Almost Everyone is Wasting Their Money On CPU Coolers

[Vishera]

@LinusTech The testing method is flawed and the results are invalid.

In gaming the more powerful the GPU The more load it will put unto the CPU,

So in 2-3 years when you upgrade your GPU your CPU will run at a higher load in games and you could find yourself in a situation where your CPU overheats if you listen to Linus's "advice".

[BlueChinchillaEatingDorito]

If Intel really wants to stick to a flower style cooler, can we at least have something like the old MSI Cyclone II so cooling is not absolute garbage? 

 

[GeorgeMKane]

On 5/4/2023 at 12:23 PM, AdamFromLTT said:

No matter where you go, you’ll see people debating the amount of cooling you need to make sure your CPU runs fast. With power hungry GPUs and computer cases that go a little too heavy on tempered glass it is important to make sure that your Intel Core or AMD Ryzen processor can run its best. But cooling can be expensive. So how much do you really NEED? Especially if you’re just gaming? Less than you think.
 

 

The Vetroo V5 is good for up to 12th gen intel and 5th gen AM4 cpu's. Outside of that, Noctua or AIO's all the way imo. Big advantage of air cpu coolers is that they typically have less worry and less maintenance.

[Bitter]

I mean I don't NEED a giant air cooler on my 2700X, the stock Prism cooler did fine but it was LOUD and wouldn't keep the temps low enough to let it self boost to it's potential. My giant air cooler is quieter running full speed than the Prism was and it keeps the hot spot temp low enough to allow the CPU to boost as high as it possibly can. 4ghz all core pulls some power and makes some heat, the stock cooler could only hit 3.6ghz for sustained work loads and was only good to shed about 130W of heat. Giant air cooler can shed 170W comfortably and allow higher clocks at lower temps with less noise. Seems like a win all around, but I'll concede that for gaming the stock cooler would have been fine.

[dogwitch]

4 hours ago, Bitter said:

I mean I don't NEED a giant air cooler on my 2700X, the stock Prism cooler did fine but it was LOUD and wouldn't keep the temps low enough to let it self boost to it's potential. My giant air cooler is quieter running full speed than the Prism was and it keeps the hot spot temp low enough to allow the CPU to boost as high as it possibly can. 4ghz all core pulls some power and makes some heat, the stock cooler could only hit 3.6ghz for sustained work loads and was only good to shed about 130W of heat. Giant air cooler can shed 170W comfortably and allow higher clocks at lower temps with less noise. Seems like a win all around, but I'll concede that for gaming the stock cooler would have been fine.

yeah a basic air flow case and a 30 buck air cool would work great.

[Ubersonic]

It's not technically a waste when you fully understand you're paying extra for less noise and other bonuses over cheaper coolers and not more performance.  Especially with a Noctua as you know it's going to be getting future mounting kit upgrades as long as the company is in business.

[LinusFanNumber 14.8M]

Oh and i3 13400 also.

[Daniel_Aldaco]

It may have passed over my head, but I was thinking of starting a PC build using the AMD Ryzen 9 7900X alongside either the AMD RX 6700 XT or the Intel ARC A770 gpu, for, at most leisure 1440p gaming. Would the Peerless Assassin be sufficient to keep the CPU within 91 C for maximum operation? The case I planned on using was the Corsair 4000D Airflow case. Any help is greatly appreciated.

[LMGcommunity]

On 5/6/2023 at 11:42 PM, Daniel_Aldaco said:

It may have passed over my head, but I was thinking of starting a PC build using the AMD Ryzen 9 7900X alongside either the AMD RX 6700 XT or the Intel ARC A770 gpu, for, at most leisure 1440p gaming. Would the Peerless Assassin be sufficient to keep the CPU within 91 C for maximum operation? The case I planned on using was the Corsair 4000D Airflow case. Any help is greatly appreciated.

This may be a better question to ask our New Builds and Planning subforum.

[yungjerky]

On 5/5/2023 at 4:28 PM, Vishera said:

@LinusTech The testing method is flawed and the results are invalid.

