Question CPU Cooler Options

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
I just installed my entire PC into a new case and I saw some pretty major temperature differences on my CPU due to more fans fitted into my new case. Now I am in the market for a CPU cooler that can fit and isn't much more than $60. I used to be up to date on coolers, but now I'm not too sure what to throw in. Any suggestions?
 

PC Tailor

Illustrious
Ambassador
Here's quite range from low budget, to $60 ish, to slightly more you could consider, all of these are pretty good. The Hyper Evo 212 and Gammaxx are excellent value for money coolers.

CPU Cooler: Deepcool GAMMAXX 400 74.34 CFM CPU Cooler ($24.85 @ OutletPC)
CPU Cooler: Cooler Master MasterLiquid Lite 120 66.7 CFM Liquid CPU Cooler ($59.99 @ SuperBiiz)
CPU Cooler: Deepcool GAMMAXX GT 29.5 CFM CPU Cooler ($39.95 @ Amazon)
CPU Cooler: Cooler Master MasterLiquid ML240L RGB 66.7 CFM Liquid CPU Cooler ($63.89 @ OutletPC)
CPU Cooler: NZXT Kraken M22 Liquid CPU Cooler ($69.89 @ OutletPC)
CPU Cooler: be quiet! Dark Rock 4 CPU Cooler ($74.90 @ Amazon)

Also great but may need to ensure it comes with the relevant backplate/adapter for your motherboard:

CPU Cooler:
Cooler Master Hyper 212 EVO 82.9 CFM Sleeve Bearing CPU Cooler ($24.99 @ Newegg)
CPU Cooler: Noctua NH-U12S 55 CFM CPU Cooler ($57.99 @ Amazon)
CPU Cooler: Noctua NH-D15 82.5 CFM CPU Cooler ($89.95 @ Amazon)

You may also find this reference useful:
https://www.tomshardware.com/reviews/best-cpu-coolers,4181.html
 

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
Here's quite range from low budget, to $60 ish, to slightly more you could consider, all of these are pretty good. The Hyper Evo 212 and Gammaxx are excellent value for money coolers.

CPU Cooler: Deepcool GAMMAXX 400 74.34 CFM CPU Cooler ($24.85 @ OutletPC)
CPU Cooler: Cooler Master MasterLiquid Lite 120 66.7 CFM Liquid CPU Cooler ($59.99 @ SuperBiiz)
CPU Cooler: Deepcool GAMMAXX GT 29.5 CFM CPU Cooler ($39.95 @ Amazon)
CPU Cooler: Cooler Master MasterLiquid ML240L RGB 66.7 CFM Liquid CPU Cooler ($63.89 @ OutletPC)
CPU Cooler: NZXT Kraken M22 Liquid CPU Cooler ($69.89 @ OutletPC)
CPU Cooler: be quiet! Dark Rock 4 CPU Cooler ($74.90 @ Amazon)

Also great but may need to ensure it comes with the relevant backplate/adapter for your motherboard:

CPU Cooler:
Cooler Master Hyper 212 EVO 82.9 CFM Sleeve Bearing CPU Cooler ($24.99 @ Newegg)
CPU Cooler: Noctua NH-U12S 55 CFM CPU Cooler ($57.99 @ Amazon)
CPU Cooler: Noctua NH-D15 82.5 CFM CPU Cooler ($89.95 @ Amazon)

You may also find this reference useful:
https://www.tomshardware.com/reviews/best-cpu-coolers,4181.html
Would you recommend liquid over air cooling or vice versa? I'm just slightly nervous about having a liquid cooler in my PC for obvious reasons.
 

PC Tailor

Illustrious
Ambassador
Would you recommend liquid over air cooling or vice versa? I'm just slightly nervous about having a liquid cooler in my PC for obvious reasons.
If you're strictly referring to AIO water coolers - there's not that much difference from top air coolers in most cases. Custom Loops however are very different .

AIO just tend to be more space effective. There is little to worry about with modern day AIO, they just have a shelf life that's all. Custom Loops are more troublesome as they require maintenance and aren't all one unit.

But as far as efficiency goes, you'll find the best air coolers are as good or better than most AIO water coolers. Just there are some fantastic, really nice looking AIO coolers nowadays for a good price.
 
At 60 dollars, there is no other cooler you can buy that will give you as good of performance as this, for the money.

PCPartPicker Part List

CPU Cooler: Scythe Ninja 5 43.03 CFM CPU Cooler ($59.99 @ Amazon)
Total: $59.99
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2019-07-23 17:55 EDT-0400


Under an overclocked load it performs within one degree of the NH-D15 and .25 degrees BETTER than the H150i Pro AIO closed loop cooler.

https://www.tweaktown.com/reviews/8925/scythe-ninja-5-cpu-cooler-review/index6.html


At only 155mm tall it will easily fit your case AND at under 20db (14.5 according to it's spec sheet) it's seriously quieter than any AIO even without factoring in the pump noise.
 

