Bitcoin Forum
June 23, 2024, 04:00:54 PM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: [1]
  Print  
Author Topic: GPU health over time... higher temps or higher fan speed?  (Read 1665 times)
kisk (OP)
Full Member
***
Offline Offline

Activity: 126
Merit: 100


View Profile
March 20, 2017, 02:41:20 PM
 #1

As the subject suggests.. over time is it better to bump the target temp up a little bit to have lower fan speeds, or lower temps with higher fan speeds?

All of my cards are RX480s -- 3 of them being ASUS cards which I hear have a higher fan failure rate. For the life of my rig I've always had them set to 70c with 3000rpm max. This has not been too much of a problem in the winter, however now that it is getting hotter I'm noticing my fans are staying near my set max (3000rpm) to keep the 70c.

I just set the target temp to 75c and that dropped my fan rate by about 600rpm. I'm sure 75c is fine but what about 80c on a constant basis?

Any pros/cons on 70c vs 75c vs 80c short -or- long term?

Thanks!
adaseb
Legendary
*
Offline Offline

Activity: 3794
Merit: 1723


View Profile
March 20, 2017, 02:46:48 PM
 #2

I got some GPUs running at 90C for the last 3 years non-stop.

While others ran at 60C and died after a few months...


Either way, you got full warranty for 3 years with ASUS. Anything happens, send it in.

General consensus is that its better to run components at lower temps because it increases component life, like the capacitors but most likely you will stop mining before they start failing.
kisk (OP)
Full Member
***
Offline Offline

Activity: 126
Merit: 100


View Profile
March 20, 2017, 03:58:35 PM
 #3

I got some GPUs running at 90C for the last 3 years non-stop.

While others ran at 60C and died after a few months...


Either way, you got full warranty for 3 years with ASUS. Anything happens, send it in.

General consensus is that its better to run components at lower temps because it increases component life, like the capacitors but most likely you will stop mining before they start failing.

Thx for the info. I suppose I'll let it run at 75c for now. This is a 7gpu rig, and as far as me stopping mining before failure -- I don't see that happening unless mining just totally becomes unprofitable. I plan on building rig #2 and #3 here in about a month Cheesy Since I started mining back in October I've become quite fond of crypto.
RentGPU
Sr. Member
****
Offline Offline

Activity: 661
Merit: 258


View Profile
March 20, 2017, 04:20:28 PM
 #4

I got some GPUs running at 90C for the last 3 years non-stop.

While others ran at 60C and died after a few months...


Either way, you got full warranty for 3 years with ASUS. Anything happens, send it in.

General consensus is that its better to run components at lower temps because it increases component life, like the capacitors but most likely you will stop mining before they start failing.
90c is way too risky, seems like 3 years means you passed this risk, but not everyone can pass it , i mean it's more safe and less risk to put some load on fans than burning everything else , fan can be replaced ,but good luck replacing everything else...

2016 GPU Miner
ccccccc7
Full Member
***
Offline Offline

Activity: 223
Merit: 100


View Profile
March 20, 2017, 05:04:08 PM
 #5

I got some GPUs running at 90C for the last 3 years non-stop.

While others ran at 60C and died after a few months...


Either way, you got full warranty for 3 years with ASUS. Anything happens, send it in.

General consensus is that its better to run components at lower temps because it increases component life, like the capacitors but most likely you will stop mining before they start failing.

I will use high fan speed to reduce the temperature. Then I will replace the fans.

+1

Or just remove the stock cooler and zip-tie high airflow or server fans to the heatsink, this provides exceptional cooling - also keeps the fans clean for resale  Grin
kisk (OP)
Full Member
***
Offline Offline

Activity: 126
Merit: 100


View Profile
March 21, 2017, 03:09:28 AM
 #6

Good idea on using different fans.

Another question: Is there a reason why I keep seeing rigs with the cards so close together? I could maybe understand if you're using a specific card with a block-type body with a specific inlet and outlet for air.. but most cards I've seen have open sides and the airflow isn't completely blown out of the top. Makes no sense why they're placed so together unless its just a space issue, which rarely is the case. I mean, what's an extra 6-12 inches for a single rig?
Qazo
Full Member
***
Offline Offline

Activity: 211
Merit: 101



View Profile
March 21, 2017, 07:05:05 AM
 #7

I was cracking my head with same question as you and then I tried to build my own rig case. You know why you must put cards so close to each other?

- lenght of riser cables.
Tasty Potato
Full Member
***
Offline Offline

Activity: 231
Merit: 100



View Profile
March 21, 2017, 06:25:23 PM
 #8

I was cracking my head with same question as you and then I tried to build my own rig case. You know why you must put cards so close to each other?

- lenght of riser cables.

Nah, that is hardly an issue. One can place 6 cards 'in one plane' with fans on the bottom (like in PC case), likely in chess pattern but this design will be x2-x3 more spacious in general. And 60cm risers are enough. Hell, maybe even 30cm too.

It needs to be tested if different placement of cards makes difference temperature wise, but most are just so conservative and build 'standard rig cases'.

