- Joined
- Feb 21, 2008
- Location
- Baltimore, MD
So what happens when you need to hard reboot one of them? Gotta climb up there to do it? Lol
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
summer is coming
- Thread starter wagex
- Start date
- Joined
- Nov 26, 2004
- Location
- 1337 Leet Street , CO 1337
lol, i am working on a little box with some toggle switches and have all the power/reset jumpers in it @ ground level
![(G{in}[AK)TION]](/forums/data/avatars/m/116/116613.jpg?1469679448){kind=avatar}
- Joined
- Apr 14, 2011
i fanboy'd at the final touches. well done moho.
- Joined
- Nov 26, 2004
- Location
- 1337 Leet Street , CO 1337
thx
sure
heres some of my major points:Coins:
currently on vertcoin, thinking of trying out that scrypt-n multipool. was on hashcows/middlecoin for quite some time, directly converting to LTC, which i am currently about 90% in. got in early on vtc, made a nice profit from it, hope it goes back in that territory
Always watching betarigs to see if there is a high demand, made some nice profit there as well. Always on the lookout for new coins, I have between .1-.75BTC(mined, i dont think i ever actually bought any of them) in about 15 different coins just hoping for them to rise(ex: galleon, Aiden, Roto, Skein, Heavy, Max, Proto, Caishen, and a few other i have the wallets backed up somewhere )Rigs:
most are windows 7, with CGminer, and CGremote as well as teamviewer. the add pool option in cgremote saves me lots of time when i try out new coins. was running BAMT as well, but i killed like 6 flash drives, and had a lot of problems getting -coldreboot to work, so i just went back to windows. just got a $30 16 port hub, attached to my router, works MUCH better than wireless USBs(for some reason i would always see it go offline/online for no reason at all(the walls are 6" cement could have something to do with it))
Power
power---this is one of my biggest problems. I can get about 2 rigs (~2500w) running on 1 breaker. so with these i need a total of 10 breakers. I just bought some 50' 13 gage 15A extension cords from amazon, they are really thick and great quality. I try to split my psus up to pull max 1300w from the wall(using a killawatt), otherwise the cords get really hot to the touch.
ALWAYS make sure you have a surge protector between the wall and the rigs, as if you have none, it can melt the cord(s) easily, i have seen it on a couple of my rigs, not fun. I have had 5 or so PSUs fail, 1 rosewill, 2 OCZ, 1 Cougar, and 1 CoolerMaster, the rosewill was really cheap quality, CM was good for about a month then didnt want to power on, the OCZs are 850w but anything over ~400w and it will power cycle, cougar just got too hot as far as i can tell. Power supplies i use are mostly 750-800W, and 2 GPUs per(280xs and such) Also, i always make sure to either use a battery backup or have the router on its own separate circuit, as if that one trips, EVERYTTHING goes down...........i know hahhah.........
Cooling:
Have had about 10 fans die on me, and ive never sent any in for replacement haha. Ive used some server fans and blowers to replace most if not all of them. If i catch a fan going out early, ive saved about 4 or 5 by taking them apart carefully and using oil to lube them
I like the blower designs, as they can be crammed pretty close together and do just as well as other designs, plus you can kinda direct the airflow with them. Ive loved the winter and its been good for mining, i moved them all to one big rig so that i can make sure they are cooled well, as in the summer there is no AC in the warehouse i have them in so they need a lot of airflow.
best mining mobo ive used is the MSI G45, 6 GPUs is nice.
another random tip, I just had a couple rigs slow the fan speed for no reason at all, the CONF file is set to 100% i made sure, but the speeds dropped to like 25%, i noticed because the temps were high 90s. I had to manually go to cgminer and change the speeds to 100%.
I usually keep all my cards in the 70-80°C range, although there are a couple cards that have really crappy coolers i let run at 95 haha, they are XFX if you were wondering
Last edited:
- Joined
- Jun 27, 2013
- Location
- Mesa, Arizona
What are the temps like in your warehouse moho?
Im wondering what summer will do to those rigs. Unless that warehouse is climate controlled, 80-90F air, no matter how much or how fast is still 80-90F air!!
Let me put it a different way...you did not get my point. Im not worried about the warehouse and its temperatures...I am worried about what that change will do to the GPUs temps.