In gaming the more powerful the GPU The more load it will put unto the CPU,

So in 2-3 years when you upgrade your GPU your CPU will run at a higher load in games and you could find yourself in a situation where your CPU overheats if you listen to Linus's "advice".

You clearly didn't even make it 3 minutes into the video or else you'd know they tested with the 3090ti to put max strain on the CPU's... All of these CPU's are already running at 100% during the whole test so upgrading the GPU won't change the result.

 

Here's the clip for ya, starting at 2:22. Pay attention:

 

[fillsky]

Hi. Kinda late to the party.

This is my first time writing a post even though I have been watching LTT for probably 11 years.

I would not recommend buying a NHD 15 for a 13700K. I switched from an 8700K OC@ 4.7GHz using the same noctua (they provided a mounting kit for the new rig). GPU: RTX 3080 and Chassis: Lian Li Lancool II. To my surprise in the stress test it was unable to keep the 13700K sufficiently cooled after even a small OC. It's fair to say, in that case, do not OC. I would also say not to buy the K version then.

After the fact, I read some topics and found out that the most the NHD 15 can handle is 180W. 13700K is 253W MTP. At the same price point you can get an AC Liquid Freezer II that can tame this beast.

I'd like to see some insight into how they measured or why they chose the 13700 with K and didn't take into account the thermal overhead available.

Obviously in games it doesn't matter (at least for now) and is fine, but when running some heavy workload, it lowers the clocks. 

 

[sqrt443556]

At 15:05, the graph shows the 600 second stress test ran at an observed temperature target of 95c.

Without claiming responsibility for anyone's actions, and only as a conversational reply, is this safe?

 

Spoiler

We've all heard that 80c is safe op temp and that tHeRmAl ThRoTtLiNg happens at 100c so it's ok but no, it's not, right?

 

Well I wonder. If the hardware thermal limit is 100c, is running at 95c for extended gaming/streaming periods (8hrs or so) consecutively okay for the health of the device? What is the measurable difference in running a device long term at 80c v 95c? Is this something labs can look into? Might be like dust-server or dust-pc video series?

 

Tldr why come 95c gpu oog oog say fire scary hurt gpu?

 

Thanks.

[AdamFromLTT]

15 hours ago, sqrt443556 said:

At 15:05, the graph shows the 600 second stress test ran at an observed temperature target of 95c.

Without claiming responsibility for anyone's actions, and only as a conversational reply, is this safe?

 

  Reveal hidden contents

We've all heard that 80c is safe op temp and that tHeRmAl ThRoTtLiNg happens at 100c so it's ok but no, it's not, right?

 

Well I wonder. If the hardware thermal limit is 100c, is running at 95c for extended gaming/streaming periods (8hrs or so) consecutively okay for the health of the device? What is the measurable difference in running a device long term at 80c v 95c? Is this something labs can look into? Might be like dust-server or dust-pc video series?

 

Tldr why come 95c gpu oog oog say fire scary hurt gpu?

 

Thanks.

Ryzen 7000 series is specifically designed to boost as high as possible until you reach the 95 degree limit (or 90 for X3D chips). AMD says it is fine to run at that temp for the lifetime of the product.

The actual hardware limit for temps on Ryzen 7000 are 115 degrees, which you can reach by manual overclocking. Any hotter and the CPU will immediately shut down to protect itself.

In terms of testing this, it would basically be impossible because we would need a sample size of hundreds to thousands of CPUs and we would need to run the test over, like, 7 years. AMD can do that. We, sadly, cannot.

[RavetcoFX]

Does anyone know why this video was privated?

[LMGcommunity]

9 hours ago, RavetcoFX said:

Does anyone know why this video was privated?

It was made private while we review the accuracy of the data in it.

[Skipple]

2 minutes ago, LMGcommunity said:

It was made private while we review the accuracy of the data in it.

Regardless of the outcome, it should probably remain private given the controversy surrounding it, for your own sanity, if nothing else. 

[mMontana]

5 minutes ago, LMGcommunity said:

It was made private while we review the accuracy of the data in it.

When the video of keyboard and dishwasher will be inaccessible?