I am prime/game/bench solid at 2.7,246x11,6-2-2-2, 1.55 volts is needed though.
watercooled so im trying to find out whats safest volts under water.
Printable View
I am prime/game/bench solid at 2.7,246x11,6-2-2-2, 1.55 volts is needed though.
watercooled so im trying to find out whats safest volts under water.
Game solid I thought I had that covered too I played EQ2 for over an hour no issues but BF2 was a different story. I might just try lowering ram/fsb to 246 and 11 multi I was hearing that sometimes multi and FSB affect OS dramatically.Quote:
Originally Posted by Usurper
At least with BF2 I can get on an empty server and test everything for stability, I don't know about the 1.55v I see people @ 1.6-1.7 but are people >1.5 just bench testing or do people leave 24/7 running at higher 1.55-1.6v and is that safe?
I'm running Prime95 right now and at 56'C 1.525v 260x10 no errors for 2 hrs in Prime95 with memory testing as well, high temp is due to ambient temp in house currently 83'F it is not that high after work when I'm home though.
What part of the stepping do you look at to see if its an E4 or E6 core?Quote:
Originally Posted by hag6br
1.55-1.6 is fine. Anything beyond that may get a little hairy. Also take into consideration that every cpu is unique and just because you saw somebody run spi1m @ 1.95V, it doesn't mean that yours can too ;)
ha, yes I know exactly what you mean that's insta-death voltage right? I am just curious because I knew the max safe volts on P4 but wasn't sure of the AMD X2, (my first AMD ever)Quote:
Originally Posted by s7e9h3n
I just don't want to toast a fairly expensive processor 1.55 even 1.6v you think 24/7 is perfectly safe, that would definately solve my stability problems for sure. I run the XP-120 with a 104 CFM fan and 3x120mm 1x92mm fans in case very good air movement.
Got a 4400+
CCBWE 0521RPMW
1264618F50205
Damn thing isn't what I hoped...
OC's to 2.57Ghz at 1.45v
Finally, last night with some loose memory timings I was able to go 2.6Ghz at 1.5v (dual prime stable for 12hrs)
I'm using an xp-90. Am considering some ducting to get fresh air to the cpu as it get's freakin' hot!! Load at about 59c!!!!
By the way, it's still primin' while I'm at work today... by the time I get home it will hopefully be 100% load at 2.6Ghz for 24hrs!!! I'm hoping to increase it a little bit... 2.64 will allow me to use another multi for my memory and keep my good timings.
Thanks all.
Hm anyone got a 3800+ X2 yet ? I wonder how they OC..
So far i've been able to get 2.6Ghz Prime stable on 1.45V but i can't go much higher then that even with 1.55V and temps start to really climb then. My load temp right now at 1.45V is around 45-46C with an XP-90.Quote:
Originally Posted by Ender17
Under Cpu-Z look at the Revision Box.. It will tell you what it is..Quote:
Originally Posted by burningrave101
it`s blanked out on mine.Quote:
Originally Posted by hag6br
http://img.photobucket.com/albums/v6...x2firstOC1.jpg
Version 1.29 of CPU-z...Quote:
Originally Posted by Rukee
http://www.cpuid.org/index.php
strange, that one shows mine as a E4, but opens with an error everytime and the multiplyer/fsb/memory are all blanked out now.Quote:
Originally Posted by [timko]
So which are better to have??? the E4 or the E6?
If I can get a 4200 oem for $499, should I????
I have a 4400+ and mine is an E6. The only difference between the E4 and the E6 is the cache size right? There aren't any other improvements in the E6 is there?Quote:
Originally Posted by hag6br
zip zoom fly has retail ver for that price, i just ordered one. Im hoping that will hold me till yohan's second gen comes out (dektop edition) or something really sweet from amd.Quote:
Originally Posted by mad mikee
I looked.. Per AMD, All 4400+ and 4800+ are E6 cores.Quote:
Originally Posted by burningrave101
But for 4200+ and 4600+, they list both an E4 and E6 core for them..
When your testing for stability make sure you do some runs with p95 on large inplace fft's... i primed 14 hours on the small inplace ffts at 2640mhz or so, no problem.. but within 15 minutes of running large ffts i errored out. And in CSS after about 35 min or so id crash.
Wow this cpu runs hot... especially with the IHS on.. here is my OC so far, just over 2607mhz 237x11. Bios says 1.5v, SmartGuardian reporting 1.45V. Is there any danger in running my PWM IC at ~65C under full tilt load? I don't know if I am going to be able to fit a fan down in there to keep them cool.
Also I reversed my fan on my xp120 (now blowing away from the cpu), and my cpu temps dropped probably 6-7 degrees, of course YMMV.
Memory is Mushkin Level 2 PC3500.. running at 2-3-3-8 @ 3.1v, 1T
Update-
Ran large fft's overnight looks like I am prime stable at 2600 *hurray*. I also dropped in a 60mm fan blowing right onto the PWM area so the max I've seen it is 59, but it tends to stay about 57. Cpu is now maxing at around 61-63 with the fan on my xp120 blowing away from the cpu.
Seems allot of early 4200 & 4600's are toledo cores that where cache crippled which would explain why there are E6 4200 & 4600's.
http://www.anandtech.com/cpuchipsets...oc.aspx?i=2484
That could mean not enough chipset voltage or unstable memory though - or perhaps your mobo just doesn't like the clock. Your CPU clock is stable I'd say after 14hours small FTT's so something else is causing that.Quote:
Originally Posted by halcy
Seems weird that they would do that considering that it's much more expensive to produce the 1mb cache with respect to yields etc but then again, big corporations have always bemused me - I'll leave that to the city people :DQuote:
Originally Posted by highoctane
I still dont know why Anandtech compared the power consumption of that 4200+ to the 3800+ even if it is an E6 core because half its cache would be disabled and not working anyways. They should of compared to a working 4400+ or 4800+.Quote:
Originally Posted by highoctane
i am wondering abt this too...........Quote:
Originally Posted by sai
Here is a benchmark everyone can use to compare their SMP performance.
http://www.hardforum.com/showthread.php?t=858272
yeah, well, Anandtech does an awful lot comparisons that don't make a whole lot of sense. Plus, they don't even run a divider on the memory when trying find a motherboards highest ocwhen they review them.