A 'normal' ambient temperature for inside a house, is around 70F or ~21C. If the ambient temp hits 90F (32C) in that warehouse because it has no environmental controls, the temps of each card will raise 11C. Considering even in a 21C environment I have a couple cards hitting 80C already, the additional 11C will make them throttle. Not to mention it is UP on a rack, heat still rises, so it could be even warmer than ambient where it is at.
Basically if that warehouse is not environmentally controlled he may have some serious issues with throttling of the cards that do not have the headroom to support the ambient temperature changes that will be incurred in such an environment.
It makes a difference, but not as substantial as some think. Look at the air circulation with no case and those big fans. I would guess that temps in the warehouse will be likely cooler than temperatures inside most people's cases.
I disagree with that assertion...
Point is, whatever the intake ambient is and his GPU temps are in C, it will raise on a 1:1 ratio with the intake ambient temps. If, for example he is at 21C and temps hit 30C, a 9C increase in GPU temps will occur. And if there is limited headroom on those cards, which with mining is usually true... as I stated I barely have headroom at 21C on a few cards, it could be a big problem, especially in such a high density setup like that. Again, XXF/C air is XXF/C air no matter how fast it is moving.
Maybe some terrible air flow cases...Mine doesn't change from ambient intake to exhaust but a few C at MOST. That said, I am watercooled, BUT most air coolers get pretty close to ambient temps... at least in most cases I had as I tested this with temp probes.
- Joined
- Jun 27, 2013
- Location
- Mesa, Arizona
I have to agree with ED on this one, which is why I was pondering the ambient temps in the warehouse.
- Joined
- Nov 26, 2004
- Location
- 1337 Leet Street , CO 1337
Pretty close to 70f but.. I have a huge door that will stay open 99% of the day.
maybe forgot to mention, I went through this last year, they stayed about 80-85C, but i didnt have the open-rack then...
maybe forgot to mention, I went through this last year, they stayed about 80-85C, but i didnt have the open-rack then...
Last edited:
Sounds like you have the headroom to play then... which is a great thing!
Yeah, it may peak there, but your averages are 10C below that at PEAK sumer with 8 months of the year being, according to you, 'sweater weather'. Usually that type of stance are from people where it is acutally 90F for a couple months out of the year, not days... so I found your comment quite funny!I usually get 92 F and the like during August.
It would usually have to be an unusually wet summer to be less than about 85 F. 2006 was one of them.
Last edited:
It will get over 100 here for a month or two. I will show my set up and keep results coming so that we see how temps will do. I am putting rigs into an outbuilding that is not climate controlled and am building in ducting to strategically bring in fresh air, but it will be 100f during the heat of summer. I have run rigs in a mostly enclosed computer room that gets up into the 90s when I shut the door. It has not had issues with throttling or shutting down due to heat.
Also, I don't necessarily think that one degree increase in ambient means a degree of temp increase at the core. It is a delta temp difference and there are other factors such as fan speed and ambient air flow.
I discuss this here, but I have also voiced the opinion in the water cooling area, that people are too worried about keeping temperatures lower than necessary and adding more radiator than necessary.
Also, I don't necessarily think that one degree increase in ambient means a degree of temp increase at the core. It is a delta temp difference and there are other factors such as fan speed and ambient air flow.
I discuss this here, but I have also voiced the opinion in the water cooling area, that people are too worried about keeping temperatures lower than necessary and adding more radiator than necessary.
Last edited:
Friend, you are welcome to take the time to confirm (what I recall to be) already confirmed science for yourself. I understand the need. If this is flawed, either I bricked on the thought (entirely possible) ever Martin and Skinee labs are wrong (which would mean so is their testing as they 'normalize' ambient temps because of the 1:1 relationship) as well as that is where I sourced the information from years ago.
As far as your situation and throttling, its likely you had the headroom versus this situation (the context in which we are solely speaking about) there us *usually* very little headroom. For example, my current PC would do fine in a 30C room because my CPU only hits 55C running P95 so I have plenty of headroom on it, however not on 2 of 4 GPUs in that I mine with as they are already sniffing 80C or so. So cheers to you that yours didn't throttle, but its clear the reason why to me.
