Overclock.net banner

1 - 20 of 23 Posts

·
Banned
Joined
·
2,667 Posts
Discussion Starter · #1 ·
To deem stable processor?tired of running prime for so long so running linx now,doing 50 runs 25335 for problem solved ram is 4917
 

·
Premium Member
Joined
·
2,289 Posts
yea pretty much but then again some machines after doing 50 passes fail blend.. just shoot a short 1 hour blend if anything
 

·
Registered
Joined
·
567 Posts
the answer is subjective, do as much testing as possible.
 

·
Registered
Joined
·
2,050 Posts
It is enough to say "this overclock is worth testing for real", but I am a stability hardliner / fundamentalist, lol. Everyone has there own ideas, and probably almost none of them, including my own, are worth much. If the flawless operation of your rig is critical, and you want to make absolutely sure your cpu won't degrade (and OMG it is a 980x), do more like ~200 runs. I recommend shooting for "factory" stable, ie. as stable as with no OC.

Oh, and just as important as doing many LinX runs, is doing a diverse array of tests. LinX (with as much memory as possible, doing a single run at "all memory" can free up more RAM for next runs.), Prime95 Blend, Small and Large FFTs, Memtest86+ v4.0, and even a combination of these run simultaneously.
 

·
Banned
Joined
·
2,667 Posts
Discussion Starter · #6 ·
Quote:


Originally Posted by Ktulu
View Post

How long did you run Prime? On this last build Linx wasn't picking up errors when Prime was.

i ran prime for 1 hour got boring so went to linx

just using a different multi and vcore to see if it is stable hjere at 1.42 vcore oppose to before using 1.46 with a different multi
 

·
Banned
Joined
·
8,014 Posts
25 on All or Very High is enough for me and has never given me problems. I got tired of the Prime 95 BS and having to run it more than 12hrs~ then having to re run it on Blend. Screw that, waste of time IMO. IBT or Linx FTW!!!!1!1
 

·
Banned
Joined
·
2,667 Posts
Discussion Starter · #9 ·
Quote:


Originally Posted by 88EVGAFTW
View Post

25 on All or Very High is enough for me and has never given me problems. I got tired of the Prime 95 BS and having to run it more than 12hrs~ then having to re run it on Blend. Screw that, waste of time IMO. IBT or Linx FTW!!!!1!1

thanks what i am going to do is after this 50 run ,ill run occt for a hour.
 

·
Banned
Joined
·
8,014 Posts
Do 25 runs, play lots of games on it, use it regularly, if you don't BSOD whithin a day or 2, it's probably fine. Has been the case for me, just my 2 cents.
 

·
Premium Member
Joined
·
9,349 Posts
50 runs should be good with LinX, but I would run the three Prime 95 tests (blend, small FFT's, large FFT's) also - for at least 8 hours each as well. That's just me though. I mean, you can never really be sure of 100% stability, but from my experiences those tests/times have proven effective for me.

EDIT: I also use a Memtest86 boot CD to check for RAM stability.
 

·
Banned
Joined
·
2,667 Posts
Discussion Starter · #13 ·
after i run linx ill run occt then tomorrow when i go out ill let prime run for a few hours

linx is almost done anyway 40/50
 

·
Premium Member
Joined
·
9,349 Posts
Quote:


Originally Posted by Rud3Bwoy
View Post

after i run linx ill run occt then tomorrow when i go out ill let prime run for a few hours

linx is almost done anyway 40/50

Oh yeah, forgot about OCCT. It's a really good stress tester as well IMO.
 

·
Registered
Joined
·
2,050 Posts
Quote:

Originally Posted by chatch15117 View Post
Lulz @ vcore. Sorry you didn't get a good chip though
. They should be binned better than that for $1k.
That is not a high vCore for 4.4 GHz with HT. Most people just don't do enough stability testing, so they think they need less vCore than they really do. This is purely my opinion, obviously.
 

·
Registered
Joined
·
567 Posts
50+ LinX runs with MAX memory
12hr+ Prime95 Blend test

Then call it stable
 

·
Banned
Joined
·
5,000 Posts
Quote:

Originally Posted by Clockadile Dundee View Post
That is not a high vCore for 4.4 GHz with HT. Most people just don't do enough stability testing, so they think they need less vCore than they really do. This is purely my opinion, obviously.
The updated Intel spec for the 32nm is 1.40v maximum - Versus the 1.55v for 45nm chips. But yeah, for 6 cores HT that is respectable.

I was expecting 1.35v or less for 4.4GHz.

And you're right about the stability testing. I let Linx go for 12+ hours before I even consider the possibility of calling it stable. I've had settings fail 6-8 hours into the runs.
 

·
Registered
Joined
·
2,050 Posts
Quote:

Originally Posted by chatch15117 View Post
The updated Intel spec for the 32nm is 1.40v maximum - Versus the 1.55v for 45nm chips. But yeah, for 6 cores HT that is respectable.

I was expecting 1.35v or less for 4.4GHz.
That is interesting. But you see my point though, in a general sense? I guess you could say I meant "for an i7 920", specifically.
 

·
Banned
Joined
·
5,000 Posts
Quote:

Originally Posted by Clockadile Dundee View Post
That is interesting. But you see my point though, in a general sense? I guess you could say I meant "for an i7 920", specifically.
Yeah, definitely. I get random BSODs at 4.45GHz, 1.408vcore(last bsod was 0x0a
). It's a lot better than my i7, although my batch is awful
.
 
1 - 20 of 23 Posts
Top