Hey guys. I was able to snag a 1080 ti MSI Armor from newegg a few days ago.
Turns out there are reports surfacing of this card not being cooled well enough. Has anyone heard about this? Also, what constitutes an "ok" temperature for a GPU to run at? I don't plan on overclocking this card at all.
80 or below under load.
Hm, that's concerning. A lot of people are reporting temperatures into the mid or high 80s.
I have a hard time separating out those who are overclocking vs. not, though. I can get people being upset at not being able to OC this card, but that's not what I intend for it. I just want to run it out of the box. I assume if it stays under 80 for me it is worth keeping, then?
Technically the chips are designed to operate without damage at temperatures close to 100c. The cards will throttle hard at 90 and will turn themselves off at 95. You won't damage the card accidentally with an overclock. Myself and a bunch of other people recommend keeping temps ~80 or lower for stability reasons though. I've found that artifacting and driver stability problems tend to start happening at temperatures above 80 so I keep my cards below that threshold.
gpu or memory overclocks any higher than that and I started to get strange artifacts on benchmarking tests
I never managed to get temps that were out of control though
This is a really normal result for cards with good cooling setups. GPU temperature is primarily a function of voltage, not clock rate. GPU voltage is set by the firmware and can't easily be overclocked on it's own, so stability thresholds being hit before thermal thresholds when overclocking gpus with a good cooler is the norm.
0
Options
Dhalphirdon't you open that trapdooryou're a fool if you dareRegistered Userregular
Ok, so follow up question. What is the best way to test this card "under load"? Just boot up a game? And how long should I run a test to make sure that the card can sustain a certain temperature for several hours?
3dmark has a stress test mode. Just endlessly loops.
Hey guys. I was able to snag a 1080 ti MSI Armor from newegg a few days ago.
Turns out there are reports surfacing of this card not being cooled well enough. Has anyone heard about this? Also, what constitutes an "ok" temperature for a GPU to run at? I don't plan on overclocking this card at all.
80 or below under load.
Hm, that's concerning. A lot of people are reporting temperatures into the mid or high 80s.
I have a hard time separating out those who are overclocking vs. not, though. I can get people being upset at not being able to OC this card, but that's not what I intend for it. I just want to run it out of the box. I assume if it stays under 80 for me it is worth keeping, then?
Technically the chips are designed to operate without damage at temperatures close to 100c. The cards will throttle hard at 90 and will turn themselves off at 95. You won't damage the card accidentally with an overclock. Myself and a bunch of other people recommend keeping temps ~80 or lower for stability reasons though. I've found that artifacting and driver stability problems tend to start happening at temperatures above 80 so I keep my cards below that threshold.
gpu or memory overclocks any higher than that and I started to get strange artifacts on benchmarking tests
I never managed to get temps that were out of control though
This is a really normal result for cards with good cooling setups. GPU temperature is primarily a function of voltage, not clock rate. GPU voltage is set by the firmware and can't easily be overclocked on it's own, so stability thresholds being hit before thermal thresholds when overclocking gpus with a good cooler is the norm.
yeah, I'm not too experienced with overclocking. GPUs especially. But I think what is actually holding me back is the power limiter at 120%. If I could go beyond that I bet my card could handle it.
I think some aftermarket cards will have higher power limits.
am I right in assuming that ultimately the power limit is preventing the card from having higher voltages?
Hey guys. I was able to snag a 1080 ti MSI Armor from newegg a few days ago.
Turns out there are reports surfacing of this card not being cooled well enough. Has anyone heard about this? Also, what constitutes an "ok" temperature for a GPU to run at? I don't plan on overclocking this card at all.
80 or below under load.
Hm, that's concerning. A lot of people are reporting temperatures into the mid or high 80s.
I have a hard time separating out those who are overclocking vs. not, though. I can get people being upset at not being able to OC this card, but that's not what I intend for it. I just want to run it out of the box. I assume if it stays under 80 for me it is worth keeping, then?
