Bitcoin Forum

Hi, I'm new to mining and I started mining just yesterday with my single 3070 founders edition in HiveOS. I wanted to know a few things but before that have a look at my settings:

https://i.imgur.com/pu83Vnd.png

https://i.imgur.com/UiYZVcM.png

Now what I want is to my card last me forever (long enough for it to lose its value and become obsolete) and for it to last that long I heard keeping it cool and making fan not running at 100% is the key because I don't want to lose my fan either. I don't really mind if its not running the maximum hashrate it can get. I'm not a professional miner, I'm just doing this as a hobby and that's the reason I want my card to be safe. Could you guys please look at the charts and tell me what needs to be changed? What is the best core temperature and best memory temperature for the card for keeping it safe? what's the best autofan settings to keep my fan safe?. Any additional tips and tricks are always welcome. Sorry if some of the question seems stupid, this is my first time mining and I learned what I could from youtube but it doesn't explain everything. Thank you very much.

Do 35 to 90 %

See if the miner has a 85C shutdown just in case of a fail of some kind.

and set card to run at 70c

U mean fan speed? min 35% and max 90%?

yeah

Start with high quality equipment, you'll need to pay a little more.
Don't push the PSU, leave lots of headroom.
Control the environment as much as possible, good airflow, low humidity, low dust.
Keep everything clean and free of any dirt buildup.
And what Phil said.

Don't overclock the memory clock too much, running at a high memory clock can bring memory errors in future, meaning memory loose due to high memory temp, the lower the temp the better, the highest I've seen my graphic cards go is 60°C and I tend to keep it this way

I can't see the uptime of your video card.
If the core temperature of the video card is 45-50C and the fan operating mode is 50-60%, then you have nothing to worry about for 1-2 years.
If the video card has a manufacturing defect, then you will find out about it in the first months of mining.
Dust off every 3-6 months.

Sooner or later the fans will start to fail.
I have cards that run for years and some not.
Zotac 1070 Minis were a nightmare.
Go for cards with a big heatsink and 3 Fans.
Temperature is a really big point.
Old founder editions with 1 fan never failed but have got the highest temperature.
My 1080 rigs run now for....mmmmh when did they come out? 2016 ? Still running, watercooled since 1. day.

ZOTAC mining fans suck on many of their models.

gigabyte has shiitty fans unless it is the AORUS

Asus Dual model cards suck. I have always had poor luck with them.

Of the three above I still buy the zotac minis and always keep some spare fans for them. sometimes that small size is hard to resist buying.

I never buy asus dual or any gigabyte cards

I probably have the same card, exactly the same hashrate, but 115 watt power consumption, 5 Celsius  less, fan at 37 %  but running in windows, because I was unable to fine tune power in Hiveos.
 

The highest failure point on most GPUs are the fans. And there is not much you can do with them. It’s also hit and miss. Sometimes you got 1 GPU and running fan at 80% and it lasts 5 years and still alive. While others you use for 6 months and fans are dead already.

With mining I think part of the reason is the way the fans are positioned. In a regular PC they are normally horizontally while in a mining rig they are vertically. And I just think that horizontally mounted fans last longer because the fan is more balanced since gravity is not pushing it down. Just something I heard.

I remember something similar about hard drives many years ago, but a hard drive has more mass which
is harder on the bearings. I don't think it mattters much for a GPU fan, the quality of the fan is probably
more important.

Maybe I've been lucky but I've had no fan failures in 7 years, but then again I don't have any of the brands
mentioned in a previous post.

And to bring this back on topic, that's why I recommended high quality equipment for a longer life.

I have more than 6x zotac mini cards though with two small fans onboard and I have no problem with overheating, they hardly hit 60°C that is if the sun if very sunny every afternoon, mainly the cards operates around 55°C to 57°C

@Philipma1957

Same here.


@Skinny48

I talked about the fan quality of the 1070 Minis. They fail first of all 1070.
Next than is gigabyte....
Not to buy....

Might have changed with the 3xxx series. But we will see in 1 year.
Anyway i buy Zotac only to watercool them. The new Zotac 3070 is perfect for this.
 

Rule of thumb, I have been mining for almost 4 years with 120 plus cards. Lot`s of expensive lessons learnt!