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
At 60 dollars, there is no other cooler you can buy that will give you as good of performance as this, for the money.

PCPartPicker Part List

CPU Cooler: Scythe Ninja 5 43.03 CFM CPU Cooler ($59.99 @ Amazon)
Total: $59.99
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2019-07-23 17:55 EDT-0400


Under an overclocked load it performs within one degree of the NH-D15 and .25 degrees BETTER than the H150i Pro AIO closed loop cooler.

https://www.tweaktown.com/reviews/8925/scythe-ninja-5-cpu-cooler-review/index6.html


At only 155mm tall it will easily fit your case AND at under 20db (14.5 according to it's spec sheet) it's seriously quieter than any AIO even without factoring in the pump noise.
Would this fit in my build listed?
 

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
At 60 dollars, there is no other cooler you can buy that will give you as good of performance as this, for the money.

PCPartPicker Part List

CPU Cooler: Scythe Ninja 5 43.03 CFM CPU Cooler ($59.99 @ Amazon)
Total: $59.99
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2019-07-23 17:55 EDT-0400


Under an overclocked load it performs within one degree of the NH-D15 and .25 degrees BETTER than the H150i Pro AIO closed loop cooler.

https://www.tweaktown.com/reviews/8925/scythe-ninja-5-cpu-cooler-review/index6.html


At only 155mm tall it will easily fit your case AND at under 20db (14.5 according to it's spec sheet) it's seriously quieter than any AIO even without factoring in the pump noise.
I didn't see that you said it would fit, that's my bad. Would ram clearance be a problem at all?
 
With two sticks, installed in the A2 and B2 slots, which is where they BELONG with only two sticks installed, and are the second and fourth slots over from the CPU socket, you should have zero problems with clearance.

Even IF you had four DIMMs installed, and HAD to slide the front fan UP the heatsink a few millimeters, it wouldn't hurt performance much if at all anyhow.

With only two DIMMs installed you shouldn't need to do that anyhow.


I did notice something on the Ninja 5 product page however that you ought to be aware of and verify.

Please note that an original backplate with screw mounting is needed for installation on mainboards with AMD sockets. Please check before buying if the backplate is fixed by screws (example)! If the backplate uses plastic-pins, mounting of this cooler is not possible.


On the Thermalright Macho rev.B, there is no such issue, as it clearly states:

the cooler is equipped with mounting material for the new AMD AM4 Ryzen processors and can be mounted on the new Intel LGA 2066 processors.

So if you have a backplate with actual fasteners, you can use the Ninja 5 but if it uses push pins, it wouldn't work.

I wasn't aware of that previously, so I apologize for that oversight. Still, those will be good options. Below, in the event those coolers don't do the trick, are my recommended list of coolers in general, which might offer you a flavor that you prefer AND they are all good coolers although obviously some are better than others.


Below is my list of preferred CPU AIR coolers, also known as Heatsink fans (HSF).

Do not look here for recommendations on water/liquid cooling solutions. There are none to be found.


They are basically listed in order of preference, from top to bottom. To some degree that preference is based on known performance on similarly overclocked configurations, but not entirely. There are likely a couple of units that are placed closer to the top not because they offer purely better performance than another cooler which is below it, but potentially due to a variety of reasons.

One model might be placed higher than another with the same or similar performance, but has quieter or higher quality fans. It may have the same performance but a better warranty. Long term quality may be higher. It may be less expensive in some cases. Maybe it performs slightly worse, but has quieter fans and a better "fan pitch". Some fans with equal decibel levels do not "sound" like they are the same as the specific pitch heard from one fan might be less annoying than another.

In any case, these are not "tiered" and are not a 100% be all, end all ranking. They are simply MY preference when looking at coolers for a build or when making recommendations. Often, which HSF gets chosen depends on what is on this list and fits the budget or is priced right at the time due to a sale or rebate. Hopefully it will help you and you can rest assured that every cooler listed here is a model that to some degree or other is generally a quality unit which is a lot more likely to be worth the money spent on it than on many other models out there that might look to be a similarly worthwhile investment.

Certainly there are a great many other very good coolers out there, but these are models which are usually available to most anybody building a system or looking for a cooler, regardless of what part of the world they might live in. As always, professional reviews are usually an absolutely essential part of the process of finding a cooler so if you are looking at a model not listed here, I would highly recommend looking at at least two or three professional reviews first.

If you cannot find two reviews of any given cooler, it is likely either too new to have been reviewed yet or it sucked, and nobody wanted to buy one in order to review it plus the manufacturer refused to send samples out to the sites that perform reviews because they knew it would likely get bad publicity.

