>>51704127
>97C
jesus christ, anon, back off of that. Keep it to 80C or so.

>>51704241
This was at 100% fan speeds. Is there anyway to reduce temps more? Or do I just suck it up and go down to 4.5ghz? Truth be told I just raised the voltage straight up to 1.25V when I got an error on a test.

>>51704127
If you want your cpu dead in a week then yes.

>>51704304
>tfw my i7-920's cooler went to shit
>ran at 100+°C
>for months
What a trooper

I've been using prime95 v28.7 which apparently gives unrealistic temperatures on new intel cpus. Is this true /g/? Should I use an older version and try again?

>>51704329
how could you not afford a $20 or something

>>51704302
That's at (or, lets face it, beyond) what the limit of what that heatsink can handle. Your options are to either get better cooling - at least one of the really enormous tower sinks, with two or three fairly loud fans on it, one of the better (2x120) closed-loop watercoolers, or a proper open loop - or you back off. Note that voltage is the bigger determinant of temps than clock speed. Heat dissipated increases (roughly) linearly with clock speed, and quadratically with voltage. Focus on finding the lowest voltage at which a chip is stable (which I usually define as 24 straight hours of prime) Many chips can manage a large increase in speed with only a slight voltage bump, but then need lots and lots of voltage to go only a little bit farther. You want to get yourself on the safe side of that inflection point.

>>51704127
Back in 2010 this thread would have 25 replies by now...

Miss the days before /g/ was all about obscure languages and consumer devices..

>>51704302
>Is there anyway to reduce temps more?
Not on air. 97C isn't a temperature you would want inside your case.

>>51704391
I miss when we had loli threads

>>51704386
I see, any comment on this however >>51704348
Because if that is the case I may be safe temperature wise

>>51704453
>"unrealistic temps"
that's kinda the point. It's a worst-case-scenario test, so you can be sure that if your CPU handles it, it can handle anything it'll see in a game, application, etc.

>97°C
Holy fuck that's bad
I was only hitting 75 on 1.44V 4.5GHz on a 2500k.
This was when my fan was blowing hot air back into the case AND my heatsink was caked in dust.

What the fuck are you fucking doing??

>>51704592

omfg, instant raging boner

>>51704591
Dunno bro its a certain test that causes that. Ran it now at 1.2V and it peaked at 91C at the same test again and for most other things its at a cozy 60C. That is why I'm asking if it's something with prime95.
You know what fuck it I think i'll download a older version and try that

"Do NOT use any Prime95 versions later than 26.6. Here's why:

Core i 2nd, 3rd and 4th Generation CPU's have AVX (Advanced Vector Extension) instruction sets. Recent versions of Prime95 run AVX code on the Floating Point Unit (FPU) math coprocessor, which produces unrealistically high temperatures. The FPU test in the software utility AIDA64 shows the same results.

Prime95 v26.6 produces temperatures on 3rd and 4th Generation processors more consistent with 2nd Generation, which also have AVX instructions, but do not suffer from thermal extremes due to having a soldered Integrated Heat Spreader and a 35% larger Die.
"

Prime95 is supposed to give you highest-possible, worst-case temps, you nimrod, and Tomshardware threads are full of idiots, which you should know.

The whole reason you run a stress test instead of just playing games and seeing if it crashes is to produce maximum-possible loads and temps, since if your CPU is stable there, it'll be stable under any lesser load.

OP just get an NH-D15

>keep my 4670k at 4.5GHz
>never had it get to a scary point of hot ever
I'm okay with this. Could probably do 4.6 but I've never tried.

>>51704648
try occt

>>51704783
if thats the case
ftp://mersenne.org/gimps/p64v266.zip

>>51706554
Yeah I tried it, now the peak was 65C. I say that's pretty good. A lot of anons are saying "durr its supposed to be unrealistic" but there's a difference between worst scenario test and completely unrealistic scenario test.

>>51704394
My laptop's gpu get's that without any overclocking

>>51704348
You should be using v26.6, Later versions of prime will nuke your CPU. I switched to the older for my 4690k and peak temps dropped about 20c.

>>51704127
Just get a better cooler even a clc water one.

I have a noctua d14 and only hit 65c at max temp and the fans are still quiet af.

>>51704127
no, 72c is the max you want to see in summer, so no

>>51704127
Lower the voltage if you want lower temperatures. 95ish is fine for modern CPUs tho (running at stock specs).

>>51704391
Overclocking is consumer crap these days. Pay more for the ability overclock.

Is there a difference between h97 and z97 or i5 4690 and 4690k besides the ability to overclock?

In the bad old days you would buy a 300 MHz Celeron. After overclocking it would perform like a Pentium II 450 that was several times more expensive.

>>51704127
prime95 is bullshit, if you can keep it below 100ºC on the most intensive test you should be good.

I have a 4690k @ 4.5ghz, it goes up to 95ºC while running prime95 yet never goes above 80 while doing literally anything else (gaming, video encoding, crypto bruteforcing, you name it)

>>51704302
Jesus fucking Christ, what makes you think an OC THIS big is acceptable with a hyper 212 evo?
Go liquid and keep the OC or stay air and turn it down not only 1 step but 2.
(4.3 - 4.4Ghz @1.18 - 1.20V is what I would shoot for, but ofc. only if your chip manages to get this OC at these Voltages)

There is a middle way though, but its plain retarded: Buy better air cooling and get down to ~ 4.5Ghz.

Also, 3770k @ 4.5V, 1.28V (0.08 offset) here, using the H100i I max out at 65°C.

>>51704302
Try 2 fans for push/pull?

>97c
Just fuck my shit up senpai

>>51704127
noob question.
does changing ram timings affect the CPU overclocking in any way?

noob question2
is it true that the more i overclock my CPU, the more i have to lower the ratio on my rams?would increasing ram voltage allow for higher ram ratio?

>oc'ing your i7-9990 to 6ghz to browse 4chan and play games

>>51708296
Ideally for ddr3 you want 1066mhz, anything higher results in poor timings and is an overclock since 1066 is the native frequency of ddr3 chips.

>>51708491
but that does not answer my question ._.

>>51708507
You can achieve higher frequencies by increasing the timing, but just decrease the ratio so it's close to the rated frequency.
Overclocking your multiplier increases ran frequency too

>>51704351
I was genuinely too retarded to know what was happening. It's got a 212 Memevo now, running about 65°c under load