Video card stopped keeping acceleration \"Core clock\"
Hello all!My graphics card msi GTX 1070 after rebuilding the PC in the new building ceased to hold acceleration core clock previously worked fine at +100mhz . Now the artifacts and crashes already at +30mhz. With memory clock no problem, how it was working at +600mhz and running.
Can offer the problem?
asus z170 pro gaming
Ural 650W
intel core i7 6700k
16 GB of RAM
Antonggghhhhllllllll
To me this situation is familiar. Solved the Twister of overlimit and voltage. Met such on the MSI Cards and Gnobit. Stupid couldn't even work in the drain, due to the factory overclocking. I.e. the instability in this case due to the lack of power on the GPU. A little stabilitywise raising limits and/or +ish the voltage and all is good. There is very little need to add one step. On average to 1.05 V are not handled and did not fall below. Perhaps the new building is a little worse was the heat. A few degrees plus on the VC, too, can affect this behavior.
Ryazancev
Hello. I have the power and temp limit stood at the peak. But if the core voltage supply 100, nothing changes.
Antonggghhhhllllllll
VK can be seen as a manifestation of Laziness (the State of consciousness is manifested in behavior, a consequence of Stupidity) in the prescription of many beech trees in the layer of the graphics card. VK Yandex is the social. network Vkontakte.
See. If I take out the voltage in half, to 100. I will also be Troubles. Tried. You try to put the beginning on normal (Zero). And add exactly one step to complete stability at the maximum frequency and the minimum voltage GPU.
It is not necessary to raise the voltage to 100 immediately. Sometimes it only makes it worse. My case Gigabytes and MSI (previously), these +100 spitalnik gave stability.
For each card, even within the same model, we need a different approach. SILICON LOTTERY.
GPU killed already. this is for you the first clear indicator. right now let's raise the powerlimit, raise the voltage of the GPU and make CPU as your video card, young regonial. 100MHz you give nothing but accelerate out of the glands of the system, that actually happened...
legusor
I doubt it, problems started after the rebuild. Or is it Holy shit what a coincidence)
Antonggghhhhllllllll wrote:
And chip is degraded only because of the high temperatures?
Not only that, and not just the chip. If green eyes without artifacts, maybe the problem is in the food. Podewell PSU and it burnt. UPS it is not fashionable to put Nino. )
I also had a gtx 1070 x gaming has a long and more than three years, eventually sold and she gradually began to hold less acceleration over time,most likely degradation of the chip.650 watt enough for her,now is that pumpkin is super rtx 2070.In General, the voltage it is dangerous to increase the power limit there.
Lodurr
But Yes he is just a Ural 650 watts,but rather in the card case the other video card works fine.Why would he even dispersal of torturing elderly 1070
The system is almost one-to-one like mine, also ural 650, 6700k and 1070, a similar situation happened today! I took the card 2 months ago, 1070 gigabytes of 1G, in a new state, I was really surprised when I saw it, immediately took excellent overclocking of 2025mhz at 9.75v, calmly drove into cyberpunk 2077 for a month, but I did not like the temperature after my strix 1080 rev2, I decided so to speak, change the thermal grease, in the end I put everything together, threw in zero thermal gaskets, arctic, stopped keeping the frequency, stupidly flies out until 1987mhz dropped it. In vain I was picking the card, it was like new inside and the overclocking dropped, obviously I have no luck with the gig, before that I poured a bunch of cards from different vendors, it always just got better, but here ..
Hmm ... and no one thought it was about the vidyahi drivers))))
Personally, it happened the same for me. 1070 Super JetStream.
I updated the motherboard bios, updated the firewood on the video - after 20-30 minutes of the game, the system stupidly hung up and went into reboot with random bsods, then it doesn't like the CPU, then there is not enough power, then the driver is somehow inoperative ... And then I remembered - Driver! BIOS!
I returned everything as it was BEFORE the changes - lo and behold! And the frequencies began to take as then, and there are no more bsods.
And also Windows updates often affect feed - now everywhere they secretly sew in protection against underclocking ... Well, they say, why give people the opportunity to overclock, when it is better to convince them to buy a new vidyahu ?! And this is not nonsense, comrades, but the harsh truth.
Otherwise - what is it for people suddenly after the release of the 30xx series from the green ones - the 10xx series suddenly began to give out "pearls". I googled the Internet - looked at the dates of people's posts, analyzed the problems encountered ... and it all came together
. New (and relatively new, ala x470 chipset and higher) motherboards are shoved with new BIOSes, where they have implemented a new method of gaining control over video memory (hmm, is it a coincidence? !) - because of this, people have from a random shutdown of computers, to the complete inoperability of the system when loading the OS ...
And after the release of 3090 and firewood for it, jambs began with the work of old video cards. (Already 2 green hotfixes were cut after that to "fix" this jamb).
The moral is this - the owners of the 10xx series sit on old wood and bios. Seriously - it's better to save yourself from unnecessary problems. And if there are no problems in new games - is there any point in updating firewood ?!