IMO, nobody out there is making better fans, overall, than Noctua, followed pretty closely by Thermalright. So if you intend to match case fans to the same brand on your HSF, those are pretty hard to beat. Of course, Corsair has it's Maglev fans, and those are pretty damn good too, but since they don't make CPU air cooling products, only AIO water coolers, they cannot join the party.


Noctua NH-D14 (Replace stock fans with NF-A14 industrialPPC 2000rpm)
Noctua NH-D15/D15 SE-AM4
Noctua NH-D14 (With original fans)
Thermalright Silver arrow IB-E Extreme
Phanteks PH-TC14PE (BK,BL, OR or RD)
Cryorig R1 Ultimate or Universal
Thermalright Legrand Macho RT
Scythe Ninja 5
FSP Windale 6
Scythe Mugen 5 rev.b
Noctua NH-U14S
Thermalright Macho rev.B
Thermalright Macho (Direct, 120)
Scythe Mugen max
BeQuiet dark rock pro (3 or 4)
BeQuiet dark rock (3 or 4)
Deepcool Assassin II
Thermalright true spirit 140 (Direct, Power, BW)
Cryorig H5
Noctua NH-U12S
Phanteks PH-TC12DX (Any)
Phanteks PH-TC14S
Cryorig H7
Deepcool Gammaxx 400

Cooler Master Hyper 212 (EVO, X, RGB. I'd only recommend this cooler if no other good aftermarket models are available to you.)


It may not be obvious, but is probably worth mentioning, that not all cooler models will fit all CPU sockets as aftermarket coolers generally require an adapter intended for use with that socket. Some coolers that fit an AMD platform might not fit a later AMD platform, or an Intel platform. Often these coolers come with adapters for multiple types of platforms but be sure to verify that a specific cooler WILL work with your platform before purchasing one and finding out later that it will not.
 

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
With two sticks, installed in the A2 and B2 slots, which is where they BELONG with only two sticks installed, and are the second and fourth slots over from the CPU socket, you should have zero problems with clearance.

Even IF you had four DIMMs installed, and HAD to slide the front fan UP the heatsink a few millimeters, it wouldn't hurt performance much if at all anyhow.

With only two DIMMs installed you shouldn't need to do that anyhow.


I did notice something on the Ninja 5 product page however that you ought to be aware of and verify.




On the Thermalright Macho rev.B, there is no such issue, as it clearly states:



So if you have a backplate with actual fasteners, you can use the Ninja 5 but if it uses push pins, it wouldn't work.

I wasn't aware of that previously, so I apologize for that oversight. Still, those will be good options. Below, in the event those coolers don't do the trick, are my recommended list of coolers in general, which might offer you a flavor that you prefer AND they are all good coolers although obviously some are better than others.


Below is my list of preferred CPU AIR coolers, also known as Heatsink fans (HSF).

Do not look here for recommendations on water/liquid cooling solutions. There are none to be found.


They are basically listed in order of preference, from top to bottom. To some degree that preference is based on known performance on similarly overclocked configurations, but not entirely. There are likely a couple of units that are placed closer to the top not because they offer purely better performance than another cooler which is below it, but potentially due to a variety of reasons.

One model might be placed higher than another with the same or similar performance, but has quieter or higher quality fans. It may have the same performance but a better warranty. Long term quality may be higher. It may be less expensive in some cases. Maybe it performs slightly worse, but has quieter fans and a better "fan pitch". Some fans with equal decibel levels do not "sound" like they are the same as the specific pitch heard from one fan might be less annoying than another.

In any case, these are not "tiered" and are not a 100% be all, end all ranking. They are simply MY preference when looking at coolers for a build or when making recommendations. Often, which HSF gets chosen depends on what is on this list and fits the budget or is priced right at the time due to a sale or rebate. Hopefully it will help you and you can rest assured that every cooler listed here is a model that to some degree or other is generally a quality unit which is a lot more likely to be worth the money spent on it than on many other models out there that might look to be a similarly worthwhile investment.

Certainly there are a great many other very good coolers out there, but these are models which are usually available to most anybody building a system or looking for a cooler, regardless of what part of the world they might live in. As always, professional reviews are usually an absolutely essential part of the process of finding a cooler so if you are looking at a model not listed here, I would highly recommend looking at at least two or three professional reviews first.

If you cannot find two reviews of any given cooler, it is likely either too new to have been reviewed yet or it sucked, and nobody wanted to buy one in order to review it plus the manufacturer refused to send samples out to the sites that perform reviews because they knew it would likely get bad publicity.

