GTX 1060 OC

[n3onis]

Seen people get higher overclocks, but rising the memory clock any further gives me green pixel artifacts.
Rising core clock also crashes either the stress test or my screen goes black.
Overall, I'm pretty happy with the overclocking capabilities of this card.

Stressed with MSI Kombustor (FurMark), which other stress testing software do you recommend?

 

[Liviu Cojocaru]

Imo the GPU clock is way to high and that is why you're getting those problems

 

[Vya Domus]

You are pushing your memory too far most likely , artifacting on Pascal is usually related to memory clocks. Most 1060s can do +500mhz plus on the memory and few up to 700mhz ( mine does +700 for example) , so nothing unusual there you were a tad unlucky.

If you want higher clocks maybe try setting a custom voltage/frequency curve in afterburner and set the voltage limit to 100%. Not guaranteed to give you higher clocks but in my case it allowed me to get an extra ~100mhz.

Stressed with MSI Kombustor (FurMark), which other stress testing software do you recommend?

None , just play games. AND STOP USING FURMARK !

Imo the GPU clock is way to high and that is why you're getting those problems

Not necessarily , my 1060 runs up to 2164mhz and doesn't crash/artifact , also those clocks reported by the specs page on GPU-z are inaccurate , you can't tell how high it goes in actual games.

 

[Liviu Cojocaru]

you can't tell how high it goes in actual games.

You can if you use MSI Afterburner

 

[P4-630]

You can if you use MSI Afterburner

You can see the maximum actual boost clock in GPU-Z Sensor tab as well..

 

[Vya Domus]

You can if you use MSI Afterburner

Yes ,but you cannot determine that by just looking at the specs page on GPU-Z , that's what I meant to say.

 

[Liviu Cojocaru]

Yes ,but you cannot determine that by just looking at the specs page on GPU-Z , that's what I meant to say.

Ok I understand.
OP,what OC software do you use? Take a screenshot of that software and upload it here. Also it would be very useful to know temps as well as the case you're using and if you have a decent airflow

 

[Laki89]

Stressed with MSI Kombustor (FurMark), which other stress testing software do you recommend?

Any 3d mark after 2008 year

None , just play games. AND STOP USING FURMARK !

What's bad with furmark, it's great stability test for GPU

 

[cdawall]

My zotac short pcb1060 35th cards all do +200mhz core (22xx boost) and +800mhz memory. Those clocks seem rather low, but it could just be a junk card.

What's bad with furmark, it's great stability test for GPU

Not really all of the aforementioned programs are automatically throttled to low power 3d clocks. So all you are doing is creating useless heat. It doesnt test peak clocks at all as the card will bever reach them. May as well downclock your cpu and run prime95 to sww how it does with an overclock.

 

[OneMoar]

the 1060 benefits more from memory clock
yours has Samsung memory it should easily be hitting +500 on the memory
core clock peaks between 2100 and 2200 are not unheard of

and furmark blows cards up do not use

 

[Vya Domus]

What's bad with furmark, it's great stability test for GPU

Go ahead and keep using it , it will be your VRMs that will blow up not ours.

You get 0 extra information about stabilty , Furmark is like prime95 but with an extra chance to start a fire.

I am exaggerating but you get the idea , don't use it.

 

[eidairaman1]

Do not furmark the card.

Stability testing is any gpu based benchmark out there

 

[MrGenius]

In a recent vid I heard der8auer say he runs Prime95 and Furmark at the same time to verify OC stability for the pre-built overclocked systems he tests for sale by Caseking.

Check the vid between 10:28 and 11:37

Anyway. All the shit talking about Furmark just makes me laugh. How many of you have actually killed a card with Furmark? How much sense does it make that MSI would contribute to the development of/sponsor, advertise with their name on, and release a piece of software through downloads on their website that is going to kill graphics cards? Kombustor is Furmark! Are they just hoping and wishing to be sued? Do they make more money through the RMA process of all their customer's cards that Kombustor kills? Is it that only MSI graphics cards are safe to run Kombustor/Furmark? Why is the latest version of Furmark now in the TPU downloads section? I mean it's going to kill graphics cards right? Yeah...not very damn likely.

 

[Vya Domus]

Anyway. All the shit talking about Furmark just makes me laugh. How many of you have actually killed a card with Furmark?

It's more than that , Furmark stresses the GPU so much that it is guaranteed to throttle. As a stability test it is essentially useless for modern cards , you never get to run the card at it's peak clocks for extended periods of time.

It's not just Furmark , it's all of these benchmarks. Best way to find out how the card behaves when you use it is to test it with what you actually use , not a synthetic benchmark.

And it does kill cards , especially older ones that don't throttle , rarely but it does. Guess what most new cards can actually detect when you run Furmark (or programs that are alike ) and automatically throttles the card , I wonder why ?

Yeah...not very damn likely.

Yeah , not damn likely , but possible.

In a recent vid I heard der8auer say he runs Prime95 and Furmark at the same time to verify OC stability for the pre-built overclocked systems he tests for sale by Caseking.

Like I said , most recent cards detect at a driver level when a load like the one Furmark produces is applied and automatically throttles the card , he is wasting his time by using it.

