Solved

Almost the worst i5 4670k chip in the world!

My 4670k need 1.21v to push 4.2ghz
and it has really high temperature!
86c under Prime95 stress test !!
Im using Corsair A50 Air cooler~~~
Did i get a really really bad chip?

i think this chip need 1.3v+ to reach 4.4ghz
and temperature will be hitting 100c++
26 answers Last reply Best Answer
More about worst 4670k chip world
  1. is there any better Air/water cooler recommended?
    Noctua NH D14 or Cooler Master V8GTS? never tried water cooler before, dun really want to try it, but if its a must, then i have to do it
    does it give a obvious temperture lower down? like 10-20c lower than what im currently having?
    i really hope it wont just giving a slightly imporvement after i have spent 80usd for the new cooler...
    is it possible to hit 4.4ghz 1.3v and below 80c after i got the new cooling system?

    thank you very much guys!
  2. For air cooling I would only look at the Noctua D-14 or D-15, or the Phanteks TC-14PE. AIO liquid coolers then look at the H105, Glacier 240L, Kraken X60.

    Haswell really does run hot.
  3. Best answer
    The Noctua is agood one. Most recommended here is the Coolermaster Hyper 212 EVO.

    1.2V should be OK according to http://www.tomshardware.com/answers/id-1834433/4670k-5ghz-good-voltage.html

    Are those core or case temps?

    http://www.tomshardware.com/answers/id-1952291/4670k-safe-overclock.html says 80C max and maybe to 1.30V max. Your turn to Google for more info.
  4. i7Baby said:
    The Noctua is agood one. Most recommended here is the Coolermaster Hyper 212 EVO.

    1.2V should be OK according to http://www.tomshardware.com/answers/id-1834433/4670k-5ghz-good-voltage.html

    Are those core or case temps?

    http://www.tomshardware.com/answers/id-1952291/4670k-safe-overclock.html says 80C max and maybe to 1.30V max. Your turn to Google for more info.



    Just because 1.2v was fine for that chip, doesnt mean its enough for the OPs chip. I would say the 212 evo wont be enough if he wants to have good temps at 4.4Ghz.
  5. NoNo Hyper 212 EVO is same level (cheap)air cooling as my Corsair A50/A70
    i'll pick Noctua NH D14
    but ND H15 is releasing very soon, should i wait for it? but it will cost over $100 usd
    omg its almost the half price of my 4670k, do u guys think its worth it?
    if D15 can extremly lower down my cpy temp and make it goes up to 4.6ghz , then i think its worth it
    but i doubt it
  6. Look at the Phanteks TC-14PE also, you might be able to get it for a better price, and I think its better than the D14.
  7. What is your mobo?
  8. Intel i5 4670k
    MSI Z87 G45 Gaming
    MSI R9 280X
    12gb 1600mhz ram
    SSD Kingston HyperX 120g
    1T HDD
    Corsair 750tx
  9. i just burnt my i5 2500k and Asus P8P67 motherboard on last sunday.... 4.5ghz 1.30v
    i had no clue how could Burnt TOGETHER?! LOL
    so i just bought my 4670k, Z87 G45 gaming motherboard and R9 280X few days ago
    i love my 2500k, its a wonderful cpu, and Fxck this 4670k, really high temperature, and i got a bad chip which required mucher higher voltage than the others do push the Ghz up
  10. what is your vrm temperature?
  11. Yes, the 2500k runs nice and cool and easily overclocks. That is why I chose it over the 3570k.

    Your setup seems good. I don't have experienece with overclocking the 4670k but I'd look out there for a safe operating voltage and find out what temps people are getting. My initial reaction is that the chip you got just isn't as easy to overclock as SandyBridge.
  12. Overclocking depends on much more than just the cpu. The motherboard and power supply are also very important factors as well as case cooling efficiency.


    Fact: The 4670k @ 4.2ghz would equal a 2500k @ 5.0ghz


    Getting high ghz numbers is meaningless or else AMD would be ruling the world right now.
  13. CTurbo said:
    Overclocking depends on much more than just the cpu. The motherboard and power supply are also very important factors as well as case cooling efficiency.

