Ryzen 5 3600 turbo speeds

[AusWolf]

I dunno, if that's just on light load then that just sounds like a mounting/mounting pressure/contact issue. Especially if it's idling at 50C, without any noticeable background tasks going on. Ryzen idle is pretty annoying but it's easy to tell when it's actually idling and when it's running something in the background. New Be Quiet mounts on the DRP4 are an improvement but sometimes take a bit of fiddling compared to Secufirm.

I don't imagine the similarly sized Shadow Rock LP performs much worse than the NH-L12S, which is stout as hell and kept my 4650G at less than 65C at all times during the short time I had it. Renoir is slightly cooler but not by much. My NH-L9x65 now keeps the 4650G at under 70C at all times, I figure between Matisse and Renoir add about 10C to account for chiplets.

When I had my 3700X it crept closer to 80C under all-out CPU benchmarks under my NH-D9L, which is worse than my NH-U9S and performs about the same as the NH-L12S. IIRC the 3600 doesn't really run any cooler than the 3700X because it runs more wattage per core, and still takes up the 88W PPT and doesn't really "save" any PPT unlike the 5600X.

The 2019 Low-Profile Cooler Shootout: Crowning a Welterweight Champ | The Tech Buyer's Guru (techbuyersguru.com)

An all-core under 1.3V between 4.0-4.2GHz probably works out best for everyday thermals, but make sure it's not a mounting issue first.

I don't think it is. I've tried 3 different coolers so far with similar results. The voltage jumps between 1 and 1.4 V even in idle. No setting seems to influence this behaviour. No such issue with the 5950X, or with the 3100 before that. Why can't this thing just simply work?

 

[Hemmingstamp]

Thanks, everyone, but I'm not planning on overclocking. I just want stable clocks with acceptable thermals - which brings me to my next problem. My be quiet! Shadow Rock LP has arrived, but this thing is still sitting at 50 C idle and hits 80 C as soon as I ask it to do something. I played around with the Windows power settings, disabled literally every performance increasing gimmick in BIOS (e.g. PBO, ASUS enhancements, etc.), still no effect. I never had this issue with the 5950X - maybe I should have gone with Vermeer again. *sigh* Any advice?

Paste, cooler not seated properly, or naff cooler fan perhaps, Bios maybe? I purchased some TG Aeronaut and used a Thermaltake UX200 cooler on my 3600. Never had problems with temps at all. That was in an ITX build with both B550 and B450 boards. I also swapped out the coolers generic fan for a Corsair SP120.

 

[tabascosauz]

I don't think it is. I've tried 3 different coolers so far with similar results. The voltage jumps between 1 and 1.4 V even in idle. No setting seems to influence this behaviour. No such issue with the 5950X, or with the 3100 before that. Why can't this thing just simply work?

I mean, being the owner of a 3100 and 5950X you should already know by now that's just normal idle Vcore behaviour when monitoring with anything other than Ryzen Master......only way is to just get used to it, eliminate any background processes that are hogging resources (e.g. CAM), or do a static OC.

We have the exact same board, no crazy thermals for me on either the 3700X, 4650G (tested once in this board), or 5900X through a lot of different BIOS revisions since last May. Sounds like the fan curve might need to be more aggressive (has been my experience with BQ case and cooler fans that they don't push much air at all unless they really spin up). But then again, the low profile cooler roundup does have the Shadow Rock LP performing barely any better than the Wraith Spire (which you'd expect to see at 80-85C in benchmarks), so it seems like that's just par for the course.

Are you using the mATX tower in your specs or the slim case in your sig? If the former, just get a 92mm or 120mm tower cooler like the U9S/U12S/212 Black and be done with it. If the latter, I'd just get a L12S and double check your paste/mounting pressure. The Shadow Rock is not a performance oriented cooler in the slightest.

Only other thing I can think of is to remember to clear your CMOS when you swap CPUs. The Asus BIOS should do it automatically but still, good practice. Ryzen 5000 and 3000 work somewhat differently and expose different settings in the BIOS.

 

[AusWolf]

I mean, being the owner of a 3100 and 5950X you should already know by now that's just normal idle Vcore behaviour when monitoring with anything other than Ryzen Master......only way is to just get used to it, eliminate any background processes that are hogging resources (e.g. CAM), or do a static OC.

We have the exact same board, no crazy thermals for me on either the 3700X, 4650G (tested once in this board), or 5900X through a lot of different BIOS revisions since last May. Sounds like the fan curve might need to be more aggressive (has been my experience with BQ case and cooler fans that they don't push much air at all unless they really spin up). But then again, the low profile cooler roundup does have the Shadow Rock LP performing barely any better than the Wraith Spire (which you'd expect to see at 80-85C in benchmarks), so it seems like that's just par for the course.