What I have in mind is even adjustable, too bad it is only in my head at the moment ;p but it is wood and metal with nuts -,-
Lauren Smith
Sr. Member
****
Offline Offline

Activity: 1050
Merit: 277



View Profile WWW
March 28, 2017, 09:05:53 AM
 #9

I wouldn't worry. Gpus are ment to last and they do. Gamers push them hard while they play so they need to be able to handle that. As long as you keeping the temperature down you are fine.

dagarair
Sr. Member
****
Offline Offline

Activity: 847
Merit: 383



View Profile WWW
April 07, 2017, 12:26:34 AM
 #10

I ride mine like a 5$ whore.   Tongue

I like in texas so keeping them cool is tough in 105 degree summer day.  as long as you have killer ventilation they stay ok in the 70's

4MW Data Center - I BUILT Tongue  - Full story below:
https://bitcointalk.org/index.php?topic=4789787.msg43227027#msg43227027
Raja_MBZ
Legendary
*
Offline Offline

Activity: 1862
Merit: 1505



View Profile
April 11, 2017, 09:22:28 PM
 #11

LOL, its a great question. Was always confused in making the decision, but never thought of asking it out here. Grin

For the last like one year, I've been running my GPU's at almost 90% fan speed with 78 degrees temperature. 90% fan is way too much by the way, I can't keep the rig in the room where I'm staying. But anyways, they're still running smoothly.

I got some GPUs running at 90C for the last 3 years non-stop.

While others ran at 60C and died after a few months...


Either way, you got full warranty for 3 years with ASUS. Anything happens, send it in.

General consensus is that its better to run components at lower temps because it increases component life, like the capacitors but most likely you will stop mining before they start failing.

That helps, will lower the fans rate right away of my all rigs. 90 C seems risky, but will still try out the speed of fans till temperature is somewhat like 85.
RentGPU
Sr. Member
****
Offline Offline

Activity: 661
Merit: 258


View Profile
April 12, 2017, 09:33:22 AM
 #12

LOL, its a great question. Was always confused in making the decision, but never thought of asking it out here. Grin

For the last like one year, I've been running my GPU's at almost 90% fan speed with 78 degrees temperature. 90% fan is way too much by the way, I can't keep the rig in the room where I'm staying. But anyways, they're still running smoothly.

I got some GPUs running at 90C for the last 3 years non-stop.

While others ran at 60C and died after a few months...


Either way, you got full warranty for 3 years with ASUS. Anything happens, send it in.

General consensus is that its better to run components at lower temps because it increases component life, like the capacitors but most likely you will stop mining before they start failing.

That helps, will lower the fans rate right away of my all rigs. 90 C seems risky, but will still try out the speed of fans till temperature is somewhat like 85.
I can't understand this temps you are talking about , my cards are always sitting at 55-60c and i live in egypt which is african country Smiley, where do you live?, i never reachs those temps , even when i dual mine my temps reachs 65 max.

2016 GPU Miner
bel123
Member
**
Offline Offline

Activity: 97
Merit: 10


View Profile
April 12, 2017, 02:48:08 PM
 #13

I got some GPUs running at 90C for the last 3 years non-stop.

While others ran at 60C and died after a few months...


Either way, you got full warranty for 3 years with ASUS. Anything happens, send it in.

General consensus is that its better to run components at lower temps because it increases component life, like the capacitors but most likely you will stop mining before they start failing.

Can you share your experience about GPU which work for you (90C) and which no work (60C and die fast)?
matmator
Sr. Member
****
Offline Offline

Activity: 487
Merit: 252

bou !


View Profile
April 12, 2017, 02:57:25 PM
 #14

It depends mainly on the gpu, the reference rx480 heat up a lot! Cooling not efficient and very noisy, on average I am a 75 c with my reference,

But with my nitro rig, is that happiness 60c or 65c max.

go to the moon !!
Keelhaul
Member
**
Offline Offline

Activity: 67
Merit: 10


View Profile
May 10, 2017, 10:51:12 PM
 #15

I try to keep both stats neat. Like having 66 degree from 40% CS intensity.
philipma1957
Legendary
*
Offline Offline

Activity: 4158
Merit: 8054


'The right to privacy matters'


View Profile WWW
May 10, 2017, 11:11:09 PM
 #16

 65c temp 65% fan is really good

but 70 c   70% fan  is still good.

if you need to do more then 80% fan  to stay under 75c  you  driving cards too high.

▄▄███████▄▄
▄██████████████▄
▄██████████████████▄
▄████▀▀▀▀███▀▀▀▀█████▄
▄█████████████▄█▀████▄
███████████▄███████████
██████████▄█▀███████████
██████████▀████████████
▀█████▄█▀█████████████▀
▀████▄▄▄▄███▄▄▄▄████▀
▀██████████████████▀
▀███████████████▀
▀▀███████▀▀
.
 MΞTAWIN  THE FIRST WEB3 CASINO   
.
.. PLAY NOW ..
Metroid
Sr. Member
****
Offline Offline

Activity: 2142
Merit: 353


Xtreme Monster


View Profile
May 10, 2017, 11:12:36 PM
 #17

I recommend leave all the fans of all cards at 90% cause you can get the best under voltage possible + more clocks on memory and core, meaning less energy and less to cool down, trust me that is the best approach to do and make sure they have cool air, so an air extractor is a must whenever your computer is. The idea is to balance everything except the fan which must stay at 90%, remember and take note, that place will be very noisy, it must not be a place you go often, watch out for your eardrums.

BTC Address: 1DH4ok85VdFAe47fSVXNVctxkFhUv4ujbR
vapourminer
Legendary
*
Offline Offline

Activity: 4368
Merit: 3658


what is this "brake pedal" you speak of?


View Profile
May 10, 2017, 11:23:49 PM
 #18

i prefer running fans on the higher side, 70-80%. cooler cards are happier cards.

did lose a fan on an asus 6870 direct cu back in btc mining days but all my other cards fans have survived, some for over a couple years at 24/7 use.
Vann
Hero Member
*****
Offline Offline

Activity: 1036
Merit: 606



View Profile
May 10, 2017, 11:53:18 PM
 #19

The fans are typically are what goes out first. Even if it's covered under warranty you still deal with the downtime. So IMO, it's better to keep the GPU fans as low as possible and use external high volume case fans or a box fan to cool.
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!