Decent Quiet cpu cooler for ryzen 7 1700 at 4ghz

[jboydgolfer]

I would hazard to guess, he would see the same temperatures he is seeing now. If not worse.

If I were you, I would stick with that D15, it's a beast ,and about as good as it's going to get

i didnt advocate he switch to mine. actually i said to stay. he mentioned the cooler i spoke of, and i said it was good for MY CPU, but that his D15 would be about as good as its gonna get.

id hazard a bet my cooler is slightly better than the D15, but SO little it wouldnt be worth a change. but thats based on my opinion, and not fact, its just my guess from the hip
now im curious,....im gonna look into it, to see if my instinct was wrong ( looks like i might have been wrong)

 

[crazyeyesreaper]

As said by everyone else.

D15 is as good as it gets on air.

If the CPU isn't throttling then its fine, Find your overclock test for stability and if it doesn't throttle the clock speed then your fine. As all these stress tests tend to use AVX and stress FPUs to extremes which results in higher temperatures. Its why in my reviews here I use AIDA64 extensively. Stressing CPU / Cache / FPU / Memory vs FPU alone is about 10'C give or take.

So in a heavy FPU load worst case scenario if your hitting 75C under an FPU load your likely only going to hit 65-70c Under a max typical load.

As others have said. Get to the Overclock you want, and test it for stability. If its stable then run a more typical stress test, Cinebench R15 continuously, Aida64 CPU / Cache / FPU / System memory etc that simulates a more realistic work load. You will see lower temps compared to Aida FPU / IBT etc etc. This is an example of not understanding modern hardware. The chance of a workload or game using FPU load continuously is extremely slim. As such its a dead end worst case this is the hottest the CPU will ever get test.

 

[sneekypeet]

i didnt advocate he switch to mine. actually i said to stay. he mentioned the cooler i spoke of, and i said it was good for MY CPU, but that his D15 would be about as good as its gonna get.

id hazard a bet my cooler is slightly better than the D15, but SO little it wouldnt be worth a change. but thats based on my opinion, and not fact, its just my guess from the hip
now im curious,....im gonna look into it, to see if my instinct was wrong ( looks like i might have been wrong)

I was pointing out the obvious flaw in the first part of your comment. You say the cooler is good, which it is on low tdp CPUs. Since he was not on such a chip, I felt it needed mentioned that your results could not be used as a stepping stone to what he would see with said cooler.

 

[jboydgolfer]

I was pointing out the obvious flaw in the first part of your comment. You say the cooler is good, which it is on low tdp CPUs. Since he was not on such a chip, I felt it needed mentioned that your results could not be used as a stepping stone to what he would see with said cooler.

I see what you were pointing out ,and I was pointing out that it was the same thing that I mentioned in that post " what's good for me, isn't necessarily good for you".

i suppose my point is the redundancy of your reply. it states what i stated already.

the 4790 doesn't generate much heat

I don't know what kind of temperatures your CPU brings compared to mine

regardless, i digress as its getting off topic.

 

[infrared]

ryzen 7 i wonder how someone can have a 4199 mhz cpu at 1.504 on air (nr 9)

Most of those cpuz validations are far from stable OC's, it's litterally a case of put voltage as high as you dare, boot with the highest frequency that'll get into windows without blue screening, open cpuz and validate. If you opened any stress test they'd likely insta-crash.

 

[sneekypeet]

I see what you were pointing out ,and I was pointing out that it was the same thing that I mentioned in that post " what's good for me, isn't necessarily good for you".

i suppose my point is the redundancy of your reply. it states what i stated already.




regardless, i digress as its getting off topic.

Because those who know are pointing him at the truth of a much larger aio being needed, if not a custom loop. Your post added a glimmer of hope which does not fit the reality of his situation. I guess my point is that outside of "keep the d15" the rest of your post has no bearing on the OPs search. I am not trying to call you out or ignore your context, I just feel that with a glimmer of hope, the OP will waste money to find out what we have all been saying was reality the whole time.

 

[the54thvoid]

Most of those cpuz validations are far from stable OC's, it's litterally a case of put voltage as high as you dare, boot with the highest frequency that'll get into windows without blue screening, open cpuz and validate. If you opened any stress test they'd likely insta-crash.

Yup. Or try gaming.... Best test.

 

[Vayra86]

I don't hate any particular pump, just those that are noise,high pitched

I don't have any magic ryzen 7 1700 or some fantastic mb where i can tweak it to something like 1.300 volt and 4ghz



I do get above 70c with ibt and that often makes the fans spin to max rpm for a few sec and then goes down.