Are you using the mATX tower in your specs or the slim case in your sig? If the former, just get a 92mm or 120mm tower cooler like the U9S/U12S/212 Black and be done with it. If the latter, I'd just get a L12S and double check your paste/mounting pressure. The Shadow Rock is not a performance oriented cooler in the slightest.

Only other thing I can think of is to remember to clear your CMOS when you swap CPUs. The Asus BIOS should do it automatically but still, good practice. Ryzen 5000 and 3000 work somewhat differently and expose different settings in the BIOS.

Sorry for not specifying earlier: I'm using the slim case. I didn't expect the Shadow Rock LP to do a miracle, just to cool a 65 W TDP CPU under normal conditions. I'll try reinstalling it with some Arctic MX-4, but I doubt it'll make any difference.

I mentioned the 5950X and 3100 because I never had any thermal issues with them. The 3100 barely went above 50 C under full load with the Wraith Stealth. I never expected such a huge difference in heat output between the 3100 and 3600.

 

[Deleted member 178884]

Somehow you didn't read that the people who degraded their chips were stressed testing with P95 AVX, Linpack, etc for the hell of it....which are absolutely pointless for gaming PC.

And yet you're pushing your CPU beyond specification without stress testing? Suppose 'gaming' nowadays consists of frequent blue screens? And not to mention, plenty of games use heavy AVX, but hey have fun degrading I guess.

 

[nguyen]

And yet you're pushing your CPU beyond specification without stress testing? Suppose 'gaming' nowadays consists of frequent blue screens? And not to mention, plenty of games use heavy AVX, but hey have fun degrading I guess.

Yeah sure whatever dude
Check out Der8aur pushing his 5800X with 1.4Vcore and it hasn't degraded in 2 months of constant Timespy Physic test
Stop talking like you know stuff please

 

[QuietBob]

My be quiet! Shadow Rock LP has arrived, but this thing is still sitting at 50 C idle

Have you checked your idle core behavior in Ryzen Master? When truly idle, they should be showing as "sleep". Also, which power plan are you using? Under advanced settings, what is your minimum processor state?

 

[Deleted member 178884]

Stop talking like you know stuff please

That's bold coming from somebody who doesn't know what a stress test is, and is claiming 1.37V is "24/7 safe" where's your evidence for this (read: you have none)

Check out Der8aur pushing his 5800X with 1.4Vcore and it hasn't degraded in 2 months of constant Timespy Physic test

And? This is a test across 3 samples out of millions made, it's not representative of all processors made and in zen2, the FIT voltage actually varies. Not to mention Timespy Physics isn't intensive as some games are and the load generally doesn't fluctuate as much.

Either way citation needed, "I stuck 1.6V into my 7980XE and it's safe because I run it at that, most certainly safe for 24/7" *provides no evidence of it being representative of all processors*"

 

[nguyen]

That's bold coming from somebody who doesn't know what a stress test is, and is claiming 1.37V is "24/7 safe" where's your evidence for this (read: you have none)

And? This is a test across 3 samples out of millions made, it's not representative of all processors made and in zen2, the FIT voltage actually varies. Not to mention Timespy Physics isn't intensive as some games are and the load generally doesn't fluctuate as much.

Just watch the video before talking nonsense please, Der8auer has much more credibility than you have.
Also I run my 3600 at that voltage for a whole year already, so that's my evidence against your nonsense.

And if you like to kill your CPU and GPU faster with Prime95 AVX or Furmark with stock voltages, go right ahead.

 

[Deleted member 178884]

Just watch the video before talking nonsense please, Der8auer has much more credibility than you have.
Also I run my 3600 at that voltage for a whole year already, so that's my evidence against your nonsense.

You run one chip at that voltage for a year and it's fine and suddenly that's representative of all CPUs made there? I guess you should replace The Silt then after all you can outdo thousands of processors being tested and proclaim you know safe voltages whilst throwing weightless claims without any citation. And der8auer has more credibility for running three processors at 1.4V on a light load? Lol okay bro, maybe you should be the one not talking nonsense.

https://www.reddit.com/r/overclocking/comments/eu3fbl

https://www.reddit.com/r/overclocking/comments/ejd5c9

https://www.reddit.com/r/overclocking/comments/es754x

https://www.reddit.com/r/overclocking/comments/ej8uqa

And if you like to kill your CPU and GPU faster with Prime95 AVX or Furmark with stock voltages, go right ahead.

Do you feel the same way about RAM stress testing, can imagine you overclock that too and run the OS into to the ground.

