Suspicious OC figures

I've just finished my first attempt at overclocking and seem to be running stable but with a lower than expected voltage and core temp.

GPU: i5 3570k
fan: Noctua NH D14
mobo: ASUS Sabertooth Z77
PSU: Seasonic Gold 750
RAM: 16Gb Corsair Vengeance 1600DDR3
Case: Cooler Master CM 690 II

I basically did a few simple steps,
1) Set RAM to XMP @ 1600DDR3 and then never touched it again.
2) Set LLC to 50%
3) Set Vcore to offset +0.01
4) Increased multiplier up to 44 in small steps

Without doing anything else, I am able to run Prime95 blend test @ 4.4GHz for 12 hours with core temps averaging 59C and peaking at 63C.
My Vcore runs at 1.125V at full load.
At idle the temps drop to 30C and vcore drops to 1v @ 1.6GHz

Is this normal behaviour? From reading other people's threads, I was expecting much higher temps and needing to use a higher Vcore as well.
Also, Have I skipped any important steps as the whole thing seemed too easy?
Aside from the fans running a little louder at idle, everything seems fine.
4 answers Last reply
More about suspicious figures
  1. Hi,

    Seems like you have a good chip, you could always run small FFT in prime95 just to make sure the CPU is stable although I would imagine it as after 12 hours of blend.

    Overclocking these days is very easy.
  2. You've got one of the best air coolers in the world on your chip, and the 3570k overclocks WAY better than the 3770k.

    Your chip is behaving more or less like I'd expect it to.
  3. Your right index finger stutters? :)
  4. Quote:
    That's completely untrue. They overclock the exact same, although hyperthreading will create a little extra heat. "Way better" implies you just have no clue what you are talking about.


    *Looks at sig*

    Yeah, no clue. It's not like I've taken multiple chips to 5 GHz+ or anything, including the 3770k (didn't get great silicon, had to disable two cores to get past 4.9 GHz).

    Look, "way" may have been overstepping things a more than bit but the 2500k/3570k does overclock better than its big brother, especially when you're limited by lack of thermal headroom on notoriously hot chips like IVB. Additionally, hyperthreading results in more work being done on each core, which results in even more heat output, so an HT enabled chip can consequently reach notably higher thermal peaks under identical workloads, further exacerbating issues relating to excess heat.

    This reminds me of a professor I had back in college, always joking about the frictionless pulleys and massless ropes the physics department used for their thought experiments. Given a perfect cooler, then sure, they'd overclock "the exact same" but this is the real world, and "a little extra heat" matters, just like in the real world engineers have to care about how much the rope weighs and how much friction the pulley has.

    [Edit] Cute, you called somebody out who knew more than you, made an ass of yourself, got schooled, then deleted your post to pretend it didn't happen. Next time you want to act like a condescending twat, be sure you know what you're talking about, okay? I mean, Christ, all you had to do was look at my sig to see that I'm an experienced overclocker. It was sitting right next to the quote button!
Ask a new question

Read More