New build, Ryzen 3600 getting 95.5c max temps

wha he said^^
DWD1961 said:
Yeah, we'll juts agree to disagree on the software polling the CPU and the CPU thinking it needs to go to boost to answer. I've already stated my argument against that type of CPU firmware behavior. It's too aggressive. The dude from AMD even said any tasks will put the 3 series into boost mode, which is not good for power usage over time. That's why he said to use the Windows powermode, not the AMD power mode-for now. And I'll bet they relax their power mode algos because they don't need full boost to start a service that takes 5ms to start. That's a ridiculous.

The OP said he was getting 95C while gaming. That's a 'normal' type of usage for a PC. Although he did say his case is not airflow optimized. So there is that.
Click to expand...
 
DWD1961 said:
The OP said he was getting 95C while gaming. That's a 'normal' type of usage for a PC. Although he did say his case is not airflow optimized. So there is that.
Click to expand...

Actually, all the OP really demonstrated is that at some point during his playing the CPU hit 95C and 1.475V. Probably initially under boost before the fan had time to spin up faster. One column over is the average and that shows 79C average and 1.35V average. This is more indicative of what you'd expect with that HSF and that kind of a load. Personally, I wouldn't worry about it if I were the OP, or I'd clean and reseat the HSF since he added some extra.

Like I said before, I notice about 70C temps under load with a 360mm AIO and 1.35V with my 3600x. I probably average in the low 60's during gaming. This is with both the Ryzen balanced and the regular Windows balanced plans. The Windows balanced plan seems to be a little cooler for my chip at idle.
 
tom_ozahoski said:
Take the plastic off the heat-sink bottom? Thermal paste? Fan plugged in? Definately too hot for your setup.
Click to expand...

Not gonna lie... I did this with a Scythe Ninja 5 the other day. The best part was, I still got reasonable temps during the stress test. That's how good of a cooler that is. Temps dropped by at least 15C when I realized my mistake.
 
NobleX13 said:
Not gonna lie... I did this with a Scythe Ninja 5 the other day. The best part was, I still got reasonable temps during the stress test. That's how good of a cooler that is. Temps dropped by at least 15C when I realized my mistake.
Click to expand...

Did you install that in your LAN Gaming rig with the 1700? That looks to be as big as my Noctua NH-D15.
 
ManofGod said:
Did you install that in your LAN Gaming rig with the 1700? That looks to be as big as my Noctua NH-D15.
Click to expand...

Nope. I have a little 92mm chinese cooler for that system. No way would that monster fit in the Cougar QBX.
 
Well........I did the -.1 offset/undervolt or whatever you want to call it a while ago at the advise given here. Temps went down to about 85 with stock cooler.

Got a Scythe Big Shuriken 3 and now maxing out about 78.3 ish (hasn't reached 79 yet)

I did also realize I may not have the best fan setup as I moved a couple fans around and saw it get up to 85 again with the Scythe so I moved them back to where I had them. Funny thing here is that my fan placement is just as effective as a $50 cooler which is lame.

I was going to return the cooler but I am ok with 78, I do want to keep a low profile down blower cooler and I am not looking to sink more money into this for an AIO or something like that.


If anyone cares my case is the Enthoo pro and my fans are all Phanteks white LED fans (got this over 5 years ago I think).

2x140 front intake
2x140 top exhaust
2x120 bottom intake
1x140 back exhaust

The case has room for a 3rd top fan but I only have 2. They are in the 2 front slots meaning not on top of CPU. Oddly enough when I moved them over on top of cpu thinking this would give better flow (thinking the air is exhausting out before the intake air even reached the cpu) that is when the temps went up. Doesn't make sense to me but it is what it is.
 
kirbyrj said:
Actually, all the OP really demonstrated is that at some point during his playing the CPU hit 95C and 1.475V. Probably initially under boost before the fan had time to spin up faster. One column over is the average and that shows 79C average and 1.35V average. This is more indicative of what you'd expect with that HSF and that kind of a load. Personally, I wouldn't worry about it if I were the OP, or I'd clean and reseat the HSF since he added some extra.

Like I said before, I notice about 70C temps under load with a 360mm AIO and 1.35V with my 3600x. I probably average in the low 60's during gaming. This is with both the Ryzen balanced and the regular Windows balanced plans. The Windows balanced plan seems to be a little cooler for my chip at idle.
Click to expand...
Try it on air. You're using a 360mm radiator liquid cooler! You'd be able to keep a Core i7 920 OCed to 4.2 with 1.5v core cool in a micro ITX case cool with that.
 
kirbyrj said:
Actually, all the OP really demonstrated is that at some point during his playing the CPU hit 95C and 1.475V. Probably initially under boost before the fan had time to spin up faster. One column over is the average and that shows 79C average and 1.35V average. This is more indicative of what you'd expect with that HSF and that kind of a load. Personally, I wouldn't worry about it if I were the OP, or I'd clean and reseat the HSF since he added some extra.