IMO, nobody out there is making better fans, overall, than Noctua, followed pretty closely by Thermalright. So if you intend to match case fans to the same brand on your HSF, those are pretty hard to beat. Of course, Corsair has it's Maglev fans, and those are pretty damn good too, but since they don't make CPU air cooling products, only AIO water coolers, they cannot join the party.


Noctua NH-D14 (Replace stock fans with NF-A14 industrialPPC 2000rpm)
Noctua NH-D15/D15 SE-AM4
Noctua NH-D14 (With original fans)
Thermalright Silver arrow IB-E Extreme
Phanteks PH-TC14PE (BK,BL, OR or RD)
Cryorig R1 Ultimate or Universal
Thermalright Legrand Macho RT
Scythe Ninja 5
FSP Windale 6
Scythe Mugen 5 rev.b
Noctua NH-U14S
Thermalright Macho rev.B
Thermalright Macho (Direct, 120)
Scythe Mugen max
BeQuiet dark rock pro (3 or 4)
BeQuiet dark rock (3 or 4)
Deepcool Assassin II
Thermalright true spirit 140 (Direct, Power, BW)
Cryorig H5
Noctua NH-U12S
Phanteks PH-TC12DX (Any)
Phanteks PH-TC14S
Cryorig H7
Deepcool Gammaxx 400

Cooler Master Hyper 212 (EVO, X, RGB. I'd only recommend this cooler if no other good aftermarket models are available to you.)


It may not be obvious, but is probably worth mentioning, that not all cooler models will fit all CPU sockets as aftermarket coolers generally require an adapter intended for use with that socket. Some coolers that fit an AMD platform might not fit a later AMD platform, or an Intel platform. Often these coolers come with adapters for multiple types of platforms but be sure to verify that a specific cooler WILL work with your platform before purchasing one and finding out later that it will not.
So, for the Scythe cooler, you're saying if I have the stock backplate from the stock cooler that uses screws instead, I could use it?
 
Sounds good. I think you'll be extremely pleased with the quality and performance of that cooler. Don't jump out of your skin when you see how BIG it is though. Most people who have never used a good aftermarket cooler tend to freak when they see how big they actually are, even the smaller of the big coolers, like the slim, single finstack Noctua NH-U14S or even the Gammaxx 400. They are all big. That particular one, is really big, although not nearly as big as the big double fin stack coolers like the D14/D15 or similar models.

Even those who "used to be familiar" with the more popular coolers, might get temporary culture shock, depending on how long ago you were familiar with the current high end coolers.
 

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
Sounds good. I think you'll be extremely pleased with the quality and performance of that cooler. Don't jump out of your skin when you see how BIG it is though. Most people who have never used a good aftermarket cooler tend to freak when they see how big they actually are, even the smaller of the big coolers, like the slim, single finstack Noctua NH-U14S or even the Gammaxx 400. They are all big. That particular one, is really big, although not nearly as big as the big double fin stack coolers like the D14/D15 or similar models.

Even those who "used to be familiar" with the more popular coolers, might get temporary culture shock, depending on how long ago you were familiar with the current high end coolers.
Yeah, I've never used an aftermarket cooler before personally, I just knew what was what. I'm honestly getting it just so my computer isn't always a heater for my room, especially in the summer. Maybe I'll crank out a little overclock too, can't say yet.
 

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
Sounds good. I think you'll be extremely pleased with the quality and performance of that cooler. Don't jump out of your skin when you see how BIG it is though. Most people who have never used a good aftermarket cooler tend to freak when they see how big they actually are, even the smaller of the big coolers, like the slim, single finstack Noctua NH-U14S or even the Gammaxx 400. They are all big. That particular one, is really big, although not nearly as big as the big double fin stack coolers like the D14/D15 or similar models.

Even those who "used to be familiar" with the more popular coolers, might get temporary culture shock, depending on how long ago you were familiar with the current high end coolers.
I just installed the cooler, and I am now a happy boy. I was getting aggravated at first, but that was due to me being an absolute moron and leaving the plastic film on the heatsink and installing it that way. Now my Ryzen sits at a nice 35 degrees celsius, and reached a max of 75 degrees on a run of Cinebench. Im assuming those are within specs?
 
Those are within specs, but they seem to be worse load temps than what your indicated in your original post. Also, 75°C while running Cinebench, on a non-overclocked machine, seems rather high.

I'd do a thermal test using Prime95 as follows. This is the standard test procedure for thermal testing. Are you using Ryzen master to monitor temps?

Final testing with Prime95

It is highly advisable that you do a final test using Prime95 version 26.6 or the latest version WITH AVX and AVX2 disabled, and run a custom configured Blend test. You can also use the Blend mode option as is, but after a fair amount of personal testing, asking questions from some long time members with engineering level degrees that have forgotten more about memory architectures than you or I will ever know, and gathering opinions from a wide array of memory enthusiasts around the web, I'm pretty confident that the custom option is a lot more likely to find errors with the memory configuration, and faster, if there are any to be found.

Please note as this is rather important, if you prefer, or have problems running version 26.6 because you have a newer platform that doesn't want to play nice with version 26.6, you can use the latest version of Prime95 with the Custom test selected but you will need to make the following change.

In the bottom of the Torture test selection popup menu there will be some options for disabling AVX. I recommend that you do so, not because we are doing thermal testing and require a steady state workload (Which AVX wouldn't affect anyhow, as Computronix explained to me), but because the last thing you need during memory testing is having to worry about CPU temperatures, and you will, with AVX enabled.

So, uncheck the option for AVX2. That will un-gray the option for AVX, and uncheck that box as well.

Now open Prime95.

Click on "Custom". Input a value of 512k in the minimum FFT size field. Leave the maximum FFT size field at 4096k. In the "Memory to use" field you should take a look at your current memory allocation in either HWinfo or system resource monitor. Whatever "free" memory is available, input approximately 75% of that amount. So if you currently have 16GB of installed memory, and approximately 3GB are in use or reserved leaving somewhere in the neighborhood of 13GB free, then enter something close to 75% of that amount.

So if you have 13GB free, or something reasonably close to that, then 75% of THAT would be 9.75GB, which, when multiplies times 1024 will roughly equal about 9984MB. You can average things out by simply selecting the closest multiple of 1024 to that amount just to keep it simple, so we'll say 10 x 1024= 10240mb and enter that amount in the field for "Memory to use (MB)". We are still well within the 13GB of unused memory BUT we have left enough memory unused so that if Windows decides to load some other process or background program, or an already loaded one suddenly needs more, we won't run into a situation where the system errors out due to lack of memory because we've dedicated it all to testing.

I've experienced false errors and system freezes during this test from over allocating memory, so stick to the method above and you should be ok.


Moving right along, do not change the time to run each FFT size.Leave that set to 15 minutes.

Click run and run the Custom test for 8 hours. If it passed Memtest86 and it passes 8 hours of the Custom test, the memory is 100% stable, or as close to it as you are ever likely to get but a lot of experts in the area of memory configuration suggest that running the extended Windows memory diagnostic test is also a pretty good idea too.

If you get errors, (and you will want to run HWinfo alongside Prime95 so you can periodically monitor each thread as Prime will not stop running just because one worker drops out, so you need to watch HWinfo to see if there are any threads not showing 100% usage which means one of the workers errored and was dropped) then you need to either change the timings, change the DRAM voltage or change the DRAM termination voltage, which should be approximately half of the full DRAM voltage.

There are also other bios settings that can affect the memory configuration AND stability, such as the VCCIO and system agent voltages, so if you have problems with stability at higher clock speeds you might want to look at increasing those slightly. Usually, for Intel at least, something in the neighborhood of 1.1v on both those is pretty safe. There are a substantial number of guides out there covering those two settings, but most of them are found within CPU overclocking guides so look there in guides relevant to your platform.

As a further measure of assurance that your WHOLE configuration is stable, you can download and run Realbench for 8 hours. If the system freezes or fails when running Realbench with your full memory amount set, try running it again but select only half your amount of installed memory.



After doing all that, do the same test again WITHOUT the side panel on, to see if there is a drop in temperature indicating that there might be an airflow problem. If there is, you may want to look at the possibility of adding or reconfiguring the case fan setup.
 

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
Those are within specs, but they seem to be worse load temps than what your indicated in your original post. Also, 75°C while running Cinebench, on a non-overclocked machine, seems rather high.

I'd do a thermal test using Prime95 as follows. This is the standard test procedure for thermal testing. Are you using Ryzen master to monitor temps?

Final testing with Prime95

It is highly advisable that you do a final test using Prime95 version 26.6 or the latest version WITH AVX and AVX2 disabled, and run a custom configured Blend test. You can also use the Blend mode option as is, but after a fair amount of personal testing, asking questions from some long time members with engineering level degrees that have forgotten more about memory architectures than you or I will ever know, and gathering opinions from a wide array of memory enthusiasts around the web, I'm pretty confident that the custom option is a lot more likely to find errors with the memory configuration, and faster, if there are any to be found.

Please note as this is rather important, if you prefer, or have problems running version 26.6 because you have a newer platform that doesn't want to play nice with version 26.6, you can use the latest version of Prime95 with the Custom test selected but you will need to make the following change.

In the bottom of the Torture test selection popup menu there will be some options for disabling AVX. I recommend that you do so, not because we are doing thermal testing and require a steady state workload (Which AVX wouldn't affect anyhow, as Computronix explained to me), but because the last thing you need during memory testing is having to worry about CPU temperatures, and you will, with AVX enabled.