I really didn't think I needed to say "all other variables remaining the same" regarding fan speed and airflow... but of course, all other variables would remain constant.
As far as your last sentence, that is for the w/c section and a bitout of scope to me within this context?
As far as your situation and throttling, its likely you had the headroom versus this situation (the context in which we are solely speaking about) there us *usually* very little headroom. For example, my current PC would do fine in a 30C room because my CPU only hits 55C running P95 so I have plenty of headroom on it, however not on 2 of 4 GPUs in that I mine with as they are already sniffing 80C or so. So cheers to you that yours didn't throttle, but its clear the reason why to me.
I really didn't think I needed to say "all other variables remaining the same" regarding fan speed and airflow... but of course, all other variables would remain constant.
As far as your last sentence, that is for the w/c section and a bitout of scope to me within this context?
- Joined
- Sep 7, 2013
Okay I'm confuzzled now.
If ambient temps rise say 10C, the GPU core temps rise 10C as well?
If ambient temps rise say 10C, the GPU core temps rise 10C as well?
I get your point that ambient temp increase equals GPU temp increase and it is likely that poorly cooled cards will throttle if all else is constant. So I think we agree for the most part. So, the difference in our statements is only the in the effects of the variables, which have been adequately addressed by anyone that has an open rack type miner that is ventilated by fresh air or a larger than usual quantity of semi fresh air (a warehouse).
Its funny now, where this went with assumptions on both sides. I was thinking the airflow would remain the same for moho, you seem to think that it will change. LOL!
Anyway, the latter part of your post I cannot prove either way but point you to a riddle... Which is hotter, 90F wind hitting your face at 5MPH or 90F wind hitting your face at 25MPH...(hint, no windchills over XXF!!!) I have no idea if this miner is going to pour on more airflow...but it really wouldn't matter too much in the big picture considering the differences won't really be much at all as the heatsinks are seemingly saturated when mining so the rate at which they shed the heat will only be so much regardless of the amount of air at whatever temperature it is going through it.
...unless to prove a point you are thinking of F3 tornado wind speeds? I don't know (not trying to be a smart ***, just trying to see where what you say would actually happen with a realistic cooling solution). Seems logical to me, that within the context of this thread and realistic cooling methods (those box fans) that temps will increase as science says it should. intake air temp is intake air temp... on a smaller scale, there is a point of diminishing returns with airflow on cooler... the heatsinks can only dissipate so much heat regardless of the airflow going through it.
EDIT:
EDIT2: Also, I found the source of where I got that 1:1 stuff...but contrary to my belief, this may not translate exactly to a 1:1 (but did with a couple of generations of CPUs and GPUs tested.. its what we do as well when measuring temps..normalize the results to a specific value)
http://skinneelabs.com/ek-supreme-hf-review/3/
The moral of the story is to be sure to check those temps (which in doing this last year he already seems to have)!!!!
Anyway, the latter part of your post I cannot prove either way but point you to a riddle... Which is hotter, 90F wind hitting your face at 5MPH or 90F wind hitting your face at 25MPH...(hint, no windchills over XXF!!!) I have no idea if this miner is going to pour on more airflow...but it really wouldn't matter too much in the big picture considering the differences won't really be much at all as the heatsinks are seemingly saturated when mining so the rate at which they shed the heat will only be so much regardless of the amount of air at whatever temperature it is going through it.
...unless to prove a point you are thinking of F3 tornado wind speeds? I don't know (not trying to be a smart ***, just trying to see where what you say would actually happen with a realistic cooling solution). Seems logical to me, that within the context of this thread and realistic cooling methods (those box fans) that temps will increase as science says it should. intake air temp is intake air temp... on a smaller scale, there is a point of diminishing returns with airflow on cooler... the heatsinks can only dissipate so much heat regardless of the airflow going through it.
EDIT:
@ moho - So with 70F, what are your temps now on those GPUs?
EDIT2: Also, I found the source of where I got that 1:1 stuff...but contrary to my belief, this may not translate exactly to a 1:1 (but did with a couple of generations of CPUs and GPUs tested.. its what we do as well when measuring temps..normalize the results to a specific value)
http://skinneelabs.com/ek-supreme-hf-review/3/
The moral of the story is to be sure to check those temps (which in doing this last year he already seems to have)!!!!
Last edited:
