• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

NVIDIA GTX 970 OWNERS THREAD

Anyone with a G1 970 does your back-plate feel very hot under load? Mine does, temps aren't that great either, when gaming with the fan on 100% I am hitting 65c...
Case is open so airflow is not an issue.
Also I get strange small line artifacts randomly, but seems to disappear if I take the fan off auto and set to 100% and decrease the voltage! This happens even if I back my overclock down to +100 core +300 mem , I have tested stable upto +170 core +600 core.
 
dean0 said:
Anyone with a G1 970 does your back-plate feel very hot under load? Mine does, temps aren't that great either, when gaming with the fan on 100% I am hitting 65c...
Case is open so airflow is not an issue.
Also I get strange small line artifacts randomly, but seems to disappear if I take the fan off auto and set to 100% and decrease the voltage! This happens even if I back my overclock down to +100 core +300 mem , I have tested stable upto +170 core +600 core.
Click to expand...

Stable and artefact free aren't the same though. The ram can be artefacting it's **** off all the way through a bench run but still complete successfully.

If it's flashes of black, white or red in polygonal shapes then it's memory. I've found on mine if the memory chips are very hot that induces artefacts and usually only happens at the end of bench runs (the lst four scenes of Unigine Valley highlight the issue nicely). When I watercool mine I'm going to put some ram heatsinks on as well.
 
amigafan2003 said:
Stable and artefact free aren't the same though. The ram can be artefacting it's **** off all the way through a bench run but still complete successfully.

If it's flashes of black, white or red in polygonal shapes then it's memory. I've found on mine if the memory chips are very hot that induces artefacts and usually only happens at the end of bench runs (the lst four scenes of Unigine Valley highlight the issue nicely). When I watercool mine I'm going to put some ram heatsinks on as well.
Click to expand...

Makes sense, but it is thin white lines 5-10cm length that flash for split second, like screen tearing. I will be taking it apart tomorrow to check contact on the heatsink and backplate, will update with pics then.
My previous 680 backplate never even felt warm to touch, but this is HOT!
I needs water block ASAP!
 
dean0 said:
Makes sense, but it is thin white lines 5-10cm length that flash for split second, like screen tearing. I will be taking it apart tomorrow to check contact on the heatsink and backplate, will update with pics then.
My previous 680 backplate never even felt warm to touch, but this is HOT!
I needs water block ASAP!
Click to expand...

u dont need to use a wblock for that , just use heatsinks on the 4 memory ic's on the back of the gcard
 
It is not the best of designs when they put 4 of the 8 memory chip on the back of 970 GPU boards. You could always try out some thermal pads between the memory chips and the back plate. Should do the same job has heatsinks.
 
Hi there

Updates
MSI Gamer: In stock
Zotac Gamer: In Stock
EVGA Superclock: In stock
Galax EX OC: In stock
Galax: In stock
Palit Jetstream: In stock
Palit: In stock
Zotac Omega: In stock

Arriving today:
Asus Strix
Gigabyte G1
Inno3D OC
Inno3D Air Boss Ultra


Today is a good day! :D
 
dean0 said:
Anyone with a G1 970 does your back-plate feel very hot under load? Mine does, temps aren't that great either, when gaming with the fan on 100% I am hitting 65c...
Case is open so airflow is not an issue.
Also I get strange small line artifacts randomly, but seems to disappear if I take the fan off auto and set to 100% and decrease the voltage! This happens even if I back my overclock down to +100 core +300 mem , I have tested stable upto +170 core +600 core.
Click to expand...

Those temps seem a little high, mine won't exceed 60 degrees under load but I'm using games that I play as my bench-marking tool.

Having said that I think there are a few driver issues messing with my OC potential as I am getting driver crashes on +150 core +500 mem, & I haven't even tested it with less yet so could be even worse (Titanfall).

Also, case open = good airflow? :/
& artifacts? Damn, keep your eye out mate
 
olliej82 said:
Just ordered a PCI-e extender to try to put some space between my cards on my matx board, has anyone used an extender, do they work?
Click to expand...

what case do you have?

I have a Matx mobo and having 2 gpu's very close is fine, but you need some good air flow to keep top gpu cool. are you gpu's reference?

if you did buy pci-e extender where would you put the other gpu?
 
My case is a zalman z11 so there's plenty of space I should be able to move the bottom card further down with the extender and mount it in one of the empty slots of the back of the case. My 970s are the msi ones, temps are ok until I use dsr and then the too hard gets way too hot. I've also ordered a new set of fans to help with the cooling but my case temp doesn't go above 30c anyway so it probably won't help
 
olliej82 said:
My case is a zalman z11 so there's plenty of space I should be able to move the bottom card further down with the extender and mount it in one of the empty slots of the back of the case. My 970s are the msi ones, temps are ok until I use dsr and then the too hard gets way too hot. I've also ordered a new set of fans to help with the cooling but my case temp doesn't go above 30c anyway so it probably won't help
Click to expand...

Reference cards would have been better choice for a Matx mobo cause heat will blow out the back of gpu.
 
amigafan2003 said:
Afterburner trace also shows 1472mhz, so I guess that's the correct "boost" clock.

ANother question, card 1 in gpu-z shows peak volts as 1.175v. I thought these cards peaked at 1.25v? Card two however peaks at 1.212v and "PerfCap Reason" shows VRel, VOp when running 3D stuff.
Click to expand...

I've got +140 on the core which gives me 1456. However, I feel the MSI boost just picks a speed it likes. If I leave core at +140 but increase voltage to +87mv the boost speed decides to go up to 1486mhz. I've also seen it vary between games and even between the cards (running at slightly different clocks).

As for the voltage, I have the same difference (I haven't seen any SLI users say they haven't yet?), there are a couple of threads on it elsewhere but nothing conclusive. I know its not my PSU as its 1300w. I am tempted to mess around with trying eg a different cable from the PSU, etc. But it means taking my pc apart (again) and I don't think it will resolve anything.
 
Legojon said:
I've got +140 on the core which gives me 1456. However, I feel the MSI boost just picks a speed it likes. If I leave core at +140 but increase voltage to +87mv the boost speed decides to go up to 1486mhz. I've also seen it vary between games and even between the cards (running at slightly different clocks).
Click to expand...

Hmm I'm no expert at OC at all but I have something similar.

+100 core +300 mem on stock volts
This takes me up to 1429MHz boost core clock as shown in GPU-Z.

However after a quick 10 min session in Tomb Raider Afterburner shows a max core clock of 1480MHz was hit :S
 
Legojon said:
I've got +140 on the core which gives me 1456. However, I feel the MSI boost just picks a speed it likes. If I leave core at +140 but increase voltage to +87mv the boost speed decides to go up to 1486mhz. I've also seen it vary between games and even between the cards (running at slightly different clocks).

As for the voltage, I have the same difference (I haven't seen any SLI users say they haven't yet?), there are a couple of threads on it elsewhere but nothing conclusive. I know its not my PSU as its 1300w. I am tempted to mess around with trying eg a different cable from the PSU, etc. But it means taking my pc apart (again) and I don't think it will resolve anything.
Click to expand...

The boost and voltage are linked.

I can check this while mining (crypto coins etc)

If I set the core to +250 and the voltage to +87 it will boost to around 1604

Sliding the voltage back to 0 will also reduce the boost clock slightly.

Even with the voltage at +87 GPUz reports that my boost clock is being limited by volts.

The Boost box on GPUz should be ignored. The card will boost further until the volts or TDP are reached, hence why it boosts higher in game
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom