14,576 posts Joined: May 2006 From: Sarawak   
|
QUOTE(8tvt @ Jun 18 2007, 03:00 PM) lol.. u need +0.175.. to get 3ghz.. and set htt to 600 (3x htt mul) Noticed that your X2 3600+ is at 3 Ghz!!! Cool!ram divider 533...  What is your OC vcore? (And what is the stock vcore BTW?) And what are you getting at full load for your core temp? |
|
|
|
|
|
Jun 18 2007, 05:20 PM
Return to original view | Post
#2
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(8tvt @ Jun 18 2007, 03:59 PM) oc = 1.4v Great proc! I guess with such a low temp, going +0.15v on vcore should be no problem.default = 1.25v EE = 1.1v stock clock.. i don't really care about the temp.. normally software reports incorrectly.. but it should around 50C on load.. maybe should try thermal sensor later.. For my toledo core, going +0.05v already cause the core temp to jump up to 55-56'c!! |
|
|
Jun 18 2007, 07:09 PM
Return to original view | Post
#3
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(8tvt @ Jun 18 2007, 06:00 PM) windsor also like that.. i even tried untill 70C... on stock cooler.. 70'c!!! i most prefer core temp.. below 60C shouldn't be a problem...  Ya, I guess below 60'c should be ok. Just that to me, the increase in vcore and temperature for that 50-100 Mhz gain is not justified. Well, at least for my proc anyway, need to pump in a lot of vcore to get a slight increase in OC....  Some forumers told me could be because of my mobo or PSU...... |
|
|
Jun 18 2007, 08:21 PM
Return to original view | Post
#4
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(8tvt @ Jun 18 2007, 08:03 PM) maybe becoz of ram? DDR and DDR2 quite different on voltage supply and heat disipation... case ventilation also another caused.. I see what you mean. Wished I was on the AM2 platform.... Then can get a brisbane.... 2.7 for toledo 4200+ seems ok... my previous windsor 3800+ also like that... but need to supply until 1.5v... brisbane is different from those two... 65W, 65nm, less default voltage.. CODE Stock TDP 89 65 Clockspeed 2700 3011 Stock Clockspeed 2000 1900 vCore 1.40 1.41 Stock vCore 1.33 1.25 Actual TDP 133.13 130.69 as u see.. even brisbane can achieve 3ghz.. still less Thermal Design Power (TDP) draw... thus less heat..  Anyway, aiming for a q6600 in the near future....  Thx for the info!  |
|
|
Jun 18 2007, 09:06 PM
Return to original view | Post
#5
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(afosz @ Jun 18 2007, 08:37 PM) Ermm I have a Q. What if SP2004 crashes after OCed, but still stable in windows and games like FEAR? That probably means that the OC is not stable.SP2004 will fully stress your proc where in games, the proc is usually not stress 100%..... Yours is a single core? |
|
|
Jun 18 2007, 09:19 PM
Return to original view | Post
#6
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(afosz @ Jun 18 2007, 09:09 PM) Nope, 3600+ Brisbane dual core have to use orthos ma.... unless you're running 2 instances of sp2004? This post has been edited by kmarc: Jun 18 2007, 09:19 PM |
|
|
|
|
|
Jun 18 2007, 09:57 PM
Return to original view | Post
#7
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(afosz @ Jun 18 2007, 09:43 PM) Thanks for the info. Still learning, hope you guys can guide me You need to stress your OC at 100% to make sure that your rig is 100% stable. Orthos will 100% stress my cpu, but what if I don't need 100% cpu stress. Just enough for my windows, games and app to load faster, does that means considered ok already? This is because you can't be sure that your rig won't run at 100% at certain times. In addition, let's say you are running at 50% most of the time, then there is no point overclocking as the reason for overclocking is to get more than what your stock CPU can provide. Do you know what I mean? |
|
|
Jun 19 2007, 02:44 PM
Return to original view | Post
#8
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(afosz @ Jun 19 2007, 01:03 PM) What is trfc? TRFC is row refresh cycle time. It is one of the RAMs settings. The most common settings that are tweaked are the: and about loose or tight timing, which timing is loose, which is tight?Currently still trying, at temp at 45~48C at BIOS, is it ok? Dont know which reliable software to monitor temp for Vista 1) Cas latency (CL) 2) Ras to Cas delay (tRCD) 3) Ras precharge (tRP) 4) Active to precharge delay (rRAS) These are memory timings, commonly written as CL-tRCD-tRP-rRAS e.g. 2-2-2-5. The lower the memory timings, the faster the RAMs run. 2-2-2-5 are tight timings which runs faster than loose timings e.g. 3-3-3-8 at the same clock speed. For DDR rams, the best is 2-2-2-5 For DDR2 rams, I think the best is 3-3-3-8..... (sorry, I don't have ddr2...  ) |
|
|
Jun 20 2007, 07:26 AM
Return to original view | Post
#9
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(lohwenli @ Jun 20 2007, 01:37 AM) Ok, I take back the I was actually researching on this "burning in" stuff to enable higher overclocking. Some people will swear by it but some experts in the field thinks it just a myth. But yeah, I would like to try burning in and see if its true. Only wish I had the cash to afford a replacement chip in case I toast mine (quite likely), which unfortunately I don't have $$  Yeah, I know about the burn-in stress testing done at the manufacturing line. Though from what I know its only to check for defects in the chip, and not to improve its capabilities (assuming its possible). Anyone would like to give their stories of burning in giving an improved overclock? Please give your own experiences only, or refer a link to someone's experience of their own. There is one factor I would like to eliminate though. Most of the burn-in before and after stories that i've heard do not actually use the same hardware components or the same settings. Even when no components are changed, the BIOS may have been updated, which in itself sometimes improves overclocking results. I would like to see a case where burning in gives an improvement in overclocking where the system is not changed at all; which is quite difficult to imagine, as overclockers will just tweak and tweak nonstop.. There was one good article about burning in by a so-call expert, will try to find it at post it here. Now have to go to work.... |
|
|
Jul 28 2007, 08:49 PM
Return to original view | Post
#10
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(likito @ Jul 28 2007, 08:26 PM) my overclock step! Looks ok to me. The reason you are increasing the vcore is because orthos fails right? And a kind reminder, don't push your proc to hard until you get your new HSF. 1.75v is way too high even for water-cooling!!!default fsb 200mhz , default vcore 1.35 , default vdmim : 1.8v increase 5mhz each time . after that run orthos 8+- minute until 230mhz , increase vcore to 1.37 after that run orthos 8+- minute 235mhz , increase vcore 1.4 after that run orthos 8+- minute any problem with my oc step ? This post has been edited by kmarc: Jul 28 2007, 08:49 PM |
|
|
Jul 29 2007, 10:25 PM
Return to original view | Post
#11
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(byfc2010 @ Jul 29 2007, 10:19 PM) tat ain't gonna go far for ur mobo.... Yeah, especially if your rams are rated for vdimm up to 2.2v! even my vdimm 2.03 cannot do anything for me...  |
|
|
Aug 20 2007, 05:56 PM
Return to original view | Post
#12
|
|
Elite
14,576 posts Joined: May 2006 From: Sarawak
|
QUOTE(likito @ Aug 20 2007, 05:29 PM) 1.45v , 250x10 2.5ghz , not stable . sien .... You sure it is your CPU? One way to know whether the CPU is the limit is to increase the vcore another notch e.g. 1.475v and see whether a higher OC is possible. If it is possible, then you know that the CPU is the limiting factor for the OC.What is your current stable OC at 1.45v? |
|
Topic ClosedOptions
|
| Change to: |  0.0328sec
 0.40
 7 queries
 GZIP Disabled
Time is now: 4th December 2025 - 08:03 AM |
Quote