Printable View
The OCCT stability test sucks(the 30 minutes one). Use the torture test. I've had a cpu pass the 30 minute stability test, but fail the torture test in less than a minute.
Prime95 takes way too long to crash just when you think you're stable. OCCT is good for quickly finding a speed that's just too high, then backing down slowly until it runs stable for a few minutes, then hours, then run your Prime95, etc.
The chip I got isn't a very good chip, 0528 E6.
290x9 @ 1.7v on a Sonic Tower (no fan) 47C. Ambient is quite cold at night though guessing round 20-22C.
This is NOT stable, I've only done 8M superpi, will do more detailed stability analysis soon.
:300x9 not stable. 1.731v on the multimeter and I'm reluctant to push it any more since its on passive.
WHOA...1.7v? What stepping? last 4 letters? I got a 0528 as well, but mine can do 290x9 at 1.4v, but it takes 1.56v to do 300x9. These are all prime stable. I was able to boot up at 330x9 with 1.65v, but when i ran prime for about 2 minutes, the temps scared me, so i stopped.Quote:
Originally Posted by ixce
Mine totally sucks! Its a week 26. Cant do more then 2550 at cas 2.5,3,6,3 mem timings.. I have to run at cas 3 just to go up to 265=2650Mhz. This Gskill TCCD LE didnt have this prob with my FX-55 chip.
I was thinking that maybe its one of the 2 512 sticks so i just tried one at a time and managed 275x10 and it was fine at the lower timings.. but just one stick.. I swaped the sticks and the other does the same so all i'm thining is that either this chip sucks or this one has a really bad mem controller or defective.
Its going back to the store for an exchange.. and after what i'm seeing i dont think i'll be jumpin on another E6 chip this time.
OMG eclypse, your avatar, what website!! I wanna buy one!
eclypse, does cpuz reports your e6 as ff2, like evets?
:D
http://www.thinkgeek.com/
http://www.thinkgeek.com/tshirts/frustrations/388b/
Ever great while i get asked this and every time its a bish to find the web site.. hehe
Enjoy!
Oh theres plenty of other shirts ext.. there.
I dont have the newest version of cpuz to see.. and that computer dont have an internet connect since the newest wb's for the vid cards took my only PCI slot for my wireless card..Quote:
Originally Posted by high5
I'll see if theres some why to get this done.
It's a 0528GPDW. I'm not on a good power supply at the moment because my new Seasonic doesn't work with it, but I will see if I can get a better OC.Quote:
Originally Posted by evetS-
I'm just not reading right or are you complaining about your chip being not so good and running it @ 1,73v with PASSIVE AIR COOLING...Quote:
Originally Posted by ixce
:slapass:
Quote:
Originally Posted by Oc-cba
If temps are under 50C I'm happy :), sonic tower is amazing....unless my thermal diode is borked.
Edit: I think PSU is holding me back, using an ancient antec smart power 20pin. Can't hit 2.385 @1.456v stable =\
Do all boards support reading from the Athlon 64's internal diode?
@Doom5:
If the S939 board can show any kind of cpu temperature, it has to read it from the A64's internal diode.
An external sensor can't be built in to the socket.
Received LBBWE 0528GPDW today. I have been testing for about 2 hours now, and this has to be the worst chip I have ever owned. 2.3ghz at 1.58v is the best it'll do, and that's not stable. Even raising fsb 5mhz at stock volts freezes. Runs fine at stock, and cool too. Idles at about 27C Primed for about 20 min at stock, temp only went to 36C. I was going to pop the top on this, but I think now it's ebay bound. I wont even offer it up here, unless someones wants a stock chip :lol:
30mn vs 9hrs come on man!Quote:
Originally Posted by evetS-
Anyway if you want extreme stability I agree prime is better, but OCCT is just realy more time efficient, even the 30min test provide a decent stability - definitively better than 32M. And I apreciate the graphs that gets out.
About torture beeing more efficient than I´ll have to look a this, but seams to me quite odd if it don´t use the same presigers for both.
Hi guys,
Just had the chance to test a revision E6 3000+ with the stepping code LBBWE 0530 EPAW. I only had a short time to test so i tried 300x9 with 1.45v and 310x9 with 1.55v, both failed to load windows. Shame really, i was hoping it would be better than my current LBBLE 0516 GPAW running at 300x9 1.45v.
I have a feeling a lot of E6 3000+ CPUs will be up for sale @ ebay soon. :lol:
hey guys, is 53 degrees C too high for a 3000+ venice under load?
try to keep it under 50C...maybe invest in a better heatsink or remove IHS will help
Holding the fort at 1.73v trying to get 290x9 stable. I put a fan on the sonic tower, temps are in the 44's but my max case temp is 53C? So if I keep it under 50C i can increase past 1.8v right?
looks like most of these E6 3000 suck sheeit! :yepp:Quote:
Originally Posted by xenolith
OMG !! E6 stepping is suck ??
Wow everyones getting crap chips now. I guess mine was a fluke huh? Oh well 117, you can't go wrong, I'm sure you'll make your money back selling it on ebay
Quote:
Originally Posted by ixce
I do not suggest going that high on air...
Oh, no, Zanr Zij said that E6 is a good stepping :DQuote:
Originally Posted by xs64