5930k IntelBurnTest results

[stayplation3]

Why does my IntelBurnTest results keep turning out like this? http://imgur.com/zNAsGw0
Instead of a steady speed its just all over the place. I was watching the clock speed it was a steady 3599MHz the whole time.

Nevermind disabled some weird settings in bios and started getting good results. http://imgur.com/x3T8Y6W

Why is my cpu speed 3599MHz? I didn't overclock it and all the OC settings on the mobo are disabled.




***************I've fixed all the problems above****************








Also my max cpu temp is 52 degrees, is that good? Thats using a thermaltake water 3.0 ultimate using the pre-applied paste it came with.

 

[xorbe]

Just curious what setting you changed.

 

[Aquinus]

The 5930k has a base clock of 3.2Ghz and a boost of 3.8Ghz. On the average (or just the core it's looking at,) may be running at 3.6Ghz. If you completely disable turbo boost you should only see 3.2Ghz.

 

[stayplation3]

the base clock is actually 3.5ghz and i figured out the problem there was another thing called eist i didnt have disabled. im in the process of overclocking now. also does anyone have an answer to my original question? is 52 degrees a good temp on a 5930k not overclocked? i got 52 degrees as my maximum temp before i started overclocking

 

[sneekypeet]

Temps are fine for IBT. As you clock heat will become a factor. While it can run all the way to throttle at around 100C, most users feel more comfy in the low to mid 80s for an OC'd run at IBT.

 

[GhostRyder]

the base clock is actually 3.5ghz and i figured out the problem there was another thing called eist i didnt have disabled. im in the process of overclocking now. also does anyone have an answer to my original question? is 52 degrees a good temp on a 5930k not overclocked? i got 52 degrees as my maximum temp before i started overclocking

In a burn in test, yes that is a good test since that is a hefty level of stress on a CPU beyond normal tasks. My 5930k in a custom water loop which includes 3 R9 290X cards when clocked at its 4.5ghz overclock on all cores it gets up to the 60's. When gaming it stays around there in games like BF4 stressing everything at 4k (I have seen it hit 70 before) which is all still way below the limit. Normally you only want to start worrying if you hit 80+ Celsius as that is starting to run a bit warm but its still below the temperature max which is 100c for throttling.

 

[stayplation3]

1. Okay well I've overclocked to 4.5GHz and left the voltage on auto, the voltage gets up to a very steady 1.244v. What do you guys think about that voltage? Does it seem low, good, high? Should I leave it on auto or do you think I can lower it if I go manual?

2. My temps with this overclock reach a maximum of 75 degrees. Does that seem normal to you with my cooler?

3. Also something weird seems to be going on with the fans on my radiator. When stress testing and temps are around 60-75, the fans will rev up for about 4-5 seconds then down for 4-5 seconds, then repeat over and over. The fan speed isnt changing drastically, only a little. From its set point of 75% to 87.5% I have it set to 75% fan speed above 50 degrees, then 87.5% fan speed above 80 degrees. It never reaches 80 degrees so it shouldn't rev up to 87.5%, the way I have it set up it should stay at 75% during the entire test.

4. I am using IntelBurnTest for stress testing. Is this safe to use with Haswell-E CPUs? I've heard that others like Prime95 are bad for these CPUs but haven't seen anything about IBT.

Okay, it would seem that my maximum overclock is 4.5GHz. I tried to push it more but going from 4.5 to just 4.7GHz requires a voltage increase from 1.244v to 1.4v. Usually that means I've reached my maximum overclock right? Is there anything else I can do to get a higher clock? All I've done is increase the multiplier.


My Titan X temp is 83 degrees running Furmark. Im about to install an H110 on it with the G10. Wish me luck!

 

[cadaveca]

IBT? Furmark? Oy vey...

Stop. Just stop. try using AIDA64's stability test, 3DMark Vantage, 3DMark (all 4 tests), and then a game like BF4 or Dying Light.

IBT and FUrmark do not in any way tell you anything about stability...they simply tell you about max temps. But any real load will have temps like 20c less than what those do... so using them to test is rather wasteful on your personal time.

BTW, you turned too much stuff off... this is a new platform, so new approach to OC is required. You loss of stabiltiy about 4.5 GHz tells me that you probably need to also adjust either power limits in BIOS, or vInput.

you can check my Haswell OC guide posted on the front page...it all pretty much applies with Haswell-E. I'll have a Haswell-E-specific guide eventually, once I've got more CPUs and mroe testing done.

 

[xorbe]

Prime95 and IBT should run without crashing though ... you never know what the CPU might be required to do ... at least check for 3-5 minutes at full tilt ...

 

[cadaveca]

Prime95 and IBT should run without crashing though ... you never know what the CPU might be required to do ... at least check for 3-5 minutes at full tilt ...

Unfortunately, software errors can mean that the app can crash, even though the system is stable.

AIDA64 does offer AVX-based testing just like IBT/P95. But I really think you should be testing a system based on the workload it will be under, not use tools and algorithms that'll never pass across those circuits ever again.