All in all , useless piece of software that has spawned and is still spawning issues when it shouldn't have been the case.

 

[Laki89]

Anyway. All the shit talking about Furmark just makes me laugh. How many of you have actually killed a card with Furmark?

One big zero: 0
and if you ask how many cards i tried with furmark- more than 100

Kombustor is Furmark!

yes, it's set of different tests similar to furmark, and one of them fury donut

Is it that only MSI graphics cards are safe to run Kombustor/Furmark? Why is the latest version of Furmark now in the TPU downloads section?

Why - to ppl test it's own cooling solutions (factory or custom) on their graphics cards

It's more than that , Furmark stresses the GPU so much that it is guaranteed to throttle.

There is reason why it throttles and it is normal in that extreme usage

It's not just Furmark , it's all of these benchmarks. Best way to find out how the card behaves when you use it is to test it with what you actually use , not a synthetic benchmark.

If graphics card pass synthetic benchmarks well (without crashing , system restart,etc.) it will work in any game very well, because games are not pushing card to the limits like synthetic tests does

 

[OneMoar]

One big zero: 0
and if you ask how many cards i tried with furmark- more than 100

yes, it's set of different tests similar to furmark, and one of them fury donut

Why - to ppl test it's own cooling solutions (factory or custom) on their graphics cards

There is reason why it throttles and it is normal in that extreme usage

If graphics card pass synthetic benchmarks well (without crashing , system restart,etc.) it will work in any game very well, because games are not pushing card to the limits like synthetic tests does

you are wrong and dangerously so and blocked

 

[DeathtoGnomes]

I love arguments like this **POPCORN**

 

[Vya Domus]

There is reason why it throttles and it is normal in that extreme usage

So what exactly are you looking for when you run it if you know what it will do ?

If graphics card pass synthetic benchmarks well (without crashing , system restart,etc.) it will work in any game very well, because games are not pushing card to the limits like synthetic tests does

You are just simply wrong. It might have been somewhat true and useful in the past but nowadays it isn't.

You don't understand how clocks work on recent GPUs , if you did you'd know that running Furmark is a waste of time if you want to check for stability. Different levels of voltages are tied with frequency steps and also to power consumption and temperatures . When the card throttles because of those two last parameters so do the voltages and subsequently the clocks. But only the voltages and clocks are what dictates stability , not the temperature and power consumption which is all that Furmark is pushing for.

When you check for stability you want to know what happens when both voltages and clocks are it's peak. You can run Furmark all day and it wont crash , yet as soon as you will start to play a game that isn't pushing the card as much and the clocks reach it's peak it will.

Not to mention that every application applies a different type of workload , sometimes it might be fine , other times not. It's doesn't matter if that application is a "benchmark" or a "game". I stand by my opinion , you test for stability with what you use.

 

[DeathtoGnomes]

Yaa but.....(just for the sake of saying it, cuz someone wants to ) "If the card is too hot, its not gonna be stable"

 

[Vya Domus]

"If the card is too hot, its not gonna be stable"

Well , they are going to have a hard time overclocking and not understanding why.

 

[cdawall]

Yaa but.....(just for the sake of saying it, cuz someone wants to ) "If the card is too hot, its not gonna be stable"

Any modern card with throttle clocks before temperature is in that threshold

 

[Laki89]

So what exactly are you looking for when you run it if you know what it will do ?

Graphics card maximum temperature,because in games it will never reach that high

It might have been somewhat true and useful in the past but nowadays it isn't.

Exactly, i use it mostly for older GPU's

Different levels of voltages are tied with frequency steps and also to power consumption and temperatures . When the card throttles because of those two last parameters so do the voltages and subsequently the clocks. But only the voltages and clocks are what dictates stability , not the temperature and power consumption which is all that Furmark is pushing for.

When you check for stability you want to know what happens when both voltages and clocks are it's peak. You can run Furmark all day and it wont crash , yet as soon as you will start to play a game that isn't pushing the card as much and the clocks reach it's peak it will.

Thanks for info

 

[eidairaman1]

The ignorance in here is strong.

You use furmark and your card dies, expect a verbal lashing for stupidity.

 

[Shou Miko]

Am I the only one who remembers what @W1zzard writes under Overclocking in all his reviews???

Every sample overclocks differently, which is why our results here can only serve as a guideline for what you can expect from your card.

U can't except all cards to overclock the same, but if u look around and check like EVGA SuperClock cards, MSI Gaming cards with their overclocked speeds and some other vendors with their features for overclock they don't do anything massively mby around 100-150mhz on the core because all cards can basically do that.

Nvidia GTX 1060 6GB FE core clock 1506 mhz
EVGA GTX 1060 6GB SC core clock 1607 mhz
EVGA GTX 1060 6GB FTW core clock 1620 mhz
EVGA GTX 1060 6GB FTW+/2 core clock 1632mhz

that's around 100/132mhz every card can do that without much tweaking and most of the cards without any voltage adjustment and in a case with decent air u can get even more, with Boost 3.0 it doesn't not only require volt or temps but a mix.

again @W1zzard made a great review: https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_1080/29.html