Technically the chips are designed to operate without damage at temperatures close to 100c. The cards will throttle hard at 90 and will turn themselves off at 95. You won't damage the card accidentally with an overclock. Myself and a bunch of other people recommend keeping temps ~80 or lower for stability reasons though. I've found that artifacting and driver stability problems tend to start happening at temperatures above 80 so I keep my cards below that threshold.
gpu or memory overclocks any higher than that and I started to get strange artifacts on benchmarking tests
I never managed to get temps that were out of control though
This is a really normal result for cards with good cooling setups. GPU temperature is primarily a function of voltage, not clock rate. GPU voltage is set by the firmware and can't easily be overclocked on it's own, so stability thresholds being hit before thermal thresholds when overclocking gpus with a good cooler is the norm.
yeah, I'm not too experienced with overclocking. GPUs especially. But I think what is actually holding me back is the power limiter at 120%. If I could go beyond that I bet my card could handle it.
I think some aftermarket cards will have higher power limits.
am I right in assuming that ultimately the power limit is preventing the card from having higher voltages?
Yes. The firmware will only deliver up to 2 volts or something close to that IIRC. In your case you need more voltage to increase your stability but the firmware won't deliver that power. It's possible to flash the firmware to one that will let you deliver more voltage but that's a bit of an involved process that I don't recommend doing unless you have one of the cards with dual BIOS and a failover switch.
Another recommendation: Don't stop monitoring your temps, especially if you're overclocking. Dust is a thing, and as the card gets older the thermal paste/TIM won't perform as well. It's better to find out that you need to clean out your card because you saw your temps creeping up than it is to find out because your card is overheating and your games are crashing.
Also, as a bit of card maintenance down the road: Removing the coolers and applying fresh thermal paste is easy. Much easier than most people expect (IMO it's easier than removing/attaching a cpu cooler). If your card is older and not performing thermally as well as it used to, some fresh thermal paste and a deep dedusting of the heatsink can breathe a bit of life into an old card.
Friend has one and has recommended it left and right previously. Good find since at Best Buy it was either that one, or another 27" Dell with GSync that was a whopping $700. I've always been using 1080p, and even then on an old brick of a monitor(whereas with this thing the screen weighs nothing and the stand has all the weight). Looks amazing... I could tell the difference when playing WoW and FFXIV immediately, and really the 2560x1440 didn't mess with my ability to read text as much as I thought. It was a bit of an annoyance getting everything hooked up with how my setup is, though now that I've got it where I want it I can just swivel the thing vertically to get to the ports.
I have a 1080 ti founders. I used EVGA XOC software to overclock it, just using the stock blower cooler.
I managed a 150mhz gpu overclock (boost takes it to ~2050 mhz) and 550mhz 508Mhz memory overclock. The temps seems to max out at 69c which i am totally fine with. The noise is a bit much though. At least i can barely hear it over my headphones.
Yeah I managed to get mine to 172mhz on the core and 375 on the memory (memory does not like being any higher then that) I get temps in the low 50's on water. Never seen it go above 56 degrees. Though now my stock cooler is back on while I wait for a replacement radiator.
I was surprised at how high the core got. I did get it to bench at 180, but it would start to artifact after about 30 minutes of benching.
@Al_wat My Zotac card will push itself to 131% power even though the limiter on their software will only let me go to 120. Maybe I just got lucky and this is out of the norm.
Bloodycow on
" I am a warrior, so that my son may be a merchant, so that his son may be a poet.”
― John Quincy Adams
Friend has one and has recommended it left and right previously. Good find since at Best Buy it was either that one, or another 27" Dell with GSync that was a whopping $700. I've always been using 1080p, and even then on an old brick of a monitor(whereas with this thing the screen weighs nothing and the stand has all the weight). Looks amazing... I could tell the difference when playing WoW and FFXIV immediately, and really the 2560x1440 didn't mess with my ability to read text as much as I thought. It was a bit of an annoyance getting everything hooked up with how my setup is, though now that I've got it where I want it I can just swivel the thing vertically to get to the ports.
