You have to risk permanent damage to an intel processor else it will run hot as shit

>You have to risk permanent damage to an intel processor else it will run hot as shit
Man fuck that, I was gonna buy an i7 8700K but this is fucking ridiculous. Realistically, how far away is zen 2?

Attached: TEMPY.png (800x600, 21K)

Other urls found in this thread:

mersenneforum.org/showthread.php?t=21462
computerbase.de/2018-10/core-i9-9900k-hohe-temperatur/
youtube.com/watch?v=yA0oo12rbiM
linustechtips.com/main/topic/850358-amd-extended-frequency-range-xfr-explained/
twitter.com/SFWRedditImages

Just wait until your CPU is really old and buy whatever is the newest when you are ready to upgrade.

Attached: x54cky6bjll01.jpg (975x590, 31K)

celcius

>over 80 CELCIUS
Intel is probably going to start manufacturing miniature stoves in the future.

I love how intelfags love to shill theur i7s but forget this inconvinient fact. Even the most high end water cooler won't get you 5 GHz stable if you don't delid.

>aida64

why is it that techtubers use that shit? are they paid by aida devs to run that shit?

My processor runs 15C hotter with prime95 small FFT compared to Aida bullshit (+ aida shit program costs money and prime95 is free and open source)

That 500mhz jump with a slight increase in volts still only 5c higher. Liquid metal scales with heat better than solder.

Tj max for 7700k is 100c before it starts throttling

Silicon starts to degrade before that. It's never a good idea to let your cpu stay above 80C.

t. ex i5-7400 owner that killed it in a year by letting it go past 90C on the stock cooler

I'd bet on Q2 of 2019. Epyc are expected to arrive earlier because of bigger margins and because AMD is likely to make Epycs from custom chips to increase yields.

after a certain version prime95 became unreliable for intel cpu benchmarks due to avx. you have to run a command or use an outdated version to get accurate results so people would rather use aida64

mersenneforum.org/showthread.php?t=21462

Also, P95 does not magically generate heat out of nothing, it generates it by usable computations so your point is void.

>*NEW* AVX doesn't matter!

Attached: i74zmsic4ziggrlkv9c5.gif (800x450, 1.87M)

calcium

>how far away is zen 2?
6 months

>prime 95 benchmarks
Nigger thats for stability testing

Attached: 1458575698021.jpg (250x182, 6K)

so you don't like knowing your cpus absolute max temperature?

That cpu is literally designed to run safely up to 100c and you killed it in a year idk man.
And no there's no 80c limit. It depends on the cpu/gpu

If your CPU isn't stable under AVX workloads then it's not stable, you dunce. The trend of Incels running back to old non-AVX versions of Prime95 because the newer ones cause their housefires to ignite is hilarious. Literally can't use all the features of your CPU without it curling up into a ball and dying.

No, it isn't. It's just the maximum tolerable temp for the silicon before it starts to suffer permanent damage in a very short time frame. You still should not exceed 80C as it will still cause degradation over time. That's the whole point of deliding intel CPUs, to extend their lifespan.

please delete this antisemitic post, AVX doesn't matter

You got to be kidding?!?

If this was true then why don't companies just run server CPUs at 100C 24/7 you stupid fuck?

not true*

Some of them probably do, others buy pre-built servers from OEMs which comply to Intel specs while designing them.

I work with servers, you braindead retard. Holy shit.

Attached: bd8.png (1190x906, 178K)

>if I run it below 80C then my CPU will last forever
brainlets in this thread lamo
bye fags

Attached: bibi.png (460x401, 425K)

>Call 9900K the best gaming CPU ever
>Release it two months before xmas
Hmm

filtered

More like the best Gaymen cpu rite XD

Can't wait for all the 720p benchmarks using an OC'd 2080 ti with half of AMDs cores turned off.

Large FFT is even better. You need to stress it as much as possible.

Anyone running tests without AVX shouldn't be taken seriously.

This.

How the fuck did you go past 90C with that CPU? The cooler's fan must died a year before that or something.

Just buy the new CPU, goy.

computerbase.de/2018-10/core-i9-9900k-hohe-temperatur/

Attached: 1-1080.3846368904.jpg (1581x960, 268K)

You have no idea how bad intel stock coolers are. There are mountain tall piles full of them in landfills.