1. Power supply, never push your power supply to it`s limit. Sweet spot is around 60% utilization.
2. Dust is a card killer, make sure you maintain dust as well as you can
3. Temperatures, place your rigs in a well ventilated room. Try remove the heat generated by the cards (Extractor fan).
4. Don`t be an IDIOT and OC your cards to the max. Loads of people are chasing hashrate instead of finding a sweet spot.. pushing a card to just under the point where it crashes is really idiotic. Your card might cope for a couple of months but it will die on you after a while.

Cheers

Yup, what he said. plus my post

After 4 years of mining on many video cards, fans have broken.
This is what most of my graphics cards look like now.
Some have 2 or 3 external 70-90mm fans
https://i.ibb.co/rs1rNRr/image.jpg (https://ibb.co/Hdk83f8)
A video card with a Palit Dual cooling system had an anti-record, the fans worked for 18 months

Whats ur memory like? maybe I'll try ur setting and see if it works better?

yup all high grade equipment's and stuff, got rm 1000x not gonna run more than 3 cards on that (if i manage to get 3 that is lol) leaves me tons of headroom

yup planning to do dusting every 1 month or so

its in my bedroom thats the problem and its under my desk so maybe heat will gather. I need to find a extractor fan but very silent ones

nah I aint gonna risk my card like that not im running it at 50 degree locked at fan around 50 both stays the same fan goes down to 35-40 so yeah I'd rather get 3-5 less hashrate than just ruin my card

you are on the same core and memory setup that I used when I was in Hiveos.
Now my memory is down to 1258,(windows 2516 in Hiveos) because the sweet spot was there . PL etc obviosly have changed

I have these type of msi brand GPUs I would like to know what fan speed you've been running them for those years before the stop working, I'm presently running mine at 72% and I think that's safe

If you are going to replace a fan with a case fan then you should remove the dead fans out of the way first. Best is if you remove the entire shroud actually. However depending where the screws are you might void warranty that way. However a cheap case fan usually even 1 can do a better cooling job than the regular fans.

Only issue with this is that you will need to spread your GPUs out more because they will be fatter. And your usb riser cables might not reach the motherboard this way. You can buy a longer cable however if the cable is too long they will usually cause stability issues.

for nvidia gpu no mod on memory applied only oc-ed the memory clock so its that safe for long term and dont mind for the fans, you can replace it for cheap

dont forget change thermal paste each year and cleaning the gpus twice a year

It depends on the quality of the fans. In winter, my fans mine at 60-70%, in summer it can be 80-100%.
I'm not even going to repair the cooling system when I sell graphics cards. The disadvantage I compensate the buyer with a lower price.
Major fan problems begin after three years

Look at the photo. The distance between video cards is 12-15 centimeters.
This does not greatly affect the cooling result.
It is enough to buy a fan with the maximum amperage 250 mA (  0.25*12= 3W)

Totally agree and don't push the card too hard. A little loss in performance is fine and running 7x24 trouble-free could be more important.

Played with ~10 1080Ti in past years, either founder's edition or Auros. It's a surprise that founder's edition seemed more stable in a mild OC with limited TDP/fan-speed, but + 10 degrees hotter (still under 80C).

Yes because you only got 2 GPUs. However try and do this on a rig with 5 or 6 GPUs. The width of the fan will be a problem because they will need to be seperated more and more and your issue will be that you won't have enough slack in the USB cable.

I had this issue and basically built a multiple level rig where one GPU was higher than the other and I got away with keeping them close by to not overheat. The easiest solution was to buy a longer USB cable HOWEVER for some reason the longer cables didn't work. They sometimes worked but eventually got hardware errors OR they would randomly crash. I think the longer the cable the more issues you will get.

Undervolt the cards and don't run them above 70c for long durations.

also it looks like you've got some sort of MSI or other terribly heat piped card..

Make sure you have airflow over those cards... Them heatsinks are dinky af for mining with. :/

Order some extra fans for them it is not expensive anymore. Also would recommend cpu fans with 4pins that way card can adjust rpm and cooling very good. i spend alot time ago when 1 fan was 20-40$ back then. i replaced with cpu fans.