OC Needs more volt manually than with auto

[Xifios]

Maybe I should have mentioned this already.
I only started ocing because I ordered a vega56 this weekend and I was kind of afraid that my cpu may become a little bit of a bottleneck. So even with this card will I not see any difference between 4.3 and 4.4ghz?

I am also still gaming on 1080p and am basically aiming for stable 60fps at ultra settings.

 

[Mr.Scott]

That 100 MHz will not even yield 2FPS.

 

[Xifios]

Ok I guess I will go back to 4.3 then just to be safe

 

[wingman99]

It's about 2% performance gain for every 100MHz over stock.

 

[freeagent]

Some processors like voltage, some don't. Feed it what it needs until you cant control the temps anymore. If I leave llc at auto on my board, it defaults to the second highest setting, which is what I use anyways On my x58 I use full llc when overclocking. No ill effects so far, just clean reliable goodness

 

[Xifios]

I did some more testing and I noticed that my cinebench score was not as high as it was when testing yesterday. I had a cpu score of about 770 and today I always got a score of around 750. So I closed every background task I could think of to see if that helps but unfortunately it did not. So I then went into my taskmanager and set the priority for cinebench to high.
Now this actually got me a much higher score of around 800! However my pc always kind of froze during the benchmark for prolonged periods of time or at least appear that way.

Because even though everything was frozen to the point where I could not even move my mouse anymore the rendering was still taking place even though it did not show visually at the time, only after like 10 seconds when my computer unfroze the image was nearly completely rendered.

is this freezing normal when putting the task priority to anything higher than normal or is there something wrong with my OC?

 

[EarthDog]

is this freezing normal when putting the task priority to anything higher than normal or is there something wrong with my OC?

Yes. You switched it to realtime so it will take priority over everything...

 

[Xifios]

Yes. You switched it to realtime so it will take priority over everything...

well actually the freezing also happened when changing it to just "higher than normal" so just one priority higher than normal.

 

[EarthDog]

Again, it is normal.

 

[trents]

If I switch the priority on Cinebench to "Realtime" I get the freezing that Xifios described but not if I only move it to "High".

 

[EarthDog]

I'll get hitching occasionally on High+. Normal, nada. Each system will be different depending on the system and what is going on with it at the time + background tasks.

 

[Xifios]

Not sure if this even belongs in this thread but since it is still a problem with my cpu I guess i post this here again.

So I got my vega 56 and I tested it's performance in gta 5 and ghost recon wildlands so far.
In Wildlands it seems like my gpu is extremely bottlenecked by my cpu. Sometimes the gpu usage goes as low as 60% and my frames go down into the 40s sometimes even 30s.
Granted this mainly happens in cities or areas where a lot is going on.
But even in areas with way less going on my gpu struggles to keep 80% usage while individual cpu cores frequently hit 100% usage.

I expected my cpu to be a bottleneck but this extreme?! I mean it is already @4.4ghz and from what I have heard people with 1070s which is performing similar to the vega 56, do not usually have such a huge bottleneck. And they sometimes still have their 3770k at stock speeds.

Also changing my graphics settings does not really seem to have a big impact on my fps. For example in GTA I always seem to be around 60-70fps with dips to as low as 40.
The thing is that at this point I am suspecting my motherboard to be faulty because I have to set my Pcie gen specifically to 2.0 otherwise I will get no video output.I noticed the same thing a few years back when I installed my gtx 970 but I suspected the 970 to be at fault here because upon researching this problem I found quite a few people who seemed to have the same proplem with their 970s and since technically pcie gen 3.0 is nowhere near noticable faster in practical use I was ok with using it in 2.0.

Could there be a general problem with my motherboard that is significantly slowing down performance or am I just simply expecting too much of my 3770k?

On a side note: I thought about updating my bios since I am currently on a release from 2012. But looking at the bios updates this one seems to be the last one that is not a beta bios. The newest is from 2017 and it is still labeled as beta so I am not to keen on flashing a bios that might make things even worse if it does not work...

Oh and I also checked my memory with memtest86 to check if my ram might be faulty but according to memtest there is nothing wrong with it.

I really don't know what else to do because right now my vega does not really perform that much better than my 970...

 

[wingman99]

I would use DDU driver cleaner and remove the Nvidia old drivers. You should not have a problem with those games on i7 3770k .

 

[Xifios]

I would use DDU driver cleaner and remove the Nvidia old drivers. You should not have a problem with those games on i7 3770k .

Thats the first thing I did, even before putting in my new card.

 

[wingman99]

I checked Bottleneck calculator with a 20% processor overclock LINK: https://thebottlenecker.com/calculator average bottleneck percentage: 2.26%. Graphic card and processor will work great together. So I would check if you have a virus or something running in the back ground. Check Task manger for processes.

 

[EarthDog]

So... how does that program work, and what is there to check against its accuracy?

 

[Xifios]

So I went over to a friend of mine who's system performance should be comparable to mine.
He has the same amount of ram(16gb) and even a more powerful PSU(850w) although I don't think that this is a factor since I have a 750w and that should be plenty.

For the CPU he has a Xeon 1231 v3 which is supposed to be as fast as a 3770k at stock speeds.

I ran a couple of benchmarks for ghost recon Wildlands and I noticed that the results on his machine were about 10% worse than on my system.

So maybe there is nothing wrong with my system after all and this is just due to bad game optimization.

Maybe someone with similar specs and a Vega 56 can post their benchmarks with this game to confirm this.

 

[EarthDog]

Same drivers? Same boost clocks on the GPUs during the test? So many things can be different, friend. Can't really compare like that unless it is the same.

Does GR: Wildlands have an integrated benchmarks or are these manual runs (generally not as accurate).

I don't think there is a problem with your system performance...

 

[Xifios]

Same boost clock on the GPU, and yes it has an integrated benchmark. Maybe the OS has an impact on this as well? Because he is still using Windows 7 while I'm using Windows 10.

On a side note: I think the in-game benchmark for this game is completely horrible because it doesn't represent the actual demand during gameplay.

While my GPU usage is pretty much at 100% during the whole benchmark and my CPU usage only hitting 70% at times, playing the game is a completely different story.
When I'm in a village the game seems to require huge amounts of CPU power and therefore my GPU usage can go down to 60% at times while my CPU usage is at 100%.

Only when I'm in open areas my GPU usage as well as my fps go up.

 

[EarthDog]

Yes, OS and drivers certainly can have an effect on FPS!!!

It may be horrible for actual gameplay, but what it does is give a test that is repeatable and reproducable. So it gives a RELATIVE performance metric.