AMD Wraith Question

killroy67

[H]ard|Gawd
Joined
Oct 16, 2006
Messages
1,572
So I have been using the AMD Wraith for awhile now with the 3900X. Typically at idle Im in the mid-30's, when gaming it stays between 55c-60c while consistently staying around 4-4.5ghz on all cores. My case is well ventilated and my ambient temps are 64f during the winter, and 68f during the summer due to central AC. So the question being how much would I benefit from a higher end cooler like a NOCTUA or one of the Be Quiet solutions, would it be a noticeable difference or am I just fine where I am.
 
Major thing I can see would be a drop in noise while maintaining better boost clocks under Prime95 all core boost situations. If your happy with noise levels don't waste money.
 
I doubt you're hovering between 4-4.5ghz all cores on a 3900x, more so with the stock air cooler. You're probably seeing single / some cores boosting to that speed but not all.
 
I doubt you're hovering between 4-4.5ghz all cores on a 3900x, more so with the stock air cooler. You're probably seeing single / some cores boosting to that speed but not all.
I’ve monitored all cores while gaming using MSI Afterburner playing Rage 2, Battlefront 2 or Borderlands 3 and it stays around 4ghz while boosting to 4500ghz frequently. I’ve also benched marked it with cinebench r20 and score between 7160-7200.
 
I’ve monitored all cores while gaming using MSI Afterburner playing Rage 2, Battlefront 2 or Borderlands 3 and it stays around 4ghz while boosting to 4500ghz frequently. I’ve also benched marked it with cinebench r20 and score between 7160-7200.

Stock boosting to 4.5 is usually on 1 or 2 cores. Not all 12.
 
Stock boosting to 4.5 is usually on 1 or 2 cores. Not all 12.
I guess I could take some pictures or make a video but I can only say what is being reported back by afterburner, I’m not making this up. I think my case ventilation and low ambient temps are what helps. But the question is if I would benefit going to another brand of air cooler or would the result for gaming be negligible. I don’t do Prime 95 type situations.
 
I guess I could take some pictures or make a video but I can only say what is being reported back by afterburner, I’m not making this up. I think my case ventilation and low ambient temps are what helps. But the question is if I would benefit going to another brand of air cooler or would the result for gaming be negligible. I don’t do Prime 95 type situations.
if it's boosting to some or most cores to that speed it due to new game engines being multi threaded. As other mentioned if you were to use prime95 or other tools that fully load cores you would barely hold over stock core speeds. To me it comes down to noise. If your happy with that then don't upgrade. If not grab a new heatsink. You might even boost a bit more consistently towards max boost across more cores for longer duration during gaming.
 
if it's boosting to some or most cores to that speed it due to new game engines being multi threaded. As other mentioned if you were to use prime95 or other tools that fully load cores you would barely hold over stock core speeds. To me it comes down to noise. If your happy with that then don't upgrade. If not grab a new heatsink. You might even boost a bit more consistently towards max boost across more cores for longer duration during gaming.
So what would be a good air cooler that is easy to mount using the existing back plate. Ive looked at the Noctua and its just massive, also considered one of the Be Quiet models.
 
Last edited:
So what would be a good air cooler that is easy to mount using the existing back plate. Ive looked at the Noctua and its just massive, also considered one of the Be Quiet models.
I thought I saw that noctua used the am4 backplate. Will check tonight
 
So what would be a good air cooler that is easy to mount using the existing back plate. Ive looked at the Noctua and its just massive, also considered one of the Be Quiet models.
Look at the scythe Mugen rev b. It uses stock am4 mounts, bit smaller than other towers and comes with a nice quiet fan
 
Look at the scythe Mugen rev b. It uses stock am4 mounts, bit smaller than other towers and comes with a nice quiet fan
first cooler i ever seen with a convex base instead of flat. aside form that it is very good cooler...i had to use a graphite pad to get good contact because of this paste was less that ideal. honestly your temps seems very good even compared to my cooler...if i had to do it over i would go with the best aio water i could fit. :) can always use the stock cooler as emergency backup in case of issues with pump
 
You have a magical 3900x if you're frequently boosting to 4.5GHz on all cores with a wraith cooler getting those temps. Magical might be an understatement in fact. I wouldn't change a thing.
 
You have a magical 3900x if you're frequently boosting to 4.5GHz on all cores with a wraith cooler getting those temps. Magical might be an understatement in fact. I wouldn't change a thing.
I decided to monitor the CPU again since I haven't done so for a while, plus it been a few bios updates so I was curious to the performance. With Borderlands 3 typically 4250-4300ghz all cores, boosting to 4450ghz occasionally while staying at around 55-60c, with Rage 2 also about 4250-4300ghz and also boosting to 4450ghz and temps are 55-60c.
 