    Fact: The 4670k @ 4.2ghz would equal a 2500k @ 5.0ghz

    Getting high ghz numbers is meaningless or else AMD would be ruling the world right now.


    Fact: LOL

    2500k @5GHz vs 4670k @4.6Ghz.
    http://www.bit-tech.net/hardware/2013/06/12/intel-core-i5-4670k-haswell-cpu-review/4
    More like 4670k @4.5Ghz = 2500k @ 5GHz.
  14. CTurbo said:



    Getting high ghz numbers is meaningless or else AMD would be ruling the world right now.


    Exactly right.
  15. CTurbo said:
    Overclocking depends on much more than just the cpu. The motherboard and power supply are also very important factors as well as case cooling efficiency.


    Fact: The 4670k @ 4.2ghz would equal a 2500k @ 5.0ghz


    Getting high ghz numbers is meaningless or else AMD would be ruling the world right now.



    I'm pretty sure thats a bit exagerrated. When comparing Ivy-Sandy you only need like a 200 mhz jump to completely close the 12% or so performance gap. Haswell is about another 5-15% over Ivy. So my guess is that a 4670k @ 4.2 ghz would = a 2500k @ 4.6 ghz. Thats if were talking single threaded performance here
  16. CTurbo said:
    Overclocking depends on much more than just the cpu. The motherboard and power supply are also very important factors as well as case cooling efficiency.


    Fact: The 4670k @ 4.2ghz would equal a 2500k @ 5.0ghz


    that is not true... 4670k 4.2gzh is equal to 2500k 4.4 or 4.5k ( approximately)
    you know how i feel? i burnt the 2500k and spend some $ to get a new chip which have same perforamnce as 2500k, that feeling make me sick, and i got my 2500k in 2011! its 3 yrs after, im buying the same performance chip! omg!
  17. I just spent 80ish USD to bought a Noctua NH D-14
    temperture dropped 10c immedately
    but its still not enough to make me go higher Ghz
    Under stress test , temp from max 82c (Corsair A50) dropped to 70c (Noctua NH D14)
    if i wanna go for 4.4ghz i must increase the voltage from 1.21v to 1.28 or even 1.30v!!
    then,, the temperature will hit 90-100c again!!!
    helpppppp~~~~
    A Step-by-Step Guide to Building a PC
    Step-by-Step Guide to Building a PC http://www.tomshardware.com/forum/274745-31-step-step-guide-building was Last Updated on February 12, 2012This thread is to provide a step-by-step guide to how to build a PC. These are the general steps required in... Read More
  18. you just have a less than average chip. You could buy the best self contained water cooler and I still bet that you couldn't get it to 4.4 or higher while staying under 80c just because you have a weak chip. The only way that I could see your processor staying cool enough while having a less than average overclocking chip is if you spend like $300+ on a custom water cooling kit, and at that point it wouldn't be worth it.
  19. is it risky if i follow the overclocking guide from this guy? http://www.tomshardware.com/forum/id-1722630/intel-god-quick-dirty-guide-4ghz-haswell.html
    if i keep testing oc with different settings, does it hurt my cpu/motherboard? * i wont set the voltage more than 1.30v
    i have never touched anything else except the multipler 44x 42x etc and the fixed cpu voltage...
    so im kinda worrying if i follow the guilde from that dude, (he changed something else beside cpu voltage and mulitipler)
    as long as i keep monitoring the prime test temp wont let it pass over 84c, is everything else will be ok?
  20. Quote:
    I'm pretty sure thats a bit exagerrated. When comparing Ivy-Sandy you only need like a 200 mhz jump to completely close the 12% or so performance gap. Haswell is about another 5-15% over Ivy. So my guess is that a 4670k @ 4.2 ghz would = a 2500k @ 4.6 ghz. Thats if were talking single threaded performance here


    If we assume that Ivy is roughly 10% faster than Sandy, and Haswell is another 10% faster than Ivy, then it's safe to assume that Haswell is roughly 20% faster than Sandy.