Discord text is tiny though :bigfrown:
I was going to recommend that Dell myself (I assume you mean the S2716dg) which I'm currently using and love. The retail price is a bit ridiculous because of g-sync, but it often goes on sale for ~$550 or less.
The one he linked doesn't have G-Sync. It *appears* it's the same panel, though.
@Arthil looks like you're covered now, but I was going to recommend you get a sub-$200 monitor to hold you over while you order something proper online; then just return the cheapo monitor.
looks like that corsair fits 300mm video cards but just barely.
Still, shame, that's like the perfect case for me.
I wonder if I could take the GPU cooling fans off and change them to something that gives me 10 mm.
hey wait a minute
isn't that ideal? no more having your GPU sag because it's so heavy: it's held up on both sides now!
(there really should be an update to the ATX/PCIe standards to establish fixed card lengths and support brackets for the other side of cards)
life's a game that you're bound to lose / like using a hammer to pound in screws
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
Ok, so follow up question. What is the best way to test this card "under load"? Just boot up a game? And how long should I run a test to make sure that the card can sustain a certain temperature for several hours?
Furmark is one of the best programs for GPU burnin and benchmarking.
First build question. I just put everything together and it seems ok, but something obviously went wrong. All the fans and LEDs power on, the USB is providing power to the mouse and keyboard, but the monitor has no signal from both the GPU HDMI or the Motherboard's. My first guess is I screwed up the mobo, but wondering if it might mean something else.
First build question. I just put everything together and it seems ok, but something obviously went wrong. All the fans and LEDs power on, the USB is providing power to the mouse and keyboard, but the monitor has no signal from both the GPU HDMI or the Motherboard's. My first guess is I screwed up the mobo, but wondering if it might mean something else.
All things being equal (as in, everything is in proper working condition), a PC doesn't boot or boot properly because something isn't plugged in correctly.
Ok, I realize this isn't a general PC help thread but I figure this thread gets the most traffic in Moe's and you're all well informed people.
I recently got an older Sony Vaio laptop from the Core 2 Duo era and I got it all set up with Windows 7 and drivers, but I think I'm having a problem with the GPU. It's a dedicated ATi Mobility HD 3470 which is supposed to have 256MB dedicated VRAM, but I think it's using system RAM because my RAM usage is always very high and Speccy sees the card as having 1.5GB VRAM.
Is there a reliable way to check if the system is allocating RAM to the discrete GPU? I don't even believe that's possible with this chip but I can't see another obvious RAM hog in task manager. I tried Google but I didn't find any useful info.
Posts
Technically the chips are designed to operate without damage at temperatures close to 100c. The cards will throttle hard at 90 and will turn themselves off at 95. You won't damage the card accidentally with an overclock. Myself and a bunch of other people recommend keeping temps ~80 or lower for stability reasons though. I've found that artifacting and driver stability problems tend to start happening at temperatures above 80 so I keep my cards below that threshold.
This is a really normal result for cards with good cooling setups. GPU temperature is primarily a function of voltage, not clock rate. GPU voltage is set by the firmware and can't easily be overclocked on it's own, so stability thresholds being hit before thermal thresholds when overclocking gpus with a good cooler is the norm.
3dmark has a stress test mode. Just endlessly loops.
yeah, I'm not too experienced with overclocking. GPUs especially. But I think what is actually holding me back is the power limiter at 120%. If I could go beyond that I bet my card could handle it.
I think some aftermarket cards will have higher power limits.
am I right in assuming that ultimately the power limit is preventing the card from having higher voltages?
Yes. The firmware will only deliver up to 2 volts or something close to that IIRC. In your case you need more voltage to increase your stability but the firmware won't deliver that power. It's possible to flash the firmware to one that will let you deliver more voltage but that's a bit of an involved process that I don't recommend doing unless you have one of the cards with dual BIOS and a failover switch.
At least not right now.
I mean I could have a fit of madness later and want to start tweaking
Also, as a bit of card maintenance down the road: Removing the coolers and applying fresh thermal paste is easy. Much easier than most people expect (IMO it's easier than removing/attaching a cpu cooler). If your card is older and not performing thermally as well as it used to, some fresh thermal paste and a deep dedusting of the heatsink can breathe a bit of life into an old card.