Attached: aHR0cDovL21lZGlhLmJlc3RvZm1pY3JvLmNvbS82L0UvNzc5MTI2L29yaWdpbmFsL1RIUk9UVExFMS5wbmc=(1).jpg (709x457, 103K)

Retard using the stock cooler on a 6 core 12 threads intel running much higher frequency. I use a locked i5 myself and the temp never gets above 65C.

i5's suck

what does CPU stability even mean?
Your OC can be stable for literally any realistic usage, but crashes in prime95. In that case, would you consider it stable or not?

I would consider it stable, since there is no realistic scenario where the CPU would be in as much load as it is in during a prime95 strese test.

Not as much as Ryzen;)

Attached: 1538419587419.png (500x970, 49K)

> would you consider it stable or not?
Of course not.
> there is no realistic scenario where the CPU would be in as much load as it is in during a prime95 strese test.
That's the point, if it hangs or stops computing in Prime95, there's a chance it will hang or crash somewhere else.

Are they using 3200MHz RAM for that because 2600 curb stomps the i5 8400 with it.

Attached: 1534326485713.png (1824x1026, 431K)

holy shit man

Attached: 1537851546430.jpg (480x580, 26K)

But most personal computing is not so important that a crash will cause significant damage.

If you're doing something and you get a BSOD with a stop code that points to CPU instability, then fix it. But if you never encounter a crash when doing even the most heavy-workload shit you normally do (3d rendering / editing etc) then I personally wouldnt worry about it. But I'm not a competitive overclocker or anything. So I can understand other perspectives on it

7% faster for 108% the price:)

Attached: 1539714671002.jpg (2620x3416, 1.74M)

>Shows a gaming benchmark
Run along child your board is /v/

Attached: 1GVSZDDC_Fs.jpg (259x360, 30K)

It's not "unreliable", it just stresses the CPU to the worst case scenario. But Intel users are generally paste eaters who want to clock their system beyond stability and thermal capacity and then pretend it's fine because in the real world you probably won't run into many cases where your CPU is loaded to 100% with AVX workload.

If you want AVX stable and thermally viable, you can't realistically clock your Intel CPU to 5GHz or above, which is the minimum you need to justify buying the CPU in the first place.

>108% the price

Yeah man it's not like you can OC to at least 4 GHz on the stock cooler and use a $60 B450 motherboard.

Attached: 2018-05-26-image-2 (1).png (1336x1460, 25K)

>using the stock cooler
the stock cooler would have been nothing
amazingly, modern CPUs still manage to run with literally no cooler youtube.com/watch?v=yA0oo12rbiM

Yeah show me that benchmark instead!

That's not "modern" feature, Pentium 4s could run without heatsink all day. Only problem is that your performance will be utter garbage and the CPU will destroy itself over a relatively short period of time. Like you have a month or couple at most.

No they don't. Servers can have massive cooling systems that runs through the whole room with an exhaustion on the top of the building.

Ryzen throttle itself to keep the temp limit like a GPU retard.

>"ree only use intel approved testing!"
COPE

Attached: chrome_20181009_102225.png (1366x768, 405K)

lol

This needs to be added to the list.

As others have said, any overclock not stable under AVX stress tests (newer versions of IBT, prime95, linpackxtreme etc etc) isn't stable at all and an unstable cpu is a liability.

>240mm radiator CPU cooler
>3600MHz RAM are b-die
>having to buy a X470 board
3000MHz can do just fine, B450 boards can do just fine 4.0+GHz overclocks and Stock cooler is excelent.

linustechtips.com/main/topic/850358-amd-extended-frequency-range-xfr-explained/
AMD unboxed is such a hypocrite when he called out PT benchmark when he manipulated the fuck out of his bench to favor amd.

What did he do to manipulate the result?

See

Why is 80C that terrible? I've had laptops that ran way hotter than that and the CPU didn't die. Either the machine lasted until it was so pathetically slow as to be unusable, or something else died.

I agree with most of your points but that stock cooler is fucking shit especially the 2600 its barely better than the intel equivalent.

Since this is an intel bashing thread I'd really like to know why intel gets curb stomped in so many games right now.

Attached: Ryzen-Deus-Ex-Mankind-Divided-1080p-Ultra-Preset-MSAA-Off-1-800x600.png (800x600, 27K)

Wait I don't get how this pertains to the benchmarks. The image is very poorly made. What are the full builds? I don't give a shit about price, the full parts lists for test benches. Do you have a link to that?
So far nothing looks wrong except 1 potential thing, but the image is so poorly composed and missing the relevant information to confirm that issue.