    10% of 4.0ghz = 400mhz
    20% of 4.0ghz = 800mhz


    I have seen in person a 4670k @ 4.3ghz consistently out perform a 3570k @ 4.6ghz and then again when both chips were bumped up 100mhz more. The two chips were about even when the 4670k was @ 4.343 and the 3570k was at a not completely stable 4.7ghz.
  21. Anything over 77 and you really will start running into problems
    for the 4670k
    Tcase (CPU Temp) = 72C
    Tjunction (Core Temp) = 77C
  22. It appears you have a less than average chip. If I were you I would buy some coolaboratories liquid pro (don't buy ultra) and delid your chip. You'll see probably a 10-20 degree difference. It still wont help with the high voltage you need though for the clock though, you just have bad silicone lottery luck!
  23. Possibly, but i would look at your case? Make sure you have clean cable management and hopefully at least 2 120's blowing in with a 120 blowing out. Maybe replace the case fans with more powerful ones? I like Cougar's. Noctua's are not bad just really expensive for the price (2 cougs for 1 noc in price). Gelid's are also nice if you want the led, but are also expensive.

    Use cable ties to bind all the unused cables together, out of the way but allowing a clear airflow over your cpu cooler. Might need to reapply your tim. I like the Phoyba HeGrease extreme but do you're own research, just please not AS5. It was good and is still decent, but is old and there are better ones out there now.

    Your gpu is gonna circulate the warm air inside your case as well making the cpu cooler work harder to cool the chip.

    As for the noctua cpu cooler, it's an awesome cooler and either one would do just fine, but it freaking huge. I can type this but until you see it, you really can't comprehand how huge it really is.

    I notice you have 12gb of ram (guessing 4gb x 2 and 2gb x 2). If you have all 4 slots loaded, removing 2 will give you slightly better performance and cool your cpu down, but like... maybe .0001 c.

    That's pretty much all i can think of right now, well there is this one other thing but i wouldn't.

    If you're really desperate and know what you're doing, you can always de-lid your cpu heatsink, and apply a decent tim there. But yeah, voids your warranty and i wouldn't do that. Oh, you can also look into something called lapping, but those last 2 suggestions are extreme measures.
  24. CTurbo said:
    Overclocking depends on much more than just the cpu. The motherboard and power supply are also very important factors as well as case cooling efficiency.


    Fact: The 4670k @ 4.2ghz would equal a 2500k @ 5.0ghz


    Getting high ghz numbers is meaningless or else AMD would be ruling the world right now.



    Really? Where did you get that info? Higher frequency will give your system fast and responsive in executing programs.

    "Fact: The 4670k @ 4.2ghz would equal a 2500k @ 5.0ghz" FALSE WHERE DID YOU GET THIS INFO AGAIN? GIVE ME SOME DATA TO JUSTIFY YOUR STATEMENT.
  25. Quote:
    Really? Where did you get that info? Higher frequency will give your system fast and responsive in executing programs.


    This is only true when comparing cpus within the same family and same generation. Geez I don't know why this is so hard to understand.
    You canNOT compare directly the clock speeds of two unrelated cpus. Do I really need to get into how much faster Intel Haswell's 3.6ghz is than AMD's 5.0ghz? Numbers mean nothing. The 4670k is TWO generations newer than the 2500k. If they were both clocked at 4.5ghz, the 4670k would be roughly ~20% faster.
  26. CTurbo said:
    Quote:
    Really? Where did you get that info? Higher frequency will give your system fast and responsive in executing programs.


    This is only true when comparing cpus within the same family and same generation. Geez I don't know why this is so hard to understand.
    You canNOT compare directly the clock speeds of two unrelated cpus. Do I really need to get into how much faster Intel Haswell's 3.6ghz is than AMD's 5.0ghz? Numbers mean nothing. The 4670k is TWO generations newer than the 2500k. If they were both clocked at 4.5ghz, the 4670k would be roughly ~20% faster.


    Roughly... The bittech benchmark that was posted already proved that your estimates in speed difference was not correct. That shows more like 4670k @4.6Ghz = 2500k @ 5.0Ghz.
Ask a new question

Read More

Chip Stress Test Temperature Intel i5 CPUs