GludDK
New Member
- Total Posts : 5
- Reward points : 0
- Joined: 2018/09/05 04:54:48
- Status: offline
- Ribbons : 0
Hi guys! Hope there's still a few left in this subforum here in the aftermath of the RTX-launch. I've been an owner of the EVGA GTX 970 SSC ACX 2.0+ (04G-P4-3975-KR) for almost three years and it has certainly served me well. However, I remember the card being quiet silent and even dead silent (0 RPM) at idle. That's just not the case anymore. I've done some testing and I've used HWiNFO to monitor my card as it took on the challenge of some PUBG and some Unigine Heaven benchmark. Being pushed to 100% utilization the card reaches 80-82 °C in no time. I know PUBG is kinda pushing GPU's hard in terms of heat, but the same results applies for BF1 and (ofc) Unigine Heaven. The clock speed seems stable, though sometimes dipping below 1300 MHz. I've monitored the card at idle as well and I feel it's still on the hot end of what's expectable. After a cold boot (yes, also physically cold) the card slowly (5 min) heats up towards 62°C from where the fans kick in at around 300rpm. Mind you, this is only from booting and loading Windows 10. The slow and inaudible 300rpm fan speed keeps the temp stable at 62°C. Now, that really isn't bothering anyone but I've heard reports of 50°C at idle for this card and 82°C at load seems quiet hot. And yes, the fans are all maxed out at load as well. I've check the card and both fans do indeed spin at the same rpm (at least it looks like that). The card is pretty clean as well (gave it some compressed air last week). My computer case is a Fractal Design Define C with stock fans (1 front @ GPU hight, 1 back @ CPU hight) pushing 1200rpm at all times. The Intel i7-6700K at 4,6Ghz has a Noctua NH-D15s (push/pull) respecting the airflow of the case fans. I really don't feel like this is an issue with my airflow, even though the Define C is not the best case in terms of airflow (it's not terrible either). Intially I though it might be because I've connected 3 monitors (on at 144hz) + a capture card, but I've done all my testing with only one monitor connected at 60hz. What do you guys think? Isn't it a bit hot? Or am I'm just being silly? Thanks! 
post edited by GludDK - 2018/09/06 06:36:59
|
Wolfpack Hybrid
New Member
- Total Posts : 40
- Reward points : 0
- Joined: 2016/01/12 02:32:34
- Status: offline
- Ribbons : 0
Re: GTX 970 SSC ACX 2.0+ running hot
2018/09/05 05:43:08
(permalink)
GludDK
Hi guys!
Hope there's still a few left in this subforum here in the aftermath of the RTX-launch. I've been an owner of the EVGA GTX 970 SSC ACX 2.0+ (04G-P4-3975-RX) for almost three years and it has certainly served me well. However, I remember the card being quiet silent and even dead silent (0 RPM) at idle. That's just not the case anymore. I've done some testing and I've used HWiNFO to monitor my card as it took on the challenge of some PUBG and some Unigine Heaven benchmark. Being pushed to 100% utilization the card reaches 80-82 °C in no time. I know PUBG is kinda pushing GPU's hard in terms of heat, but the same results applies for BF1 and (ofc) Unigine Heaven. The clock speed seems stable, though sometimes dipping below 1300 MHz.
I've monitored the card at idle as well and I feel it's still on the hot end of what's expectable. After a cold boot (yes, also physically cold) the card slowly (5 min) heats up towards 62°C from where the fans kick in at around 300rpm. Mind you, this is only from booting and loading Windows 10. The slow and inaudible 300rpm fan speed keeps the temp stable at 62°C. Now, that really isn't bothering anyone but I've heard reports of 50°C at idle for this card and 82°C at load seems quiet hot. And yes, the fans are all maxed out at load as well.
I've check the card and both fans do indeed spin at the same rpm (at least it looks like that). The card is pretty clean as well (gave it some compressed air last week). My computer case is a Fractal Design Define C with stock fans (1 front @ GPU hight, 1 back @ CPU hight) pushing 1200rpm at all times. The Intel i7-6700K at 4,6Ghz has a Noctua NH-D15s (push/pull) respecting the airflow of the case fans. I really don't feel like this is an issue with my airflow, even though the Define C is not the best case in terms of airflow (it's not terrible either). Intially I though it might be because I've connected 3 monitors (on at 144hz) + a capture card, but I've done all my testing with only one monitor connected at 60hz.