Not shure what max temps are (throttle) mabye it's 75, theres luckily no temp offset with a ryzen 7 1700, what i see is what the temps are. Im trying with my asus gtx 1070 strix to get 4k gaming pc score both in time spy and firstrike (standard) im very close to my gpu's limited

Time spy 1.0

Fire strike 1.1

30% is where i feel my cpu cooler is very silent, not just under avarage or okay silent, 60% is just under 1000rpm

View attachment 97329

Change (case) fans. I can strongly recommend beQuiet! SilentWings 3. If you need high performance cooling at will, get the High Speed versions. They are whisper quiet at 900 RPM, and a hurricane at 2200 RPM, like seriously, I get a GPU that is 5 C cooler (!!) if I turn those up a bit. CPU, similar drop. I've got 3x SilentWings 3 HS 120mm as front intakes (and just a simple low-speed fan as outtake, bc positive case pressure = beneficial).

You already have the pinnacle of air cooling so its either better case airflow, or water. When you go water, you can skip the simple AIO's right away, go custom or get something serious at least, or the temp drop is not noticeable. Most AIO's are similar or worse than the best air.

We share quite similar power/heat to dissipate - I cool a 8700k @ 1.34v (120-140w to dissipate) on a pretty similar air cooler (BQ Dark Rock Pro 3). You should be getting solid results on this cooling, not a temperature struggle, so you saying you do struggle points towards bad case airflow/a higher ambient temp/case temp.

 

[John Naylor]

The Scyther Mugen Max ($37) and Fuma ($45) meet of beat the $90 Noctual NH-D15 and Cryorig R!. All4 beat the Corsair h100i thermally while the AIO will be orders of magnitude louder.

As you can see here, replacing the Noctua 1500 rpm fans with Phanteks 1200 rpm fans can drop temps 3C or at same 1200 rpm 6C.

http://www.overclockersclub.com/reviews/phenteks_f140/3.htm

The Phanteks MP series 1500 shud get you a bit more

I ya going to go water, AIO I can't recommend anything but the Swiftech H series and Phoenix series from EQ with all copper componentry (EK also has an all aluminum series) ... not none of the CLcs with copper blocks and aluminum rads.

 

[ClockworkOranges]

New here, found this while trolling around some overclocking advice and thought id make an account and throw my two cents in.

Im surprised no one here has mentioned thermal paste yet, what are you using OP? I was a huge fan of Arctic Silver for many years because "Gold Standard" right? Well then I stumbled upon this paper.

Basically, if you're using AS, its not worth spit and you should consider replacing it right away. They advertise it as having a thermal conductivity of 8.7W/mK, when in reality its closer to 0.94W/mK
Its only transfering approximately a 9th of the energy per mm^2 its "rated" for.

I replaced mine with some Shin-Etsu G751. You can't find the paste mentioned in the paper, but the G751 is the next best thing, tested at 4.5W/mK, its literally 5 times more conductive than AS5 all things condsidered. I replaced my CPU and GPU paste with this stuff to my extreme pleasure. Asus 970 used to sit above 40 when idle and overclocked, with the fan speed at 40%. Now it Idles around 30C, fan @40%.

I didn't notice as much as a difference with cpu and my H100IV2, but testing with that is a lot harder as the ambient temperature and coolant temperature variables make it a PITA to get true numbers.

Anyhoo, best of luck op!

 

[Relayer]

Maybe try reapplying TIM and reseating the cooler?

 

[gasolin]

noctua tim

 

[Vayra86]

New here, found this while trolling around some overclocking advice and thought id make an account and throw my two cents in.

Im surprised no one here has mentioned thermal paste yet, what are you using OP? I was a huge fan of Arctic Silver for many years because "Gold Standard" right? Well then I stumbled upon this paper.

Basically, if you're using AS, its not worth spit and you should consider replacing it right away. They advertise it as having a thermal conductivity of 8.7W/mK, when in reality its closer to 0.94W/mK
Its only transfering approximately a 9th of the energy per mm^2 its "rated" for.

I replaced mine with some Shin-Etsu G751. You can't find the paste mentioned in the paper, but the G751 is the next best thing, tested at 4.5W/mK, its literally 5 times more conductive than AS5 all things condsidered. I replaced my CPU and GPU paste with this stuff to my extreme pleasure. Asus 970 used to sit above 40 when idle and overclocked, with the fan speed at 40%. Now it Idles around 30C, fan @40%.

I didn't notice as much as a difference with cpu and my H100IV2, but testing with that is a lot harder as the ambient temperature and coolant temperature variables make it a PITA to get true numbers.

Anyhoo, best of luck op!

If you get more than 3 C between the regular OEM and best paste, its already a lot. No single thermal paste is going to make or break an OC or regular use. Only bad application of it does; such as leaving a plastic foil in between, overtightening or too loose, and yes it all happens.

 

[gasolin]

2 new fans 2c better tim 2 -3c right?