 

[nguyen]

You run one chip at that voltage for a year and it's fine and suddenly that's representative of all CPUs made there? I guess you should replace The Silt then after all you can outdo thousands of processors being tested and proclaim you know safe voltages whilst throwing weightless claims without any citation. And der8auer has more credibility for running three processors at 1.4V on a light load? Lol okay bro, maybe you should be the one not talking nonsense.
Do you feel the same way about RAM stress testing, can imagine you overclock that too and run the OS into to the ground.

You really have no statistics beyond anecdotal evidence of some chumps putting their CPU under unrealistic load.
Also your knowledge is really lacking it is quite worrying.
What kill the chip are not voltages, it's power and heat. Power = Voltage x Current

Here is the quote from your beloved Stilt:
"According to FIT, the safe voltage levels for the silicon are around 1.325V in high-current loads and up to 1.47V in low-current loads (i.e ST), depending on the silicon characteristics." - The Stilt

 

[freeagent]

What might be unrealistic to you could be a daily thing to someone else.. as in real world.

 

[Deleted member 178884]

You really have no statistics beyond anecdotal evidence of some chums putting their CPU under unrealistic load.
Also your knowledge is really lacking it is quite worrying.
What kill the chip are not voltages, it's power and heat. Power = Voltage x Current

Link me to the AMD datasheets showing loadline specification and ICCMax current.... oh wait.
Also, yes the formula is correct, but you're ignoring the fact you're shoving 1.37V when there's high current games or similar occurring which can more than easily degrade the CPU under hundreds of hours of gaming, not to even mention this is one CPU out of hundreds that have been tested & degraded.

Here is the quote from your beloved Stilt:
"According to FIT, the safe voltage levels for the silicon are around 1.325V in high-current loads and up to 1.47V in low-current loads (i.e ST), depending on the silicon characteristics." - The Stilt

Now learn to read and you'll read "depending on the silicon characteristics."

So either way, how this is singular chip representative of all of them? Care to enlighten us? Or are you not reading anything you post like some sort of clown?

 

[AusWolf]

Have you checked your idle core behavior in Ryzen Master? When truly idle, they should be showing as "sleep". Also, which power plan are you using? Under advanced settings, what is your minimum processor state?

I never managed to make Ryzen Master work for some reason. When I tried to install it last time, it said it's already installed, but I still can't find it anywhere.

Instead, I use HWinfo for reporting, which never reports clocks going below around 3-3.3 GHz, although the effective clocks are way down, so I suspect everything is OK in that regard. Now I tried disabling PBO, and it's sitting at 3.6 GHz reported idle with a 21 W power consumption and 43 C.

A random thought: is there any way I can influence PBO/PPT from BIOS to keep the processor under 65 W?

 

[nguyen]

What might be unrealistic to you could be a daily thing to someone else.. as in real world.

Understandable, if I were to do rendering (which i don't) or video editing (which I also don't) then I would be using a lower PPT and Voltages.
Still I don't like people who just stress testing for fun and break the crap outta their hardware and just return them to the store, it's scream dishonesty.

 

[WHDS]

Can we not just agree to disagree? There is plenty of information about these things for someone to make a reasonable judgement. At the end of the day, it's your cpu, do with it what you will. Personally, stock with a mem oc is what I run and am very happy with it.

 

[tabascosauz]

Sorry for not specifying earlier: I'm using the slim case. I didn't expect the Shadow Rock LP to do a miracle, just to cool a 65 W TDP CPU under normal conditions. I'll try reinstalling it with some Arctic MX-4, but I doubt it'll make any difference.

I mentioned the 5950X and 3100 because I never had any thermal issues with them. The 3100 barely went above 50 C under full load with the Wraith Stealth. I never expected such a huge difference in heat output between the 3100 and 3600.

I understand that the NH-L12S isn't cheap, but if you can get that cooler I can at least give you a bit of a benchmark with the L12S that I had and experience from the old-ass D9L I own. Then we can tell if something seems off about the thermals. When applying MX-4 make sure the spread ends up covering the heatspreader - AM4 has a big IHS and the ol' LGA115x pea method often is inadequate. Fair sized rice grain is good usually.

The 3100 is known for somewhat abnormally being one of the coolest running Ryzens, it's worlds away from the 3300X which behaves thermally much like a 3600 or 3700X would. The power draw is pretty conservative, and thermal density is as low as can be. It won't be much a useful metric when comparing to any other Matisse CPU.

I looked up the Aerocool case, and it looks like case airflow is quite poor. That's going to have quite the impact, as my temp data on the Noctuas is gathered either in a Cerberus with 2 x A14 intakes and 1 x A9 exhaust, a NCASE with 3 x A12x25 intakes and 1 x A9x14 exhaust, a HT5 with 2 x A8 exhausts directly adjacent to the CPU cooler, or an open bench. So if the Shadow Rock LP is running in the low to mid-80s during a CPU benchmark, I'd say that's not far off the mark given the airflow situation in that case. Any more and you gotta look into your cooler mounting or paste.