What do you guys think? Isn't it a bit hot? Or am I'm just being silly?
Thanks!
It's not the heat you should be worried about with your card. It's possibility that you like me got a lemon EVGA card where your R33 Inductors are running hot and will fail sooner or later. My 980ti HYBRID watercooled card had no heat problems and it fried. I suggest you watch this video. These issues persist through 900/1000 series. If you do not wish to watch whole video skip to 18:00min EVGA's VRM Thermals Not the Killer of Cards - Final Test https://www.youtube.com/watch?v=mpxQaSjQclo
|
bob16314
Omnipotent Enthusiast
- Total Posts : 8048
- Reward points : 0
- Joined: 2008/11/07 22:33:22
- Location: Planet of the Babes
- Status: offline
- Ribbons : 761
Re: GTX 970 SSC ACX 2.0+ running hot
2018/09/05 06:03:41
(permalink)
Seems a little on the 'hot' side to me, especially at idle, but under certain conditions/configurations it will not downclock at idle and run a little 'hot'..Is it downclocking into power-saving states at idle?..What's the GPU clock speed at idle? If it's not downclocking at idle, see post #11 here..You might consider taking the card apart and replacing the thermal paste on the GPU chip (I would), which is easy to do and will not void any EVGA warranty you might have as long as you don't damage the card in any way in the process. The Maximum GPU Temperature for a GTX 970 chip is 98C, but will thermal throttle at 80C or whatever the Temp Target says in EVGA Precision. See also: Why does my EVGA graphics card run hot? in the EVGA FAQs for more information.
* Corsair Obsidian 450D Mid-Tower - Airflow Edition * ASUS ROG Maximus X Hero (Wi-Fi AC) * Intel i7-8700K @ 5.0 GHz * 16GB G.SKILL Trident Z 4133MHz * Sabrent Rocket 1TB M.2 SSD * WD Black 500 GB HDD * Seasonic M12 II 750W * Corsair H115i Elite Capellix 280mm * EVGA GTX 760 SC * Win7 Home/Win10 Home * "Whatever it takes, as long as it works" - Me
|
GludDK
New Member
- Total Posts : 5
- Reward points : 0
- Joined: 2018/09/05 04:54:48
- Status: offline
- Ribbons : 0
Re: GTX 970 SSC ACX 2.0+ running hot
2018/09/05 06:49:04
(permalink)
Wolfpack Hybrid
It's not the heat you should be worried about with your card. It's possibility that you like me got a lemon EVGA card where your R33 Inductors are running hot and will fail sooner or later. My 980ti HYBRID watercooled card had no heat problems and it fried.
I suggest you watch this video. These issues persist through 900/1000 series. If you do not wish to watch whole video skip to 18:00min
EVGA's VRM Thermals Not the Killer of Cards - Final Test
Interesting video - typical Gamers Nexus  Watched it all the way trough and the conclusion is that there's really no way knowing if you have at bad-capacitor/inductor-card and if it's going to pop. However, my issue is specificly related to the GPU-temperatur reported by the GPU itself and the minor thermalthrottling as a result of this. I feel that the issues presented in the video is more universally related to every EVGA GPU owner and that it isn't the cause of my higher GPU temperatures. Thanks for sharing the video though bob16314
Seems a little on the 'hot' side to me, especially at idle, but under certain conditions/configurations it will not downclock at idle and run a little 'hot'..Is it downclocking into power-saving states at idle?..What's the GPU clock speed at idle?
If it's not downclocking at idle, see post #11 ..You might consider taking the card apart and replacing the thermal paste on the GPU chip (I would), which is easy to do and will not void any EVGA warranty you might have as long as you don't damage the card in any way in the process.
The for a GTX 970 chip is 98C, but will thermal throttle at 80C or whatever the Temp Target says in EVGA Precision.
See also: in the EVGA FAQs for more information.
Thank you, Bob. I'll check the idle clocks upon my return to my PC. Reading through post #11 in the topic you mentioned I saw the thing about "multi display power savings". Never heard of it, so better check up on that. Also, the post tells me to change the power managment mode inside Nvidia Control Panel from "Prefer Maximum Performance" to "Adaptive". I'm pretty sure mine is set to Maximum Performance since virtually every tweaking guide for PUBG (and other games) recommend this. Might be some kind of placebo voodoo though, since I can't find any reason why the card would throttle gaming performance on full load because of this option. Might check up on this as well. I'll report back with my findings soon! Thank you!
|
Dr.Death
FTW Member
- Total Posts : 1011
- Reward points : 0
- Joined: 2016/06/29 19:17:44
- Status: offline
- Ribbons : 9
Re: GTX 970 SSC ACX 2.0+ running hot
2018/09/05 07:39:43
(permalink)
''I've check the card and both fans do indeed spin at the same rpm (at least it looks like that).'' have you checked the RPM's ? maybe them powerlogic fans aint going up to speed tp properly cool anymore ? I;ve had to replace that brand fan in a few cards and was the trick you can see they use electronics in the fan motor trhat I think is failing on them and may not be the cards [seeing how known good working fans fixed / resolved in there place ]  may or may not be but just something else to conceder when scratching your head on this issue p.s could also the fan telling the card senders it doing 2000 rpm but its really doing 800 ..lol.. ???? just some extra electronics to tend to fail and I now wonder how many guys get told its the 300-500$ card and junk it over a 15 buck fan in the end evga don't sell fans theres some at ebay claimed as new ? https://www.ebay.com/sch/i.html?_from=R40&_trksid=m570.l1313&_nkw=evga+gtx+970+fans&_sacat=42000
post edited by Dr.Death - 2018/09/05 07:46:33
|
GludDK
New Member
- Total Posts : 5
- Reward points : 0
- Joined: 2018/09/05 04:54:48
- Status: offline
- Ribbons : 0
Re: GTX 970 SSC ACX 2.0+ running hot
2018/09/06 06:36:10
(permalink)
Thank you everyone for your feedback on my situation! Really appreciate it! bob16314
You might consider taking the card apart and replacing the thermal paste on the GPU chip (I would), which is easy to do and will not void any EVGA warranty you might have as long as you don't damage the card in any way in the process.
Rocking the third year of it's service, I guess my card is way out of its warranty period anyway. So yesterday I went ahead and took it apart, wiped the old thermal compound of and applied a fresh grain of Noctua NT H1. Unfortunately, my temps didn't change at all. Still pushing 80-82°C on load and exactly 61°C at idle. However, I did change the power management option inside Nvidia Control Panel from "Prefer maximum performance" to "Adaptive". Before that the clock speed at idle was around 1150Mhz and I'm now seeing a decrease to around 950MHz. That took the idle tempereatur down below 60°C and my fans are now off. Yay! Might try the Nvidia Inspector trick (hack?) to enable multi display power savings as well, though I feel like this could be a SLI-related setting. Dr.Death
Have you checked the RPM's ? maybe them powerlogic fans aint going up to speed tp properly cool anymore ?
I've had to replace that brand fan in a few cards and was the trick
you can see they use electronics in the fan motor trhat I think is failing on them and may not be the cards (seeing how known good working fans fixed / resolved in there place)
may or may not be but just something else to conceder when scratching your head on this issue
p.s could also the fan telling the card senders it doing 2000 rpm but its really doing 800 ..lol.. ???? just some extra electronics to tend to fail and I now wonder how many guys get told its the 300-500$ card and junk it over a 15 buck fan in the end
HWiNFO reports around 2500rpm at 82°C (full load) and I'm pretty sure that's about true since it sounds like a small jet engine. However, since I'm only able to confirm this by hearing and seeing the fans, I can't be certain that they are reporting the correct rpm. I don't think I'll go for some spare fans though. I'm not even sure my fans are broken. Might opt for another cooling solution down the road though. I just realized that my card has the dual BIOS option, so I might play around with this feature to investigate heat, noise and performance further. Who knows, maybe the currently enabled BIOS is to agressiv on the GPU voltage? Is this a thing? Although I'm considering upgrading the card before the end of this year, I do want to get to the bottom of this. I plan on selling the card at that time and I do not wish to have it thrown back at my face because of thermal throttling. Yesterday, it did dip to 1270MHz running PUBG, so there's a bit of that nasty throttling going on.