Maybe games are starting to take advantage of multithreading more now. AMD's strong suit is many threads. Intel's is single thread performance.

cool

Attached: bclk2.png (500x616, 80K)

You know where that bench comes from you sly shill..

I'm guessing hardware unboxed, but I don't watch that channel so I don't know what his benchmark rig is you retard. You sound like an Intel shill because you're not willing to give me simple information that probably takes you less time to type than it did for me to type this post.

The issue I'm guessing is that he used different cooler between Intel and AMD invalidating his benchmark result.

>GPU bottleneck

I don't think 80-86c is going to cause permanent damage. I have been running my old ass intel 2500k on 82-84c for over 7 years. It's going strong.
Too much power is what will kill your cpu, if your chip is bad don't overclock into danger zone even if you can cool it.

Attached: IMG_2336.jpg (600x600, 75K)

temps on my 6700k are fine desu

Attached: 159019283799.png (1920x1080, 882K)

I doubt you have 28C at full tilt even with a custom loop. You need phase change for that.

Whats your overclock?

the hottest my cpu ever got was in pubg when it was unoptimized and it hit 67 degrees. i'm using a scythe kotetsu

you can't be serious

That's actually alarmingly high for just a game if that's legitimately the highest it has ever gone. Small FFT AVX in Prime 95 would be problematic I would imagine.

you shouldn't be happy with an unstable system.. I don't know what else to tell you

it's not

Of course you wouldn't consider that stable, retard. If there's any potential for your CPU to crash because of its overclock, then it's by definition an unstable overclock. Now, maybe it DOES make it through your gay men sessions without shitting the bed, and maybe you don't care about the odd crash here and there because you do literally nothing important on your machine anyway. But it doesn't change the fact that you're running an unstable overclock if it doesn't pass at least a few hours of Prime95 and IBT.

this, except I let that shit run over night

It should release on March.

My birthday is in that month.

Attached: 53f.jpg (196x250, 7K)

>Since this is an intel bashing thread
this is an intel bashing board, if you have problems with that take it up with president gamergate.

>67 degrees is high
sorry lad i'm not throwing away a perfectly good cpu to buy amd

Easy bake ovens with cpus

Where did I tell you to buy AMD, brainlet? I'm telling you that's a fucking high temp for just running a game. My 6800K reaches that temp doing small FFT AVX loops in Prime 95.

Why isn't direct die more common? It always baffled me going to such lengths to delid and change the paste just to then reapply the IHS on it. I get that IHS are made out of conductive materials like copper, but no conductive material at all will always be better than some to begin with. It's also way less of a pain to reapply your paste 6-12 months down the line. I've seen people selling stamped aluminium spacers with the exact thickness and holes for the surface mounted components.
What's the catch? Does liquid metal rape the aluminium in heatsinks? I could find some full copper ones, I guess...

Attached: 23.png (725x571, 70K)

You can't direct die cool Intel CPUs. The IHS is part of the mounting mechanism.

I think the spacer took care of that...

>highly unoptimized game
why do you think running a game isn't cpu intensive?

Because it's not, comparatively speaking. Specially if it's an unoptimized game. You need very good optimization to get extremely high CPU utilization, because you need to take advantage of the execution resources that exist on all cores.

look up most pubg benchmark videos around may to august of last year they all hit around 55-60 degrees in game. they also almost all use aios when you ask them.

i'm basing this on the max i would hit in game, i didn't monitor temps the entire time while playing. if it's not over 80 degrees, i don't care.

Maybe the CPU is just shit, then. That's way higher than I'm used to.

fucking 90 degrees. Your mins show that your ambients probably arent bad but either your mount job is shit, you got intel timmed, or you have zero case flow. intel must have to jack that vcore up like crazy to keep that shit stable. fucking lol.

ranges from 4790ks to ryzen 1700s.
temps are better now but it's pretty common for cpus to hit those temps.

which cooler are you using?

>when my i7 950 has been working for three years on a Hyper 212 with no fan
>and the temps are cooler than that

Attached: 1537651201.jpg (593x584, 69K)

It's really bad but not a record breaker.
The Geforce 480 runs "normally" at 105c.