Like I said before, I notice about 70C temps under load with a 360mm AIO and 1.35V with my 3600x. I probably average in the low 60's during gaming. This is with both the Ryzen balanced and the regular Windows balanced plans. The Windows balanced plan seems to be a little cooler for my chip at idle.
Click to expand...
Is that monitoring program taking an average temp every so often or is it taking one current, one high, one low, and then averaging the three? If it is truly just a spike, then it's no problem. Until we know how the program polls for its average, we have no idea how long it ran at 90+, right?
 
Maddnotez said:
Well........I did the -.1 offset/undervolt or whatever you want to call it a while ago at the advise given here. Temps went down to about 85 with stock cooler.

Got a Scythe Big Shuriken 3 and now maxing out about 78.3 ish (hasn't reached 79 yet)

I did also realize I may not have the best fan setup as I moved a couple fans around and saw it get up to 85 again with the Scythe so I moved them back to where I had them. Funny thing here is that my fan placement is just as effective as a $50 cooler which is lame.

I was going to return the cooler but I am ok with 78, I do want to keep a low profile down blower cooler and I am not looking to sink more money into this for an AIO or something like that.


If anyone cares my case is the Enthoo pro and my fans are all Phanteks white LED fans (got this over 5 years ago I think).

2x140 front intake
2x140 top exhaust
2x120 bottom intake
1x140 back exhaust

The case has room for a 3rd top fan but I only have 2. They are in the 2 front slots meaning not on top of CPU. Oddly enough when I moved them over on top of cpu thinking this would give better flow (thinking the air is exhausting out before the intake air even reached the cpu) that is when the temps went up. Doesn't make sense to me but it is what it is.
Click to expand...
Scythe Big Shuriken: 47.00
Enermax Liqmax III 120mm AIO 60.00
EVGA CLC 120mm All-In-One CPU Liquid Cooler 55.00
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
DWD1961 said:
Try it on air. You're using a 360mm radiator liquid cooler! You'd be able to keep a Core i7 920 OCed to 4.2 with 1.5v core cool in a micro ITX case cool with that.
Click to expand...

My girlfriend has a 3600 on my old scythe fuma ;).
 
DWD1961 said:
Is that monitoring program taking an average temp every so often or is it taking one current, one high, one low, and then averaging the three? If it is truly just a spike, then it's no problem. Until we know how the program polls for its average, we have no idea how long it ran at 90+, right?
Click to expand...

Ya really need to read up on monitoring and observer effect on Ryzen cpus. Most monitoring apps don't know what sensor to read on Ryzen so you see individusl core temp instead of die average. If don't you like see spiking temps and get nervous, stop monitoring how you are doing it. Use hwi with AB and RTSS, and use hwi to monitor die average instead. As for voltage, again read up on how quickly Ryzen cpus can perform context switching. Intel cpus are abysmal in this area, that's why it looks like their cpus don't react to anything and on the flipside Ryzen cpus are always jumping all over the place. This gets worse with monitoing apps that are polling based which is like 99% of them. I don't even bother monitoring voltage. I let the system do it's thing. Another point of understanding is effective clocks vs discrete clocks. Again most if not all current monitoring apps are discrete clock based which is not the way you want to monitor Ryzen cpus.
 
thesmokingman said:
Ya really need to read up on monitoring and observer effect on Ryzen cpus. Most monitoring apps don't know what sensor to read on Ryzen so you see individusl core temp instead of die average. If don't you like see spiking temps and get nervous, stop monitoring how you are doing it. Use hwi with AB and RTSS, and use hwi to monitor die average instead. As for voltage, again read up on how quickly Ryzen cpus can perform context switching. Intel cpus are abysmal in this area, that's why it looks like their cpus don't react to anything and on the flipside Ryzen cpus are always jumping all over the place. This gets worse with monitoing apps that are polling based which is like 99% of them. I don't even bother monitoring voltage. I let the system do it's thing. Another point of understanding is effective clocks vs discrete clocks. Again most if not all current monitoring apps are discrete clock based which is not the way you want to monitor Ryzen cpus.
Click to expand...
Agreed about not montoring voltages. I have in the past to see what changes in the BIOS made back in my OCing days, but I always pay attention to CPU temperatures. The discussion isn't, for me, about voltages, but heat on a stock 3600.
 
DWD1961 said:
Agreed about not montoring voltages. I have in the past to see what changes in the BIOS made back in my OCing days, but I always pay attention to CPU temperatures. The discussion isn't, for me, about voltages, but heat on a stock 3600.
Click to expand...

The OP had a bad mount, like real bad as in he didn't have the backing plate mounted, ie loose.
 
You must log in or register to reply here.
Back
Top