post edited by GludDK - 2018/09/06 06:40:39
|
Dr.Death
FTW Member
- Total Posts : 1011
- Reward points : 0
- Joined: 2016/06/29 19:17:44
- Status: offline
- Ribbons : 9
Re: GTX 970 SSC ACX 2.0+ running hot
2018/09/06 08:27:41
(permalink)
I been using fans ou of a old sapphire card for testing
check under the fans blades for any dust grime caked on that may been missed from blowing it out . or pull the shroud and look over all the coolers fins as well for any stuck in a hard to get to blow out places ? I gusee if you go that far may as well re due the paste as said above ?
still could be the card and all that does nothing in the end then it just run it as is and let it bop till she drops. may just run fine with high heat for another year or 2 ?? or something comes along you want to replace it with anyway ..
I know my old evga 8800 cards ran so hot they melted the thermal pads used in them ,but still worked great I just forgot they had lifetime warrantee and ended up giving them away to a kid in need ...
good luck
|
joeymir
SSC Member
- Total Posts : 732
- Reward points : 0
- Joined: 2007/04/09 16:45:56
- Location: Oregon
- Status: offline
- Ribbons : 4
Re: GTX 970 SSC ACX 2.0+ running hot
2018/09/06 17:11:11
(permalink)
It's possible I missed it, but what about ambient room temps? Being summer and all, i'd suspect a good 5-10c increase in a room with little or no A/C which in itself would explain the increased temps. But if your room is temperature controlled, ignore what I just said.
You applied new thermal compound, so check that off the list. I've heard others talk about cards "deteriorating" thermal performance overtime, however i suppose I've never taken note of this ever happening in my PC history(possibly because I upgrade mildly frequently). All in all 82C is well within normal operating temps of this card, however anytime you see temps you generally haven't it can raise questions. For now, i'd try completely removing Nvidia drivers, and reinstalling as it never hurts. i'll wait to hear back before posting further.
Gigabyte X570 Aorus Master | Ryzen 9 5900X | 32GB Crucial Ballisitx DDR4 3600MHz | MSI Radeon 6900 XT | Samsung 950 Pro 512GB | 2x Crucial MX500 1TB Raid-0 | 2x WD Velociraptors 500GB Raid-0 | Seasonic Focus Platinum 850W | Lian-Li PC-O11 
|
GludDK
New Member
- Total Posts : 5
- Reward points : 0
- Joined: 2018/09/05 04:54:48
- Status: offline
- Ribbons : 0
Re: GTX 970 SSC ACX 2.0+ running hot
2018/09/07 04:28:48
(permalink)
Dr.Death
I been using fans ou of a old sapphire card for testing
check under the fans blades for any dust grime caked on that may been missed from blowing it out. Or pull the shroud and look over all the coolers fins as well for any stuck in a hard to get to blow out places ? I gusee if you go that far may as well re due the paste as said above ?
still could be the card and all that does nothing in the end then it just run it as is and let it bop till she drops. may just run fine with high heat for another year or 2 ?? or something comes along you want to replace it with anyway ..
I know my old evga 8800 cards ran so hot they melted the thermal pads used in them ,but still worked great I just forgot they had lifetime warrantee and ended up giving them away to a kid in need ...
good luck Already did a pretty good clean, but I'll check again. Thermal paste has been reapplied as said in my last post, unfortunately without any results. However I did notice a strange miscolored spot of the copper heat spreader of the cooler. See my attached photos. Can't really see how this should affect anything, but I did find it quite odd. Rubbing alchohol on it intensively didn't do anything.   I might end up just living with the high temps since they are kept beneath critical levels. I just find it hard to accept that i might be unable to fix it. joeymir
It's possible I missed it, but what about ambient room temps? Being summer and all, i'd suspect a good 5-10c increase in a room with little or no A/C which in itself would explain the increased temps. But if your room is temperature controlled, ignore what I just said.
