Post your speccy and CPU-Z shit.
>>52869212
I would try pushing it further, but there's not much point as I'll be upgrading to Broadwell-E when it's released. Temperatures start getting out of hand at 1.3V+ anyway.
Should i try to overclock and would i even gain anything?
>>52869412
You need the CPU-Z screenshot as well, this is an overclocking thread.
>>52869429
Nothing substantial really. But you could if you wanted to.
>>52869212
>/COG/ CPU Overclocking General.
This is a really shitty idea.
>>52869513
Thanks for the bump faggot.
>>52869532
OK, kid.
>another general
Sup reddit.
Fuck off back to plebbit with your retarded "Generals"
You clearly have not understood how 4chan works if you want to make one.
Also I like how reddit is filtered now because the reddit mods are butthurt how we always tell them to fuck off.
I'm running a 2500k at 3.2 Ghz, which means it is underclocked by 100Mhz.
Why? Because that gives me a seamless ratio multiplier of 4, considering I have 1600 MHz DDR3 RAM.
If you're a gamer, you should try that too, for pure number crunching it's not as recommendable though.
Why am I doing this? This prevents tearing by a fuckton when playing vidya.
Reminder: The pentium anniversary edition is also designed like this, which is the sole reason why everybody says "even on stock it's so fast".
If you have a memory multiplier of, let's say 16 (which means 8 if it's DDR), and you have a multiplier of say 17 for the CPU, every 16 RAM cycles a pause of 1 cycle has to be put in to make up the difference. That's what happens when you use unfitting multipliers.
When you need pure CPU power (rendering ray tracing or video editing) this effect doesn't matter, however, for a nice and smooth computing experience this is a must.
Note: Benchmarks will not proof this, you need your own eyes to see the benefit.
YOUR WELCOME, FUCKERS
>>52869718
>Note: Benchmarks will not proof this
lel
next upgrade in 2018
>>52869765
Can you benchmark the difference V-Sync does?
No, you can't, you have to move the mouse yourself to see the difference and what that fuzz is about.
Can you benchmark or proof freesync, g-sync? No, you have to see it with your own eyes.
Can you really tell the differences from AA without trying them out for yourself with the game you're playing? E.g. with screenshots? No, screenshots don't show the "noise" no AA gives and videos are always compressed. You have to try it for yourself.
Same applies here, m8.
>>52869718
>>52869832
It's good to know that we have some fucking crazies posting here.
>>52869819
>have worse single thread score than a fucking X5470
JUST
Can I get mt 3570k to 4.5 with a 212 meme cooler or do I need water? Also what sort of voltage should I be aiming for?
>>52870214
That would be easy to do. You'd probably sit around 75c at full load.
>>52870331
>>52870331
How much power does this thing draw out of the power outlet?
Has anybody considered passively cooling a skylake ITT?
>>52870696
Linus Shill Tips did a video where he passively cooled a 5960x, so I think its possible.
>>52869412
If youre gunna try to lie about the fact that you have a 980Ti, atleast do the research to know that they don't come in 2GB configs, standard is 6Gb.
When will it become worth it to upgrade from my i5 2500k?
>>52871060
When Zen comes out.
>>52871005
Hello newfriend!
Unfortunately I cant OC any higher, just not stable.
using a pentium 3 at 933 mhz and 256mb ram. too lazy right now for speccy
>>52871060
When I say 'worth' it, I mean like a night/day difference.
Kinda like the difference between onboard graphics and a GTX 570, or booting from an SSD vs a 5400rpm HDD.
>>52869362
>Wow my 4790k at 4.6 beats the 6700k at stock speeds
>>52869212
Fuck I wish I was home.
i7 [email protected] ghz on my desktop.
>>52869212
Muh Dual X5690s are old, but gold.
