they kinda do, especially the gen 5 ones (on the controller) but hard to do much about that
all components have started pushing the power consumption up bit by bit
1 Like
Because in most desktops motherboards NVMe 5.0 slots are stupidly placed 3 mm above main 16X PCIe slot where GPU gets plugged. So imagine 300W GPU hot air blowing on that NVMe heatsink.
In laptops some of this 5.0 NVMe not even have a heatsink, or is just a 0.3 mm Al foil painted black 
oh sorry not painted “Anodized” cause both are enhancing the cooling.
Fans got Rotors. You put the question wrong. How many stops and starts can a rotor sustain before breaks. And is way less than you think. How many spins can do before brakes? Millions, with one condition you don’t start and stop it that often. The high usage of the fans are not the spins count but, the stops and the starts counts?
Ask an engineer your self. I did that already. 2 of them, hydraulic engines engineer and automotive engineer.
I worked with servers and they have fans running all the time, the only fan that breaks was on the offices were people smoke, was more dust and more humidity. All the fans was working all the times none of them stopped start doing to many db and than needs to changed.
No, they will make unbearable noise and than you’ll want to change them.
What wares fans the most are repeatedly stops and starts of the rotors.
A fan consumes 1.8-2.5 W rarely 3 Watts 3x2 Watt(GPU)=6 Watts. But max of 3 W will be achieved by a fan 2.5cm thick . GPU got usually 1cm -1.5 cm thickness which means consume less.
This for fans 92-100mm thickness which is the kind of diameter you see on your GPU fans.
On low RPM ~ 400 you have around 1-1.2 Watt consumption/fan.
So where is the power consumption you talk about? Think. Your card consume maybe 250Watts how much matter another 7 Watts in this equation?
Isn’t wiser to cap the game in NCP and get 70 watts less on your GPU than stopping the fans which consume 7 Watts?
I was talking about desktop cards clearly. Than who’s fault it is?
The card does that if you don’t tweak it. At 50 % load even the GPU is at 51 C and VRAM at 76 some GPU not even starting the fans. Seen that on 1660 Ti client.
Another one, 3070- card at 49C GPU, Hotspot 68 C- VRAM 70 C >>>>fans 0 % than it started fans for like 2 min when GPU reached 55C and than stopped again the fans.
2080 similar with 3070
6600XT similar with 1660 - VRAM was 72 only.
I’ve see cards starting and stopping the fans 6-7 times in 20 minutes. That my friend will kill you fans.
All cards that(like 50% in the office) was having stupid noise from their fans- whining, growling, buzzing. I deshroud them, put case fans and built fan curves(without stopping the fan ever)- those fans never developed those noises, one got over 4 years now. The temps was much better on the GPU with less noise. 25mm vs 10 mm fans they push much more and quieter. Brands- Asus, MSI, Gigabyte, PNY. In 3 instances PNY lasted longer with with the stock fans.MSI and Giga the worst.
You don’t say.
NVMe 4.0 runs around 40 C, that is not hot. Rarely they reach 50C when copy hundred small files.
So what exactly are you talking about?
If NVMe runs too hot that is on the MBO manufacturer, stupid design. They just placed NVMe above or under the main PCie 16x where is the GPU. And the GPU fins orientation is just towards your NVMe slot blowing 200-400Watts heat on them.
Not everybody got modern GPUs. A lot of people still runs 1000 and 1600 also low level entries like 3060 series from Nvidia and counters from AMD and those will be taxed heavily by AOE 4, that video is mostly for them but, also for those who want to reduce power consumption of the GPU or just want lower temps.
You can’t deny people cards are affected by high temps in AoE4, just because you have a more powerful card and have a different experience doesn’t invalidate their experience of a sizzling GPU.
Just remember GPU and CPU throttle down, VRAM does NOT! Will fry - nobody wants a dead GPU now days with this GPU prices.
My card does very good for a 2017 card and runs 2K, but got deshrouded to reach those good temps without noise.
Yeah, no, that was exactly my point. They’re proofed to much higher temps and can handle the motherboard placement you’re criticising (especially with a decent cooler).
By decent cooler meaning that aluminum anonized heatsink- how will that suffice to the heat from 200-400 watts card hot air coming towards it? It won’t. Look at the orientation of GPU rads now days it blows hot air on the sides(one side being towards the heatsinks of NVMe), not outside the case, unless is a blower card.
I haven’t had a single problem with my Crucial NVMe drive I’m using as a boot drive and the 4070 Super I recently installed on top. Stock cooler on the NVMe as well.
Heck, the main problem was fitting the Super in the case. That and being surprised it worked first time. My cabling skills are atrocious.
And your NVMe the temps are?
Also transfer hundreds of file on to your NVMe for 1hour or so and we talk about your temps after.
That NVMe suppose to handle big raw pictures, films and what not for studio works, and it won’t specially if is 5.0 and is positioned above or under GPU I’ve seen them going around 65 C easy and is throttling than.
Just booting in windows are fine. But NVMe are not only for that, like any other drive suppose to hold onto extensive IO operations and often without craping the speeds.
For those who still doesn’t understand the meaning of my video or my tweak or they like to troll some more
I’ll give them another opportunity
I’ve redone the test this time with max shadows as well to be no more silly discussions why I’m using no shadows.
Maxed out setting AoE4 1440P - 1080 Ti, max power draw for this card 290Watt
1vs1 -destroying enemy base - fire, smoke, effects, screams and all that
So is simple as that, you don’t cap AoE4 in Nvidia Ctrl Panel this card draws 88 Watt more and is 18 C hotter.