You applied new thermal compound, so check that off the list. I've heard others talk about cards "deteriorating" thermal performance overtime, however i suppose I've never taken note of this ever happening in my PC history(possibly because I upgrade mildly frequently). All in all 82C is well within normal operating temps of this card, however anytime you see temps you generally haven't it can raise questions. For now, i'd try completely removing Nvidia drivers, and reinstalling as it never hurts. i'll wait to hear back before posting further.
Thank you for giving your two cents, Joey. Appreciate it! My room is quiet small and not temperatur controlled It's usually around 23°C, however the room temp does increase after some gaming (usually 30 min). I think I'll do some testing with the ambient room temeperatur, and the timing couldn't be better with the outside temps coming down to 15°C/59°F. Havn't experienced any deteriorating regarding thermal performance myself, but it would explain a lot in my situation. Will try the DDU-wipe the drivers and do a reinstallation and monitor the voltages afterwards. Thanks, guys! Have a great weekend!
|
Dr.Death
FTW Member
- Total Posts : 1011
- Reward points : 0
- Joined: 2016/06/29 19:17:44
- Status: offline
- Ribbons : 9
Re: GTX 970 SSC ACX 2.0+ running hot
2018/09/07 07:00:16
(permalink)
if you take your fingernail or something does it feel pitted ??? ... I don't know , but I feel that should be a ''clean'' finish . look up lapping and review steps to resurface or polish up that cold plate http://www.overclockersclub.com/guides/lapping/ plenty of more info on that around to review what about the top of the gpu chip ? same thing on it ? thing is it was running cool for years , but you don't know if that's been there all that time as well one thing I say is over time the cards PCB may warp in heating up and cooling down and could break the cooler to chip and paste contact with each other. evga has that [full pcb] stiffing plate to help in preventing that do as you can . and run it till it spits her bits . may still ''run hot'' but do so for a few more years or just kick the bucket in a week ? good luck
|
GludDK
New Member
- Total Posts : 5
- Reward points : 0
- Joined: 2018/09/05 04:54:48
- Status: offline
- Ribbons : 0
Re: GTX 970 SSC ACX 2.0+ running hot
2018/10/01 22:53:23
(permalink)
Alright guys, I've been traveling for a few weeks, so my troubleshooting has been on hold. I hope it's okay to revive the thread or at least give you an update. Dr.Death
if you take your fingernail or something does it feel pitted ??? ... I don't know , but I feel that should be a ''clean'' finish .
look up lapping and review steps to resurface or polish up that cold plate
plenty of more info on that around to review
what about the top of the gpu chip ? same thing on it ?
Hmm... it doesn't feel pitted. It's completely flat and looks more like a miscolored spot. I'm really considering a polish of the copper, but on the other hand I'm so glose to pulling the trigger on a 1080Ti. The only reason would be to make sure the card isn't thrown back at my face after selling it. I really do feel that this particular cards value is really low at this point. I've been optimizing my case airflow and fixed the multi monitor power savings in NVIDIA Inspector which has help tremendously on the clocks and temps at idle. The card sits around 135 MHz to 200ish Mhz and just shy of 50°C which keeps the fans a 0 rpm. I do have some GFX-glitches though, but i guess that's the reason why the multi monitor power savings i disabled for more than one monitor in the first place. Last night I did a few ours of PUBG and I noticed that my fans over time reached an even more insane level of noise. Unfortunately I wasn't able to pull any numbers from HWiNFO during the heat (no pun intended) of the moment. Because of this I did a guick FURMARK this morning - just for the fun. However, it turned out to be no fun at all. In a matter of seconds my card spiked up to a new all time high 91°C with the fans spinning at an insane 4700 rpm. I know FURMARK pushes heat alot, but MY GOD that is one hot GPU right there. I don't know what to think of this, really. What about you guys?
|