>>52871378
@4.5
>>52871803
>i7 5820K
>8GB Single Channel
POOR
>>52871842
It sucks, have another 8gb stick in the mail. :(
>>52871259
What are you trying to say, you fucking retard? CPU-Z offers profiles to guage performance against. I chose that one because it's actually kind of interesting to see that a Haswell chip needs to run about 400MHz faster than a Skylake one for equal performance.
You sound like a Skylake owner who's incredibly insecure about his purchasing decision. Try killing yourself and see if your anxiety goes away.
>>52871870
You should have another three in the mail. Who buys X99 and doesn't take advantage of quad channel memory?
>>52871878
You swallowed that bait, deep
>>52869212
My 2500k is at 4.6Ghz, 100Mz with a 46 multiplier. I'm away from home so I can't screen shot it.
I posted in a CPU-Z benchmark thread a while ago and people commented on my multi-core performance saying it was abnormally low. I couldn't find a solution, but when I upgraded to Windows 10 my score went up to normal. Not sure what was going on there.
Does this make me one poorfag?
>>52873408
yes.
yes it does.
>>52869947
it gets worse, anon.
I've managed to get 4.6 out of it, but at 4.2 I never get above 65c under load. Stable for months now.
>>52875152
What cooler?
You should be able to get it way past 4.2 with a decent cooler.
default 3.8ghz turbo fucked to 4.2ghz
I aint havin none of that so I disabled all the cpu faggotry, I mean features and made it 4.5ghz 24/7
>>52875460
just noticed something.
fucking speccy man...
Posting
>>52871005
I bet you also think AMD CPUs truly run at 12C on air.
6600K @ 4.7Ghz
>>52875200
GeminII S524 Ver.2
4.4ghz is right around 1.3v, it ran at 4.6 but the heat very quickly shot up to around 76C under load. I'm trying to keep it under 70C load even though I've seen plenty of people mention runnin around 80C with no issues.
>>52871231
probably 3-5 more years.
>>52875522
either the guy is an idiot or has never ran speccy or both.
Speccy has false readings for a lot of hardware.
Basically it it might get the product label right, but from there on, it's a fucking trainwreck.
>>52875627
https://www.youtube.com/watch?v=CRNkK6v2PGk
>>52869718
This guy just gave everyone the secret sauce. 1:1 ratio on ram is the real deal. Try it on an old c2d machine, makes shit real fuckin snappy.
>>52871803
wtf guy
see
>>52871842
>>52870841
Link? i cant find the video whatsoever
>>52875564
>an 6600k OCed beating an 6700k stock at multi thread
wath
>>52871005
thanks friend
>>52876356
Despite what you have heard, hyperthreading is only useful in certain situations, like extremely narrow sets of situations. Faster clock speeds will beat out more threads any day.
>>52875988
Its the silent PC build. Look under the PC builds playlist.
>>52876657
so would that oced 6600k beat the 6700k at video encoding ?
>>52876677
thank you anon!
>>52876697
Not at video encoding no. That is highly multi threaded.
And you're welcome
Have OCd to 4.5 and 4.6, but it is never stable.
>>52876727
Also Speccy always reads the 980 Ti as having 2Gb of VRAM for some reason.
>>52876727
Try increasing your voltage anon.
>>52876744
I did, pumped it up to 1.35 and it still wasn't stable at 4.5. Other people were claiming stable at 1.3 so I'm unsure what to think.
>>52876769
Please think about memes and disable every feature the CPU has, except virtualization and thermal throttling
>>52871231
when the 2500k starts causing visible lag in games
I'm a poorfag and I'm running this FX 6300.
Don't have any thermal paste to apply, nor do I have money to buy an aftermarket cooler.
I used to get over 80° under load, resulting on the PC getting shut down when hitting 86 or so.
I tried undervolting it to 1.23v and now it hovers around the 70° mark under heavy load.
Is there any other way of making it run cooler, that doesn't involve buying anything?
I tried underclocking it to 3.2GHz (3.5 is stock), and it made little to no difference.