Please give me one reason why I shouldn’t use this tweak or show to other how to??!!
This works for any card, soon I’ll test it on an Asus OC 1070. I just tested 3 weeks ago on 4060(only 45 Watt less but still)
Buddy, it’s my boot drive. It sees very substantial file activity generally.
Within the proofed temperature range for the drive. Currently 55C.
Nobody has ever said you couldn’t, just that 55 - 70C is in no way hurting your card.
My last GPU lasted until the thermal paste wore out, at which point airflow in my case, or the demands of any video game I was playing, stopped being a concern (because any activity was enough to raise the temp, nevermind playing AoE IV).
Hey there,
It has many factors and I see you have 1080ti which is a good graphics card.
The card is simply getting older and might be affecting the performance.
Also, you need to maintain that card and apply thermal paste at least every 12 months.
Turn off the Anti-Alliasing option in the graphics settings. It blurs and drains performance.
Additionally, turn off V-sync. This is the most useless tool to use because it simply drops all your FPS.
Lastly, try to change your resolution to 1080p. Older graphics cards have huge issues with higher resolutions.
Hey there,
Thanks for your input.
Don’t know what factors.
The card is old but, got a memory bus of 352 bit that’s why can handle 1440P very well. Better than 2060 and 3060 cards with 128 bit bus. I had cards with 256 Bit bus in 2007 I don’t see my self buying now days(17 years later) 128 bit cards specially when I’m running 1440 P.
My card was repasted recently.
As you can see from the test she will stay around 37 - 43 C on the GPU which very good. Will stay like that in 4vs4. 43c on GPU - means around 60 C on VRAM which again is very good.
Yes, card was deshrouded and I’m running the fans all the time, I even have a 4th 92mm fan pulling air from the back of the card towards top of the case running ~ 900-1000 RPM.
I will not change resolution to 1080p, I love my 1440P 
But I will not run AoE4 uncapped. Period. And I 'm not advising nobody to do it.
Even If I’ll get a 4090 today I will still cap AoE4. Don’t see why would I waste watts/hour - here the electricity ain’t cheap. With 88 watts I can run easy another screen and youtube on it ,why flushing down that power down the drain.
First of all AoE4 is an RTS and is no point in running it over 60 FPS.
Secondly I have no reason whatsoever to run the game with my GPU 18 C hotter and pulling 80 watts more.
However the video is addressed to people that doesn’t go to deshrouding their GPU, not even repaste their cards cause their are afraid of disassemble their GPU, got dust collected in their radiators, negative pressure in the case, case fans configured to ramp up only when CPU is hot ignoring GPU temps, exhaust fans in the case barley reach 1000RPM, not enough to pull in time the heat from GPU.
That is the usual picture in prebuilt I see often.
So, are a lot of people having their GPUs around 80 -85 C after 1 hour in AoE4 all other games from various reasons.
The clients above had those temps in 10-20 min test in various games including AoE4.
in 1 hour gaming most of them reached 85-88 C on the GPU - which resulted in VRAM being around 100 C and above. Always VRAM is hotter than GPU with 15-20C - is a fact.
On VRAM 100 C is critical specially on Micron and Hynix chips, Samsung chips tend to do a bit better on the heat stress.
The worst chips being Microns produced in 2018 found on 1000 and 2000 cards and AMD counters.
If ONE single memory chip gave up on you because of prolonged heat stress you gonna have dead card.
VRAM temps are the main reason here. Cooling them properly will do give longevity to your card.
1 Like
No.
You just can’t compare windows logging, updating, telemetry, security with copying files for hours at 2000-5000 mb/sec. Is plain unrealistic.
Transfer hundreds of files, like pictures, films, zips for an hour or so - that is a good test I can promise you.
None of NVMe used as boot heat up on my workplace only those with a lot of file transfers. Samsung PRO, ADATA, Crucial, Kingstons - those with file transfers not the windows drives. Those are reaching 65 C. A lot of those with high temps are arund GPU positioned.
Is yours a PCIe 5.0? and as you said is underneath GPU right?
Mine is 4.0 and is away from GPU as you expected.
My fastest slot is above GPU 5.0. Anyways I don’t need that fast PCIe for now so went for 4.0. Even that I like Crucial went for Samsung 970 EVO + (better offer)
As you can see 34 C average.
Yes, nobody. But that doesn’t mean I should run it with 55C when I can with 38 C? 55C on my GPU will mean more heat in my room. Why would I want that in summer?
Maybe I can run it at 60 C in the winter and warm up my room. 
More wasted energy also. Why? I didn’t even measured wasted energy on the CPU and add it up to the GPU wasted 80 watts.
We can run AoE 4 more Green, ECO and less CO2… mmm the cows are still farthing though and melted away the glaciers. Soon water cooling for free 
With respect, you have no idea what my I/O usage is like. Trust me when I say it gets a battering. I can give specifics but outside of specific software disciplines it’s mostly just a waste of words. A lot of the traffic is in AppData (which more and more Windows apps are using as per MS guidelines - direct use of Documents is in the past, especially since OneDrive came in).
My old SSD is still serving as a secondary (games) drive (for games that really need the read / write). Crucial’s software says it’s still healthy.
Nobody’s telling you whether or not you should do anything. It’s entirely the end user’s choice.
Information is good. You providing it is more than fine. Telling people how to run their PCs, assuming everything is in a healthy range, is less good. But I’ve said all this before, I think.