So, uncheck the option for AVX2. That will un-gray the option for AVX, and uncheck that box as well.

Now open Prime95.

Click on "Custom". Input a value of 512k in the minimum FFT size field. Leave the maximum FFT size field at 4096k. In the "Memory to use" field you should take a look at your current memory allocation in either HWinfo or system resource monitor. Whatever "free" memory is available, input approximately 75% of that amount. So if you currently have 16GB of installed memory, and approximately 3GB are in use or reserved leaving somewhere in the neighborhood of 13GB free, then enter something close to 75% of that amount.

So if you have 13GB free, or something reasonably close to that, then 75% of THAT would be 9.75GB, which, when multiplies times 1024 will roughly equal about 9984MB. You can average things out by simply selecting the closest multiple of 1024 to that amount just to keep it simple, so we'll say 10 x 1024= 10240mb and enter that amount in the field for "Memory to use (MB)". We are still well within the 13GB of unused memory BUT we have left enough memory unused so that if Windows decides to load some other process or background program, or an already loaded one suddenly needs more, we won't run into a situation where the system errors out due to lack of memory because we've dedicated it all to testing.

I've experienced false errors and system freezes during this test from over allocating memory, so stick to the method above and you should be ok.


Moving right along, do not change the time to run each FFT size.Leave that set to 15 minutes.

Click run and run the Custom test for 8 hours. If it passed Memtest86 and it passes 8 hours of the Custom test, the memory is 100% stable, or as close to it as you are ever likely to get but a lot of experts in the area of memory configuration suggest that running the extended Windows memory diagnostic test is also a pretty good idea too.

If you get errors, (and you will want to run HWinfo alongside Prime95 so you can periodically monitor each thread as Prime will not stop running just because one worker drops out, so you need to watch HWinfo to see if there are any threads not showing 100% usage which means one of the workers errored and was dropped) then you need to either change the timings, change the DRAM voltage or change the DRAM termination voltage, which should be approximately half of the full DRAM voltage.

There are also other bios settings that can affect the memory configuration AND stability, such as the VCCIO and system agent voltages, so if you have problems with stability at higher clock speeds you might want to look at increasing those slightly. Usually, for Intel at least, something in the neighborhood of 1.1v on both those is pretty safe. There are a substantial number of guides out there covering those two settings, but most of them are found within CPU overclocking guides so look there in guides relevant to your platform.

As a further measure of assurance that your WHOLE configuration is stable, you can download and run Realbench for 8 hours. If the system freezes or fails when running Realbench with your full memory amount set, try running it again but select only half your amount of installed memory.



After doing all that, do the same test again WITHOUT the side panel on, to see if there is a drop in temperature indicating that there might be an airflow problem. If there is, you may want to look at the possibility of adding or reconfiguring the case fan setup.
I am indeed using Ryzen Master to monitor temps. On my last machine, I didn't use Cinebench for stress testing, because that maxed my thermals near immediately. If my thermals are high, what could the culprit be? It's not exactly cool in my room, so the ambient temps may be a factor, and I just reconfigured my case and have as many fans as I have headers for (2 front intake, one rear exhaust). What else could the issue be if my temps remain high?
 

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
Those are within specs, but they seem to be worse load temps than what your indicated in your original post. Also, 75°C while running Cinebench, on a non-overclocked machine, seems rather high.

I'd do a thermal test using Prime95 as follows. This is the standard test procedure for thermal testing. Are you using Ryzen master to monitor temps?

Final testing with Prime95

It is highly advisable that you do a final test using Prime95 version 26.6 or the latest version WITH AVX and AVX2 disabled, and run a custom configured Blend test. You can also use the Blend mode option as is, but after a fair amount of personal testing, asking questions from some long time members with engineering level degrees that have forgotten more about memory architectures than you or I will ever know, and gathering opinions from a wide array of memory enthusiasts around the web, I'm pretty confident that the custom option is a lot more likely to find errors with the memory configuration, and faster, if there are any to be found.

Please note as this is rather important, if you prefer, or have problems running version 26.6 because you have a newer platform that doesn't want to play nice with version 26.6, you can use the latest version of Prime95 with the Custom test selected but you will need to make the following change.

In the bottom of the Torture test selection popup menu there will be some options for disabling AVX. I recommend that you do so, not because we are doing thermal testing and require a steady state workload (Which AVX wouldn't affect anyhow, as Computronix explained to me), but because the last thing you need during memory testing is having to worry about CPU temperatures, and you will, with AVX enabled.

So, uncheck the option for AVX2. That will un-gray the option for AVX, and uncheck that box as well.

Now open Prime95.

