Find fields & stores near you!
Find fields and stores
Zipcode
PbNation News
PbNation News
Community Focus
Community Focus

Reply
 
Thread Tools
Old 05-10-2014, 02:19 AM #1
ttimberlak443
aaaaaaaand its gone
 
ttimberlak443's Avatar
 
Join Date: Apr 2012
Location: Stephens City, VA
ttimberlak443 supports Team VICIOUS
ttimberlak443 posts videos on PbNation
ttimberlak443 plays in the APPA beginner division
ttimberlak443 has achieved Level 3 in PbNation Pursuit
4670k OC questions

System Specs:

Intel 4670k
NZXT Phantom 410
Corsair CX600
ASRock Extreme 4 Z87 mobo
Corsair H100i
EVGA GeForce GTX 660 SC
8GB G.Skill Trident X F3-2400C10D-8GTX
Samsung 840 120GB SSD
Seagate Barracuda 1TB
ASUS Xonar DGX

I HAVE UPDATED THIS TO CURRENT SETTINGS
Current BIOS settings:

Multiplier is at 46
VCore is fixed at1.240v
CPU Input is fixed at 1.9v

It has passed 31 1/2 hours of AIDA64 stress testing and 20 passes of IBT on max. IBT was used for checking for basic stability and thermal output. AIDA64 was used to check for longer term stability.

I'm getting about 85C max on IBT and around 65-70C during AIDA64. It sits around 60C during long gaming sessions on more CPU intensive games like Total War.

I had two BSODs during ARMA III free weekend while I was recording using FRAPS. That is the only game that I had issues with and it worked fine most of the time. I haven't had any issues with any other programs or games.

-Could the ARMA III BSODs be a result of the OC or could something else have been causing the BSODs?
-Is there any other sort of stress testing that could be done to check for long term stability for Haswell? I have heard that Prime95 is no good for Haswell.

This OC was going fine before the ARMA III incident and has been perfectly fine since afterwards.

Last edited by ttimberlak443 : 05-25-2014 at 09:30 PM.
ttimberlak443 is offline   Reply With Quote
Old Sponsored Links Remove Advertisement
Advertisement
Old 05-25-2014, 05:10 AM #2
fatesdemise
 
 
fatesdemise's Avatar
 
Join Date: May 2014
88C, seems high for that speed, but then again ive never cared for those premade water cooling system in a box coolers.. often times a regular fan cools better.

I have a lower model haswell running at 4.2ghz and I don't break 40C on stress test
ambient temp about 70F

I would get some artic silver thermal paste and see if that helps. If not then you need a better cooler to get your temps down

Last edited by fatesdemise : 05-25-2014 at 06:14 AM.
fatesdemise is offline   Reply With Quote
Old 05-25-2014, 07:39 PM #3
ttimberlak443
aaaaaaaand its gone
 
ttimberlak443's Avatar
 
Join Date: Apr 2012
Location: Stephens City, VA
ttimberlak443 supports Team VICIOUS
ttimberlak443 posts videos on PbNation
ttimberlak443 plays in the APPA beginner division
ttimberlak443 has achieved Level 3 in PbNation Pursuit
Quote:
Originally Posted by fatesdemise View Post
88C, seems high for that speed, but then again ive never cared for those premade water cooling system in a box coolers.. often times a regular fan cools better.

I have a lower model haswell running at 4.2ghz and I don't break 40C on stress test
ambient temp about 70F

I would get some artic silver thermal paste and see if that helps. If not then you need a better cooler to get your temps down
I was able to get stable (from what I can tell) at 4.6ghz 1.24v. I set a more agressive fan curve since then and I didn't hit higher than 85C during IBT. I also lowered the RAM frequency to 1600mhz with 7-8-8-24 timings, per G.SKILLS recommendation.

What baffles me is this OC passed 31 1/2 hours of AIDA64 and 20 passes of IBT on max, but I got two BSODs (0x124) during the last day of ARMA III free weekend while playing ARMA III. It didn't BSOD during any other game or stress test. Granted, I'm pretty sure I was recording with FRAPS during both BSODs, so perhaps something else was causing the BSOD.

EDIT: This current OC has been going for about a week now perfectly fine (after the ARMA III incident) and the temps while gaming never really go above 60C

Last edited by ttimberlak443 : 05-25-2014 at 07:44 PM.
ttimberlak443 is offline   Reply With Quote
Old 05-25-2014, 07:58 PM #4
DayoftheGreek
I hate you
 
DayoftheGreek's Avatar
 
Join Date: Jul 2004
Location: Burlington, NJ
 has been a member for 10 years
One of the things people don't tell you about stability testing is that there is no substitute for time. No matter how hot, or how much load some program puts on the CPU, when you start talking about the math and statistics behind very low fail rates on the edge of stability, it takes a LONG time to draw them out.

Want to be really stable? Use something like folding at home for a week straight. I did every stability benchmark under the sun, but when I tried to get some constant folding going, it would crash after two or three days every time. I ended up bringing my voltage from something like 1.2 to 1.35, just to sort out that very last bit of stability. Since then, I haven't seen a blue screen in 2+ years.
__________________
FEEDBACK: http://www.pbnation.com/showthread.p...hreadid=665897

1/30/08: I was there when the plane flew but everyone was still stupid.
DayoftheGreek is offline   Reply With Quote
Old 05-25-2014, 09:10 PM #5
Promagistrate
such motto
 
Promagistrate's Avatar
 
Join Date: Oct 2010
Location: CA
Promagistrate owns a Planet Eclipse Geo
Promagistrate has achieved Level 1 in PbNation Pursuit
Promagistrate has achieved Level 2 in PbNation Pursuit
Promagistrate supports DLX Technology
Quote:
Originally Posted by DayoftheGreek View Post
One of the things people don't tell you about stability testing is that there is no substitute for time. No matter how hot, or how much load some program puts on the CPU, when you start talking about the math and statistics behind very low fail rates on the edge of stability, it takes a LONG time to draw them out.

Want to be really stable? Use something like folding at home for a week straight. I did every stability benchmark under the sun, but when I tried to get some constant folding going, it would crash after two or three days every time. I ended up bringing my voltage from something like 1.2 to 1.35, just to sort out that very last bit of stability. Since then, I haven't seen a blue screen in 2+ years.
bingo. this is why IBT isn't a very good indicator of OC stability.
__________________
wat
wat
wat
wat
wat
Promagistrate is offline   Reply With Quote
Old 05-25-2014, 09:22 PM #6
ttimberlak443
aaaaaaaand its gone
 
ttimberlak443's Avatar
 
Join Date: Apr 2012
Location: Stephens City, VA
ttimberlak443 supports Team VICIOUS
ttimberlak443 posts videos on PbNation
ttimberlak443 plays in the APPA beginner division
ttimberlak443 has achieved Level 3 in PbNation Pursuit
Quote:
Originally Posted by Promagistrate View Post
bingo. this is why IBT isn't a very good indicator of OC stability.
I was just using IBT for basic stability and heat output. I was using AIDA64 for the longer term stability. What other stress tests should I look into for long term stability besides AIDA64? I heard Prime95 is no good Haswell.
ttimberlak443 is offline   Reply With Quote
Reply


Thread Tools

Posting Rules
Forum Jump