Sorry for not responding to any of the suggestions, threw some towards some peeps in Discord and got a major "GETITGETITGETIN" for this one : https://www.amazon.com/Dell-Computer-Ultrasharp-U2415-24-0-Inch/dp/B01M0LDXTL/ref=sr_1_2?ie=UTF8&qid=1491211406&sr=8-2&keywords=Dell+IPS&th=1
Friend has one and has recommended it left and right previously. Good find since at Best Buy it was either that one, or another 27" Dell with GSync that was a whopping $700. I've always been using 1080p, and even then on an old brick of a monitor(whereas with this thing the screen weighs nothing and the stand has all the weight). Looks amazing... I could tell the difference when playing WoW and FFXIV immediately, and really the 2560x1440 didn't mess with my ability to read text as much as I thought. It was a bit of an annoyance getting everything hooked up with how my setup is, though now that I've got it where I want it I can just swivel the thing vertically to get to the ports.
Discord text is tiny though :bigfrown:
Yeah I managed to get mine to 172mhz on the core and 375 on the memory (memory does not like being any higher then that) I get temps in the low 50's on water. Never seen it go above 56 degrees. Though now my stock cooler is back on while I wait for a replacement radiator.
I was surprised at how high the core got. I did get it to bench at 180, but it would start to artifact after about 30 minutes of benching.
@Al_wat My Zotac card will push itself to 131% power even though the limiter on their software will only let me go to 120. Maybe I just got lucky and this is out of the norm.
― John Quincy Adams
I was going to recommend that Dell myself (I assume you mean the S2716dg) which I'm currently using and love. The retail price is a bit ridiculous because of g-sync, but it often goes on sale for ~$550 or less.
@Arthil looks like you're covered now, but I was going to recommend you get a sub-$200 monitor to hold you over while you order something proper online; then just return the cheapo monitor.
I wanted to build one of those slick portable cases
The NCASE M1 will fit some 1080 cards, there's a list here
All this just to get people to search for glory holes?
I wanted something more like this:
https://www.newegg.com/Product/Product.aspx?Item=N82E16811139052
Which I guess they don't make anymore?
Also my GPU is apparently 298mm
https://reddit.com/r/pcmasterrace/comments/5weabc/when_your_gpu_is_too_big/
Still, shame, that's like the perfect case for me.
I wonder if I could take the GPU cooling fans off and change them to something that gives me 10 mm.
Added to the OP.
hey wait a minute
isn't that ideal? no more having your GPU sag because it's so heavy: it's held up on both sides now!
(there really should be an update to the ATX/PCIe standards to establish fixed card lengths and support brackets for the other side of cards)
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
-- Ben Franklin
Nikola Tesla, probably
Furmark is one of the best programs for GPU burnin and benchmarking.
(He didn't actually say that, contrary to legend. It's from a play.)
Also, how have I never see Casually Explained before? Holy shit.
Steam: Elvenshae // PSN: Elvenshae // WotC: Elvenshae
Wilds of Aladrion: [https://forums.penny-arcade.com/discussion/comment/43159014/#Comment_43159014]Ellandryn[/url]
Can you connect anything else to the monitor? Like a game console or other PC?
Thanks for the help anyways Santa.
Did you plug in the cpu power connector?
I recently got an older Sony Vaio laptop from the Core 2 Duo era and I got it all set up with Windows 7 and drivers, but I think I'm having a problem with the GPU. It's a dedicated ATi Mobility HD 3470 which is supposed to have 256MB dedicated VRAM, but I think it's using system RAM because my RAM usage is always very high and Speccy sees the card as having 1.5GB VRAM.
Is there a reliable way to check if the system is allocating RAM to the discrete GPU? I don't even believe that's possible with this chip but I can't see another obvious RAM hog in task manager. I tried Google but I didn't find any useful info.
I don't know a way to actually check the details, though.
PSN: ShinyRedKnight Xbox Live: ShinyRedKnight