Click on "Custom". Input a value of 512k in the minimum FFT size field. Leave the maximum FFT size field at 4096k. In the "Memory to use" field you should take a look at your current memory allocation in either HWinfo or system resource monitor. Whatever "free" memory is available, input approximately 75% of that amount. So if you currently have 16GB of installed memory, and approximately 3GB are in use or reserved leaving somewhere in the neighborhood of 13GB free, then enter something close to 75% of that amount.

So if you have 13GB free, or something reasonably close to that, then 75% of THAT would be 9.75GB, which, when multiplies times 1024 will roughly equal about 9984MB. You can average things out by simply selecting the closest multiple of 1024 to that amount just to keep it simple, so we'll say 10 x 1024= 10240mb and enter that amount in the field for "Memory to use (MB)". We are still well within the 13GB of unused memory BUT we have left enough memory unused so that if Windows decides to load some other process or background program, or an already loaded one suddenly needs more, we won't run into a situation where the system errors out due to lack of memory because we've dedicated it all to testing.

I've experienced false errors and system freezes during this test from over allocating memory, so stick to the method above and you should be ok.


Moving right along, do not change the time to run each FFT size.Leave that set to 15 minutes.

Click run and run the Custom test for 8 hours. If it passed Memtest86 and it passes 8 hours of the Custom test, the memory is 100% stable, or as close to it as you are ever likely to get but a lot of experts in the area of memory configuration suggest that running the extended Windows memory diagnostic test is also a pretty good idea too.

If you get errors, (and you will want to run HWinfo alongside Prime95 so you can periodically monitor each thread as Prime will not stop running just because one worker drops out, so you need to watch HWinfo to see if there are any threads not showing 100% usage which means one of the workers errored and was dropped) then you need to either change the timings, change the DRAM voltage or change the DRAM termination voltage, which should be approximately half of the full DRAM voltage.

There are also other bios settings that can affect the memory configuration AND stability, such as the VCCIO and system agent voltages, so if you have problems with stability at higher clock speeds you might want to look at increasing those slightly. Usually, for Intel at least, something in the neighborhood of 1.1v on both those is pretty safe. There are a substantial number of guides out there covering those two settings, but most of them are found within CPU overclocking guides so look there in guides relevant to your platform.

As a further measure of assurance that your WHOLE configuration is stable, you can download and run Realbench for 8 hours. If the system freezes or fails when running Realbench with your full memory amount set, try running it again but select only half your amount of installed memory.



After doing all that, do the same test again WITHOUT the side panel on, to see if there is a drop in temperature indicating that there might be an airflow problem. If there is, you may want to look at the possibility of adding or reconfiguring the case fan setup.
Also, is this testing my thermal capacities or my ram? I did a windows extended mem. diagnostic when I first got the machine, and it passed @ 3200mhz.
 

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
Those are within specs, but they seem to be worse load temps than what your indicated in your original post. Also, 75°C while running Cinebench, on a non-overclocked machine, seems rather high.

I'd do a thermal test using Prime95 as follows. This is the standard test procedure for thermal testing. Are you using Ryzen master to monitor temps?

Final testing with Prime95

It is highly advisable that you do a final test using Prime95 version 26.6 or the latest version WITH AVX and AVX2 disabled, and run a custom configured Blend test. You can also use the Blend mode option as is, but after a fair amount of personal testing, asking questions from some long time members with engineering level degrees that have forgotten more about memory architectures than you or I will ever know, and gathering opinions from a wide array of memory enthusiasts around the web, I'm pretty confident that the custom option is a lot more likely to find errors with the memory configuration, and faster, if there are any to be found.

Please note as this is rather important, if you prefer, or have problems running version 26.6 because you have a newer platform that doesn't want to play nice with version 26.6, you can use the latest version of Prime95 with the Custom test selected but you will need to make the following change.

In the bottom of the Torture test selection popup menu there will be some options for disabling AVX. I recommend that you do so, not because we are doing thermal testing and require a steady state workload (Which AVX wouldn't affect anyhow, as Computronix explained to me), but because the last thing you need during memory testing is having to worry about CPU temperatures, and you will, with AVX enabled.

So, uncheck the option for AVX2. That will un-gray the option for AVX, and uncheck that box as well.

Now open Prime95.

Click on "Custom". Input a value of 512k in the minimum FFT size field. Leave the maximum FFT size field at 4096k. In the "Memory to use" field you should take a look at your current memory allocation in either HWinfo or system resource monitor. Whatever "free" memory is available, input approximately 75% of that amount. So if you currently have 16GB of installed memory, and approximately 3GB are in use or reserved leaving somewhere in the neighborhood of 13GB free, then enter something close to 75% of that amount.