I found some stuff out tonight, by putting my CPU VDDP voltage to Normal, and enabling PBO under AMD CBS the 3900X hovers around 4325ghz while boosting to 4475-4600ghz while playing BL3. Temps are staying around 55-60c with Ambient temps of 64F.
 
You have a magical 3900x if you're frequently boosting to 4.5GHz on all cores with a wraith cooler getting those temps. Magical might be an understatement in fact. I wouldn't change a thing.
i dont think some people have the same idea of all core clocks is my guess?
 
After testing tonight all the cores do hover around 4325ghz, some (4-5 cores) boost between 4475-4600ghz.

My previous testing was done when I got the 3900X and a number of bios updates back when I was seeing 4500ghz across all the cores, this may of been bad information from Afterburner or a combination of things.
 
Are you now using HWINFO 64 for monitoring? I think that's pretty much the standard.
 
Are you measuring max core while in game or during cut scenes, etc. Does max core last longer than a ms?
 
Are you measuring max core while in game or during cut scenes, etc. Does max core last longer than a ms?
While in game on Borderlands 3. As I said all 12 cores are typically running at a consistent 4325ghz, 4 to 5 of the top cores will boost between 4475-4600ghz. Typically its more consistently boosting to 4450-4475ghz and will stay there for a couple seconds.
 
Last edited:
While in game on Borderlands 3. As I said all 12 cores are typically running at a consistent 4325ghz, 4 to 5 of the top cores will boost between 4475-4600ghz. Typically its more consistently boosting to 4475ghz and will stay there for a couple seconds.

the 4-5 cores that are boosting to 4.4-4.6Ghz are likely the cores the games using while the others are parked or idle. most applications(this includes hwinfo64) can't properly detect the core clock speeds on idle cores for ryzen so it'll just display the last reported clock speed.
 
the 4-5 cores that are boosting to 4.4-4.6Ghz are likely the cores the games using while the others are parked or idle. most applications(this includes hwinfo64) can't properly detect the core clock speeds on idle cores for ryzen so it'll just display the last reported clock speed.
The bottom half of the cores seem active according to Afterburner and HWinfo, they are fluctuating up and down, hovering around 4300-4325ghz and sometimes dropping below 4ghz......every now and than one might boost to 4500ghz but not often. Ill try testing it with Ryzen master later on maybe.
 
Screenshots would be nice. I’m a 3900x owner myself and do not see that kind of performance happening on a wraith cooler. But you may just have the one golden chip.
 
 
Last edited:
in my mind an all core clock would be when maxing all the cores at once like with handbrake or prime95 another program that uses all the cores to max at he same time...mine will do just under 4ghz all cores but will only boost 2 or 4 cores to about 4.4
I wouldn't change a thing if you're happy. I would like to see you set Cinebench R20 to run 300 seconds and see where your chip boosts after the first minute with HWINFO64.
way less than that lol;)
 
Cinebench R20 at 300sec first half of the test all cores stayed around 4025ghz at times some cores dropping to 3975ghz, later in the run it was at 4000ghz also dropping to 3975 on any of the cores.
 
Cinebench R20 at 300sec first half of the test all cores stayed around 4025ghz at times some cores dropping to 3975ghz, later in the run it was at 4000ghz also dropping to 3975 on any of the cores.
there is your all core speed:) Pretty much same as my 2700x (clock wise)
 
Yes but I dont pay much attention to Cinebench or Prime 95, none of it translates to games, no game that I know of will max out all 12 cores 100%.
what max temp does it reach with 5 min of prime95 max heat? mine hits around 79-80 at max temp with most extreme usage case. Im just curious what that stock cooler gets to? If i had to guess it be high 80s or worse
 
I can run it later on, what kind of test blended test or something else?
 
I ran the blended test and all the cores stayed around 4225GHZ and heat was between 60 to 68F at its highest. All the cores were maxed out at 100%, same as the small FFT test.
 
Well strange, I ran the tourture test with small FFT's and each core was just under 3800GHZ, the heat never even broke 66F. Is this because I have PBO enabled in the bios?
certainly looks like its throttling the clocks to keep temps and power down right?
 
I put PBO back to Auto and same result, all the cores throttle back on Prime 95 Small FFT's.
 
Alrighty a little bit better result....I ran the same Small FFT's but also checked round off checking. All cores held at 3850-3875GHZ and the highest my temps went is 71c after five minitues.
 
Alrighty a little bit better result....I ran the same Small FFT's but also checked round off checking. All cores held at 3850-3875GHZ and the highest my temps went is 71c after five minitues.

For how long did you run each of those Prime 95 tests? It takes a while before Prime 95 starts actually taxing the cpu. It sounds to me like you ran it 5 minutes at a time. Let it go 15 minutes and see what happens.
 
Back
Top