So, I guess I'm really just saying that real-world-based stability testing is the way to go. Another GOod app to try is ASUS RealBench, which offers different encoding workloads.

 

[sneekypeet]

RealBench will show instability quickly too....Handbrake is no joke for usage!

 

[vega22]

xtu aint a bad app to test oc and stability too.

but realbench is very much more real world usage.

 

[stayplation3]

well im using those programs to get max temps. even though they will never reach those temps again, i like the feeling knowing that even under the most extreme circumstances that they will still be nice and cool.

I just installed the H110 on the Titan X with near complete failure!!!! The stupid hoses on the H110 are too short! i cant mount the radiator properly so its just sitting in the case zip tied down. AND to top it off, the 2 fans the H110 came with are not only both 4 pin, but it didnt come with a Y connector too. So this is what happened. I plugged the fan on the kraken G10 into the mobo, and the H110 pump into the mobo, then all i had left was 1 4-pin spot on the mobo. So i put 1 of the 4 pin fans the h110 came with on that last spot on the mobo, and put that fan on one side of the radiator, then on the other side of the radiator i connected 2 fans, one on top of the other, and plugged both of those right into the PSU, they are silent fans so its still quiet even at max speed.

what i am going to do is order a 4 way splitter so i can plug all 4 radiator fans into that last 4 pin slot i have left on the mobo, that will fix my fan problem. as for mounting, right now i am thinking i might just leave it zip tied down forever...

the only way i can mount it is if i buy a 140mm closed loop cooler and i can install that in a different spot that i can mount for sure.

right now my titan x is idling at 24c, im going to run furmark and let you know what temp it got to.

 

[cadaveca]

RealBench will show instability quickly too....Handbrake is no joke for usage!

That's why I've been using it for so long in my reviews. I've been encoding the same video clip since the very beginning, and what used to take 7 minutes now takes 3...so I now dual-pass wit hevery run. I was really glad to see ASUS use it as well in their test.

right now my titan x is idling at 24c, im going to run furmark and let you know what temp it got to.

Just be sure to watch clockspeeds, since furmark can load a card in such a way that the clocks it runs at either decrease, or are lower than what real-world will give you, making it rather useless overall. My 780 TI's, even under water, will hit 853 or whatever stock is, but then load up a game, and they hit 1020.... Thankfully the new NVidia cards have much higher power limit adjustments available, so maybe you'll not experience it, but it's still something to look for. Both temps and power usage can cause these cards to adjust clockspeed.

 

[stayplation3]

i ran it for about 30mins and it got up to 40c and i watched the clock speed it was normal 1200mhz like its supposed to be the whole time. ima OC it now and see what i can do with it.


anyone know a great program to OC gpu's? im using msi afterburner
ok i need a different program to OC. anyone know a good one? msi afterburner isnt letting me adjust the voltage, even after i checked the box "unlock voltage control"

How do I overclock my GPU more? I got msi afterburner set to +200 core clock and +200 memory clock and the power limit is set at 110% which is the max. If I go more than 200 it crashes I think I need to raise the voltage. Am I not allowed to raise the voltage on the Titan X? Or is there just a problem with msi afterburner not letting me do it?

 

[RealNeil]

Sapphire Trixx does voltages.
As for the fans, try using one of these for your fans.
It will run them at max speed, but if you have quality fans, you will probably not notice the noise and the cooling will be better overall.

Phobya 4 Pin to 6X splitter

 

[stayplation3]

ive got a 4 pin CPU fan connector on the mobo i can use. i want something that i can plug into that, then splits off into 4 4-pin male connectors so i can have all 4 fans plugged into the 4-pin CPU spot on the mobo then they wont run full speed all the time. anyone know what i am talkin about? i need to find one to buy that i can connect at least 4 fans to



Mwuhauahauahahaha just found this > http://www.newegg.com/Product/Product.aspx?Item=N82E16812311001

 

[RealNeil]

Bad Idea.
The MoBo's CPU fan connector doesn't provide enough power for four fans.
The fourth wire in a four pin connection provides fan PWM speed to the connected fan.

Multiple fans will throw this off.


I have good results with Cougar Vortex PWM fans connected to the splitter that I linked to. They run at full speed, but they're quiet as can be and they move a shitload of air.

EDIT: Just saw your link to the Evercool Cable. That will work, but PWM readings will be off.

 

[xvi]

The fourth wire in a four pin connection provides fan PWM speed to the connected fan.

Multiple fans will throw this off.

+1. Not a great idea. The more fans you add, the less control you'll have over fan speed. I'm not sure of the technical reasons, but splitting a PWM signal seems to degrade it to the point where all fans end up running at full speed anyways (usually about 4-5 fans).

 

[stayplation3]

well my radiator on my cpu has 3 fans connected to a 3 way splitter and it runs flawlessly, i dont see having 1 more fan giving me all the trouble you guys are thinking about.

 

[RealNeil]

Well, you should do it then. It may work without any problems.

The splitter that you linked to will certainly provide plenty of power to your fans, But I'm not sure how accurate your sensor readings will be.