So if you have 13GB free, or something reasonably close to that, then 75% of THAT would be 9.75GB, which, when multiplies times 1024 will roughly equal about 9984MB. You can average things out by simply selecting the closest multiple of 1024 to that amount just to keep it simple, so we'll say 10 x 1024= 10240mb and enter that amount in the field for "Memory to use (MB)". We are still well within the 13GB of unused memory BUT we have left enough memory unused so that if Windows decides to load some other process or background program, or an already loaded one suddenly needs more, we won't run into a situation where the system errors out due to lack of memory because we've dedicated it all to testing.

I've experienced false errors and system freezes during this test from over allocating memory, so stick to the method above and you should be ok.


Moving right along, do not change the time to run each FFT size.Leave that set to 15 minutes.

Click run and run the Custom test for 8 hours. If it passed Memtest86 and it passes 8 hours of the Custom test, the memory is 100% stable, or as close to it as you are ever likely to get but a lot of experts in the area of memory configuration suggest that running the extended Windows memory diagnostic test is also a pretty good idea too.

If you get errors, (and you will want to run HWinfo alongside Prime95 so you can periodically monitor each thread as Prime will not stop running just because one worker drops out, so you need to watch HWinfo to see if there are any threads not showing 100% usage which means one of the workers errored and was dropped) then you need to either change the timings, change the DRAM voltage or change the DRAM termination voltage, which should be approximately half of the full DRAM voltage.

There are also other bios settings that can affect the memory configuration AND stability, such as the VCCIO and system agent voltages, so if you have problems with stability at higher clock speeds you might want to look at increasing those slightly. Usually, for Intel at least, something in the neighborhood of 1.1v on both those is pretty safe. There are a substantial number of guides out there covering those two settings, but most of them are found within CPU overclocking guides so look there in guides relevant to your platform.

As a further measure of assurance that your WHOLE configuration is stable, you can download and run Realbench for 8 hours. If the system freezes or fails when running Realbench with your full memory amount set, try running it again but select only half your amount of installed memory.



After doing all that, do the same test again WITHOUT the side panel on, to see if there is a drop in temperature indicating that there might be an airflow problem. If there is, you may want to look at the possibility of adding or reconfiguring the case fan setup.
Link to screenshot
 
Last edited:
You are right. I am sorry. I gave you the wrong test procedure. I appologize.

I gave you my Prime test for memory testing. This is the correct procedure for verifying thermal compliance.


Run Prime95 version 26.6 or the latest version, in which case you will want to disable AVX and AVX 2 in the options at the bottom of the torture test options screen and choose the "Small FFT test option". Do not choose "Smallest FFT", just "Small FFT". Run this for 15 minutes while monitoring your core/package temperatures to verify that you do not exceed the thermal specifications of your CPU.

(This should be considered to be 80°C for most generations of Intel processor and for current Ryzen CPUs. For older AMD FX and Phenom series, you should use a thermal monitor that has options for "Distance to TJmax" (Like AMD Overdrive, (which is NOT compatible with Ryzen platforms) and you want to NOT see distance to TJmax drop below 10°C distance to TJmax. Anything that is MORE than 10°C distance to TJmax is within the allowed thermal envelope.)

Take a screenshot of the Ryzen master window about five minutes into the test as that is long enough to saturate the package and heatsink in most cases when running a stock configuration. If you are overclocking, I would wait until between 10-15 minutes, because I've seen thermal peaks happen occasionally after ten minutes but never beyond 15.
 

advancedmixedgaming

Distinguished
Nov 5, 2015
468
4
18,865
You are right. I am sorry. I gave you the wrong test procedure. I appologize.

I gave you my Prime test for memory testing. This is the correct procedure for verifying thermal compliance.


Run Prime95 version 26.6 or the latest version, in which case you will want to disable AVX and AVX 2 in the options at the bottom of the torture test options screen and choose the "Small FFT test option". Do not choose "Smallest FFT", just "Small FFT". Run this for 15 minutes while monitoring your core/package temperatures to verify that you do not exceed the thermal specifications of your CPU.

(This should be considered to be 80°C for most generations of Intel processor and for current Ryzen CPUs. For older AMD FX and Phenom series, you should use a thermal monitor that has options for "Distance to TJmax" (Like AMD Overdrive, (which is NOT compatible with Ryzen platforms) and you want to NOT see distance to TJmax drop below 10°C distance to TJmax. Anything that is MORE than 10°C distance to TJmax is within the allowed thermal envelope.)

Take a screenshot of the Ryzen master window about five minutes into the test as that is long enough to saturate the package and heatsink in most cases when running a stock configuration. If you are overclocking, I would wait until between 10-15 minutes, because I've seen thermal peaks happen occasionally after ten minutes but never beyond 15.
No worries, I'll post a screenshot once my test runs.