[SOLVED] 9900K OCed to 5Ghz - Smoking Hot

[Regev]

Hey guys, I tried to OC my CPU for the first time. My PC:

9900K + Noctua D15 + Gigabyte Aorus Z390 I Mini ITX, inside a Silverstone LD03. No GPU. PSU is a Platinum+ 750W Silverstone.

I set the ratio to 50, ring ratio to 47, voltage to 1.265V, but the system fails (blue screens) when I run Intel BurnTest (30 Loops, Standard). Ramped up the voltage to 1.3V, still fails. Set it to 1.4V, still fails. HOWEVER, when I set the voltage on Auto, it does pass the test. The problem? All cores' Max Temp recorded is 95 to 100.
On stock speeds they reached a maximum of 80 on that test.

What am I doing wrong, is the chip simply bad lottery (I see folks hitting 5Ghz on 1.265 no probs), and is there anything I can do?

Thanks <3

 

[Karadjgne]

You need an education it sounds like. Just setting vcore, ring voltages and the multiplier doesn't always work. There's LLC to consider, vid, vccio, vccsa, turbo current power limits short and long, and a heap of other possible settings. It's also important to know where to put the settings. Just slapping LLC to extreme or 1 or 6 isn't going to work.

 

[Regev]

That's just natural variation, at least I think so, the crystals and circuits aren't that exact to keep the same exact clock at all time.

Yeah, just why isn't it on 3.6 when it's idling ?

 

[Regev]

You need an education it sounds like. Just setting vcore, ring voltages and the multiplier doesn't always work. There's LLC to consider, vid, vccio, vccsa, turbo current power limits short and long, and a heap of other possible settings. It's also important to know where to put the settings. Just slapping LLC to extreme or 1 or 6 isn't going to work.

Any guide for these? That guide I was linked to previously on this thread I noticed only talks about the main settings

 

[TerryLaze]

As I said, speedstep and or turbo, it could be called something else for your mobo.
Maybe if you manually change multiplier your mobo ignores these settings.

 

[Regev]

Ok guys, I followed the OC guide on TomsHardware to the letter, here's what I got so far, it's so weird:

On stock settings, the CPU package gets to 86c on HWiNFO after 15 mins of Prime. That, remind you - on Auto voltage, and everything else, etc. It also runs on 4.7K all cores it seems. However, lowest voltage I could get it to run stable on (for 15 mins of Prime + 1 hour of RealBench) is 1.230V. Runs for 4.7K all cores - CPU package tops at 75c (11c less!) . Then I tried to up the freq to 4.8K, managed to do it with 1.27V minimum - reached 80c on the CPU package - stable after 1 hour of RealBench.
HOWEVER,
Today I ran another test on the 4.8K settings - failed this time! And failed on the 15mins Prime temp testing. And kept failing till I increased voltage to 1.29V I believe, reaching 83c.
So I don't understand two things:

1. Does the motherboard by default pump way too much voltage "just in case"? How come reducing the voltage from Auto (which was about 1.248V) to 1.23V result in 11 less celsius degrees, while keeping the same all-core frequency?

2, How come settings that result in stable 15mins Prime + 1 hour RealBench fail after 4 minutes of Prime 95 the very next day, even with 0.005 or 0.010V extra?

By the way - opening the case glass doors reduces another 7-9c

 

[zero_l0gic]

you need better cooling. i9-9900k OC'd can go up to 250W TDP and your D-15 is rated for 180-190ish W.
as i had the simmular problem, had overclocked my i5 to 5.0ghz (aircooled on Zalman CNPS10X PERFORMA dual fan) idle was around 60 and on loads it would go to 90+(thermal trottles alot). changed it for corsair hydro h100x (240mm) and boom idle on 35ish and load on 70ish. i woudnt Delid the CPU if you dont want to do EXTREAM OC due to problems stated above.

 

[Regev]

But how would it be stable on X settings (for 15 mins of Prime + 1 hour of RealBench), then the next day fail under the exact same settings after a few mins of Prime ?

 

[Afro_ninja199]

Hey guys, I tried to OC my CPU for the first time. My PC:

9900K + Noctua D15 + Gigabyte Aorus Z390 I Mini ITX, inside a Silverstone LD03. No GPU. PSU is a Platinum+ 750W Silverstone.