>>52869212
Here goes nothing
>>52876862
Get a 20$ cooler dude. Like even the poorest of poorfags can afford.
>>52876862
If you can't afford a $7 tube of thermal paste and a $30 Hyper 212 you need to seriously rethink your life choices
>>52876929
>>52876916
>third world
>student loans
>tech taxes
>no job
I seriously can't afford a $20-$30 luxury that might not work right now.
>>52876744
Increased voltage slightly, and OCd to 4.5. Solid gains honestly.
>>52877000
Oops, wrong i7.
>>52875062
why do you let yourself suffer. why not just end it all
>>52876991
If it doesn't work it's broken and can be returned
>>52869718
Hey what if I have a 2400 with 1333 ram instead what should my multiplier ratio be then? Pls
>>52877223
downclock the RAM to 1066MHz, and set the CPU to a fixed clock of 3.2GHz.
>>52869659
Speecy/CPU-Z/Overclock threads were very common in /g/ before reddit was even a 'thing'.
'Generals' don't predate, but certainly were around in reddits infancy. Get off your LE REDDIT horse.
Got a 6700k skylake build i'll be doing once my tax return arrives. Going to drop it on an MSI Z170A Gaming pro board.
OC isn't my main push, but if I can keep it rock stable, I may consider it.
What kind of results should I expect?
>>52878208
4.4 is easy, 4.6 can be finicky with voltage bumps etc. , 4.8 is hard.
>>52878237
May try the 4.4. Never played with one before, so I got a Corsair AIO close loop water cooler with radiator, too. Should be quieter if anything.
Still researching if it's best to have it pull into case, push through fins out of case, or pull through fins out of case for heat dispersion, though.
amd wizardry on a 50$ asus mobo, i wonder how my shitty mobo hasnt blown up after three years of abuse, seriously, i had this thing up to 4.6ghz before realizing it was cancer for the mobo.
currently saving/waiting for zen and polaris
5960X @4.5 GHz on 1.31V.
>>52876744
I had to set mine to 1.34v to be stable at 4.55Ghz. I have to set it to 1.5v to be stable at 4.75Ghz
http://valid.x86.fr/4pymlv
>>52869718
SEND HELP
>>52871223
>1.616V
Madman
>>52878124
>>52877223
The answer you got is correct, however I think that reducing RAM speed isn't really recommendable. E.g. I'd always recommend faster RAM over faster CPUs (only to a certain degree of course).
That's only my opinion and/or guess about how it would feel in front of your machine. You need to try it out. Because reducing 1333 to 1066 is like 1/4th speed loss, instead of the 1/33th speed loss I have on my CPU.
Not sure if my point gets across, explaining maths in english is pretty hard, but try it out for yourself, maybe you will notice a difference, maybe to the better, maybe it's gets slower, maybe there's no difference to be felt at all.
You could also get faster RAM so that the multiplier fits again.
But what's a 2400 anyway? Did you mean 2500? If yes, I'd highly recommend getting 1600 Mhz RAM. Also you could try oc'ing the RAM, I've made good experiences doing so (remember to test the RAM with memtest) in the past. In my experience RAM doesn't break earlier when OC'd, it just breaks when it feels like it, no matter what speed it runs on.
>>52878280
If you have an AIO, you should be set to run for 4.6-4.7 as the major concern is thermals.
As long as you're not retarded and don't kill the chip by overvolting it, you should be fine.
Also, don't use Prime95 for stress testing overvolted chips. Recent versions have been known to destroy overclocked chips even though they are stable in other stress tests.
>>52869212
What kind of clocks can I expect on an FX-6300? Have one sitting at 4.0GHz at the moment, don't want to go higher until I know what's safe.
>>52878886
4.3-4.4 on stock cooler, ~1.40-1.45V
It's not the clock frequency that's dangerous, it's the voltage that you throw at it.
Stressing 4790K at 4.8GHz with x264 stability test.