Thermal Kryonaut (-5.6c that's alot)

Do i need better fans or Scythe Fuma ? I would like to know since i might save a little if new fans or a cpu cooler could be bought from the same place as the tim,thermal paste,grease

 

[ClockworkOranges]

If you get more than 3 C between the regular OEM and best paste, its already a lot. No single thermal paste is going to make or break an OC or regular use. Only bad application of it does; such as leaving a plastic foil in between, overtightening or too loose, and yes it all happens.

That rule generally holds true, I'll give ya that. In the end, it is about the watts the cooler can dissipate. Im not saying its a magic wonder cure, but it is something to keep in mind as thermal paste has come a long way since 2010 even, let alone when that paper was published in 08. Making sure you've got a good paste and application will get you far.

Op, what are the ambient temps like where you live, and whats the size of the space your pc is in? I know after playing a few hours of kingdom come, my little computer room gets really stuffy without a window open. Thankfully I'm graced by canadian winters, with the window open the room will hit 13c on the floor sometimes, allowing me to keep my fans at near-stall speed and still have decent cooling.

If you're anywhere hot, you may be SOL.

 

[crazyeyesreaper]

Noctua TIM is already pretty decent

The best tims on the market vs worst TIM your looking at 4 C

Using pastes available today stuff like Noctua NT-H1 or Arctic MX2 or MX4 IC Diamond etc the gap shrinks to 1-2C with the Majority being 1C apart or less.

1C is not worth spending $10+ to gain.

Not to mention too much or too little paste etc will influence those results. meaning for the majority of situations you would need to be using something like Ceramique 2 and them upgrade to a top tier paste to really see a benefit.

 

[gasolin]

ambient temps about 22 c pc case define r5 3 x 140mm noctua pwm fans

$10 for 1c? It's more like 5 c going fron noctua's NT-H1 to Thermal Grizzly Kryonaut

I personally have tried delidding my i7 7700k i don't have today, i didn't noticed any change in temps, people told me that i should have used a liquid metal thermal paste instead of Gelid GC extreme.

I don't want to take a change of using a conductive thermalpaste

I remounted my notuca a few days ago since my ryzen 7 1700 is new, that gives my a chance of remounting,reseating my cooler and at the same time try better tim,thermal paste

Crazyeyereaper only to little makes a difference (or going from avarage thermal paste to one of the best)

(Just for the fun of it)

 

[kastriot]

Maybe when 2700 is out you can sell this one and get better oc-able cpu?

 

[gasolin]

Oc is not a problem it's the temps at 4ghz,it's after all a 8 core cpu (3.9ghz is high, 3.8ghz is approved)

 

[londiste]

OP, what exactly are you looking for? D15 is as good as it gets for air cooling. As has been said several times already, you can get a degree here or a dba there but nothing significant. Thermal paste an get you a degree or two unless you go for liquid metal that does have some problems of its own. Try to get it off after a while for example.

AIO does not improve thing very much on noise front. You can get lower temps, sure, but that will come at the cost of noise and even then you would need AIO with 360mm radiator to match or exceed your current D15.

If you really want lower temps at same or lower noise level, custom loop is pretty much your only realistic option. D5 and large radiator (at least 360mm) is what you want.

 

[gasolin]

lower temps so i can oc to 4ghz, i will start with the thermal paste and go from there

It's the least expensive i could try

 

[HTC]

lower temps so i can oc to 4ghz, i will start with the thermal paste and go from there

It's the least expensive i could try

I could be wrong but it seems you lucked out on the silicon lottery. If so, then even with custom water cooling you'll be hard pressed to achieve your goal.

 

[gasolin]

It's not a problem getting it to 4.0ghz, problem is temps at 4ghz and to tweak it so it's 100% stable, not just making shure temps are low.

Even if better thermal paste only gives me 2-3 c lower temps, it's still a good start even when i only can run my cpu at 3.9ghz.

I can run it at 3.9ghz hoping for 4.0ghz.

IF i had a real clc like something JayTwoCents would use in a define r5 for a ryzen 7 at 4ghz im pretty shure i could run it stable at 4ghz if it can handle more than 1.500 volt

I do have to say i would never try to run my cpu above 4ghz not even 4.025 ghz since i do need around 1.500volt for 4ghz

 

[infrared]

since i do need around 1.500volt for 4ghz

Time to give up on the idea of it being daily usable at 4.0ghz.. that voltage is too high for daily use, 1.50v is benchmarking territory. If you care about that chip at all I'd recommend that you hold back at 1.45v or less, and just adjust clocks to suit. Your temps will be noticeably better too.

 

[R00kie]

hate high pitched aio pumps

Wat? Had a Corsair H100i for like 3 years, and I have never heard anything coming from the pump, same can be said with the other AIO's that I've been installing for people - quiet as a cemetery.