I set the ratio to 50, ring ratio to 47, voltage to 1.265V, but the system fails (blue screens) when I run Intel BurnTest (30 Loops, Standard). Ramped up the voltage to 1.3V, still fails. Set it to 1.4V, still fails. HOWEVER, when I set the voltage on Auto, it does pass the test. The problem? All cores' Max Temp recorded is 95 to 100.
On stock speeds they reached a maximum of 80 on that test.

What am I doing wrong, is the chip simply bad lottery (I see folks hitting 5Ghz on 1.265 no probs), and is there anything I can do?

Thanks <3

leave the ring ratio to stock, bump volts to 1.3 and then work backwards

 

[mdd1963]

5.0 is not a magic number so why must it be 5.0? Your trying to pump up the voltage for a just average binned chip is what it sounds like.

From looking at most OC related posts, ever since the 8700K, there seemingly has been a general mindset of 'if it ain't 5 GHz all-core, it's a sack of useless dung'...despite stock operation of 4.7 GHz all-core delivering ~98% of the same frame rates at 13C cooler temps... (my own 7700K jumped from 73C to 85C in Prime95 /small FFTs when jumping from 4.7 GHz all-core to 4.8 GHz, so I settled for 'only' 4.7 GHz all-core MCE operation)

 

[conofroc75]

As for why one setting works for one thing and nit another its hard to tell when you get up near the edge of stability weird <Mod Edit> happens you could've had a background update run or something or a different program was running in thew background. But as far as it not downclocking make sure to check on your windows power setting and put it on balanced. If its on high performance it will tend to stick the higher cpu ratio set. It also matters in the bios to what core mode is set in bios if it's turbo ratio or package or per core. But now if you have it set up to run a bit higher depending on your use case you can try and set it up to get the best single core ratio or have t set to run 4-6 cores at 5.0 since depending on your daily worklode not all of the programs will utilize all 8 cores all the time. I wish there was a better guide for getting like a 75% to 80% OC instead of just YOLO dew the dew!! your not living unless your hitting 5.0 or 5.1 all time. Or a decent guide towards what level of oc will be better for what your using on a daily basis.

 

[Karadjgne]

you need better cooling. i9-9900k OC'd can go up to 250W TDP and your D-15 is rated for 180-190ish W.

The Noctua NH-D15 is rated at 250w+. The only aircooler I can think of offhand that's rated higher is the Deepcool Assassin3, but that cooler isn't as efficient as the Noctua, so ends up slightly behind in cooling ability, even if having a higher cooling capacity, which are not the same thing.

 

[alceryes]

However, lowest voltage I could get it to run stable on (for 15 mins of Prime + 1 hour of RealBench) is 1.230V. Runs for 4.7K all cores - CPU package tops at 75c (11c less!) .

This is close to where I have my 9900K. It sounds like you need better cooling, to be able to pump the voltage higher while keeping temps lower, to be stable at 4800MHz+.
Is there a reason you want to go higher? Are you experiencing a performance deficit somewhere that you think a couple/few hundred MHz could help with. I have run my CPU at 5+GHz - some extra points in synthetic benchmarks (and A LOT of extra heat) is all your gonna see.

HOWEVER,
Today I ran another test on the 4.8K settings - failed this time! And failed on the 15mins Prime temp testing. And kept failing till I increased voltage to 1.29V I believe, reaching 83c.
So I don't understand two things:
1. Does the motherboard by default pump way too much voltage "just in case"? How come reducing the voltage from Auto (which was about 1.248V) to 1.23V result in 11 less celsius degrees, while keeping the same all-core frequency?

There could be several things causing this. From too low voltage setting (including vcore, vccio, vccsa, etc.), to LLC settings, to your memory controller being clocked too high, to insufficient cooler, etc.
Yes, motherboard (in general) push more voltage than needed when everything is set to auto.

2, How come settings that result in stable 15mins Prime + 1 hour RealBench fail after 4 minutes of Prime 95 the very next day, even with 0.005 or 0.010V extra?
By the way - opening the case glass doors reduces another 7-9c

Because you're giving it more voltage - simple as that.
It really seems like you are at the limits of your motherboard and/or CPU at whatever settings you have at that specific voltage, with that specific cooler. You haven't even gotten into the more advanced CPU settings like overcurrent protection, PWM switching, over temp protections, etc. But messing with these is like taking the safety off - you can easily kill your CPU if you adjust these wrong. You should stop at around 4.7/4.8 until you get better cooling and are able to keep your CPU temps lower.