 

[Hemmingstamp]

I never managed to make Ryzen Master work for some reason. When I tried to install it last time, it said it's already installed, but I still can't find it anywhere.

Never had that issue at all. I suggest you try again by searching in your C: drive for the AMD folder and see if it really is installed.
I usually grab the latest chipset drivers from the AMD site also because the one's from the mobo vendor can be slightly out of date.

 

[nguyen]

Link me to the AMD datasheets showing loadline specification and ICCMax current.... oh wait.
Also, yes the formula is correct, but you're ignoring the fact you're shoving 1.37V when there's high current games or similar occurring which can more than easily degrade the CPU under hundreds of hours of gaming, not to even mention this is one CPU out of hundreds that have been tested & degraded.

Now learn to read and you'll read "depending on the silicon characteristics."

So either way, how this is singular chip representative of all of them? Care to enlighten us? Or are you not reading anything you post like some sort of clown?

Well my nephew had like couple thousand of hours playing on my 3600 rig, no problem so far. Show me which game use high current please
Sure if it dies in a few years I will make a keychain outta it, happy?

 

[Deleted member 178884]

Show me which game use high current please

BF V, Warzone? There's literally plenty out there, dying they don't exist is quite the cope.

 

[Deleted member 205776]

3600 Silver sample owner here. Don't tinker with the thing manually. I pushed mine to 4.3GHz and it became unstable. Leave it at stock and use Ryzen Master to boost it.

I used ClockTuner to find out it was a silver sample. I was sure I could squeeze some more MHz out of the CPU only to find out it wasn't worth my time after all.

All ClockTuner did was shove 1.55v down my CPU's throat then have the audacity to run a light AVX test. Never gonna use anything made by 1usmus again.

 

[AusWolf]

I understand that the NH-L12S isn't cheap, but if you can get that cooler I can at least give you a bit of a benchmark with the L12S that I had and experience from the old-ass D9L I own. Then we can tell if something seems off about the thermals. When applying MX-4 make sure the spread ends up covering the heatspreader - AM4 has a big IHS and the ol' LGA115x pea method often is inadequate. Fair sized rice grain is good usually.

The 3100 is known for somewhat abnormally being one of the coolest running Ryzens, it's worlds away from the 3300X which behaves thermally much like a 3600 or 3700X would. The power draw is pretty conservative, and thermal density is as low as can be. It won't be much a useful metric when comparing to any other Matisse CPU.

I looked up the Aerocool case, and it looks like case airflow is quite poor. That's going to have quite the impact, as my temp data on the Noctuas is gathered either in a Cerberus with 2 x A14 intakes and 1 x A9 exhaust, a NCASE with 3 x A12x25 intakes and 1 x A9x14 exhaust, a HT5 with 2 x A8 exhausts directly adjacent to the CPU cooler, or an open bench. So if the Shadow Rock LP is running in the low to mid-80s during a CPU benchmark, I'd say that's not far off the mark given the airflow situation in that case. Any more and you gotta look into your cooler mounting or paste.

Thanks for helping so much.

I loved my 3100 to bits for being so economical. Maybe I'll just dig it out of my drawer, pop it back into the mobo and call it a day. I'm sure it would be plenty powerful for the low profile GTX 1650 I bought for it.

Speaking of cases, I bought this one originally for a HTPC, but I thought why not use it as my main PC as it has a huge grate on its side for CPU air intake. It always had the 3100 running around 45-50 degrees with the tiny Wraith Stealth even at full load. Maybe that grate doesn't help with the 3600 as much as I thought.

Other thing: I managed to find the PPT setting in the BIOS. With it set to 65 W, PBO enabled, the CPU runs at the low 70s under a Cinebench run. Weird thing that now HWinfo reports clocks between 4-4.1 GHz, while Windows Task Manager got stuck at 3.6 GHz. Idle clock reporting is also all over the place.

 

[freeagent]

My cooling is set up to take the heat the most intense apps have to give. I wouldn't give my XT more than 1.3 for hardcore stuff. The CPU just cant take it. I'm sure it can but for how long? Some CPUs love voltage and others hate it, depends on which part of the wafer it originates from. That is probably what The Stilt meant about silicon characteristics..

But I am a little out of my element here so

 

[Deleted member 205776]

I wouldn't give my CPU more than stock operation. I want this thing to last. Giving it 1.3v or more is a surefire way of ensuring it doesn't last.

 

[nguyen]

BF V, Warzone? There's literally plenty out there, dying they don't exist is quite the cope.

3600 @ 4.2Ghz all core OCed

3600 pulling 60W at 1080p is hardly anything to worry about vs 200W stress testing under Prime95 AVX.