High CPU usage, so called problem.

[Mick Naughty]

Best my 1080 has seen. Just max voltage and power. On heaven though. 

[done12many2]

2 minutes ago, stealth80 said:

1080ti is known to run cooler - i'm not doubting, my issue is MY temps on my 1080. Are your cards FE? What clockspeeds are they running? 

 

You're right, but a pair of overclocked 1080 Tis runs a hell of a lot hotter then stock.  That's why I said I'd run them overclocked. I pull just short of 900w from the wall between the CPU, system and 1080 Ti SLI overclocked when running Superposition.  I can up voltage and pull right at 1100w.

 

I really am doing my best to make this worst possible conditions.  There is no worse benchmark load then Superposition at this time.  I'm maxing voltage at 1.093v.  If you want me to run the voltage higher, sure I will.

[done12many2]

1 minute ago, Mick Naughty said:

Best my 1080 has seen. Just max voltage and power. On heaven though. 

 

Try doing the same with Unigine Superposition in 4k or 8k.  It wrecks GPUs.  

[stealth80]

See I think there's something wrong here, I have the following set up:

 

1700 @ 3.95ghz + Ek Supremacy

1080GTX @ ~ 2.1ghz (Palit boost - not modded) with EK 1080 block

Alphacool XT45 360mm crossflow rad + 3x Thermaltake Riing 12 fans

Alphacool UT60 240mm rad + 2x EK Varder 

Alphacool XT30 120mm crossflow rad + 1x EK Varder

 

My card could hit 63C when I had dual loops in the X9, the GPU had a 360 + 240mm to itself in that thing

 

I use AIda 64 to monitor GPU and CPU temps, starting to wonder if it's mis reporting as by rights - the fluid goes to the CPU after the GPU and the Ryzen sits around 40C whilst gaming

[done12many2]

1 hour ago, stealth80 said:

 

How the hell are your GPU (1080 non ti) temp's so low?  I have a 1080 in my loop with the CPU (Ryzen 1700) with a 360, 240 and 120mm radiators and I've seen mine as high as 63C

 

 

1 hour ago, Mick Naughty said:

I too would like to see that. 100% usage on the gpu's and some on the CPU. Some sort of real load. I'd even take heaven a year that point though. 

 

 

42 minutes ago, stealth80 said:

See I think there's something wrong here, I have the following set up:

 

1700 @ 3.95ghz + Ek Supremacy

1080GTX @ ~ 2.1ghz (Palit boost - not modded) with EK 1080 block

Alphacool XT45 360mm crossflow rad + 3x Thermaltake Riing 12 fans

Alphacool UT60 240mm rad + 2x EK Varder 

Alphacool XT30 120mm crossflow rad + 1x EK Varder

 

My card could hit 63C when I had dual loops in the X9, the GPU had a 360 + 240mm to itself in that thing

 

I use AIda 64 to monitor GPU and CPU temps, starting to wonder if it's mis reporting as by rights - the fluid goes to the CPU after the GPU and the Ryzen sits around 40C whilst gaming

 

Yeah, I don't really know what's up with your setup without diving a little deeper into it.

 

For this run, my ambient temps were at 25c (you can pause the HWInfo64 in video to see it and water temps) and I think I saw the highest card hit 36c.  Normally my ambient temp is 20 to 21c when I'm running air conditioning, but we have the windows open today.  Subtract 4 to 5c due to my current ambient and that puts my hottest card at 31 to 32c under one the the heaviest loads you can put on a GPU at this time.  Temps on the cooler card are obviously even lower.  

 

The video may take a while for YouTube to finish "processing"

 

By the way, this is roughly what my regular 1080 SLI does on a much smaller loop as well. 

 

I should add that this was recorded with Nvidia Shadowplay with the same cards running the benchmark so the score has no real comparative value.

 

 

[stealth80]

26 minutes ago, done12many2 said:

 

 

 

 

Yeah, I don't really know what's up with your setup without diving a little deeper into it.

 

For this run, my ambient temps were at 25c (you can pause the HWInfo64 in video to see it and water temps) and I think I saw the highest card hit 36c.  Normally my ambient temp is 20 to 21c when I'm running air conditioning, but we have the windows open today.  Subtract 4 to 5c due to my current ambient and that puts my hottest card at 31 to 32c under one the the heaviest loads you can put on a GPU at this time.  Temps on the cooler card are obviously even lower.  

 

The video may take a while for YouTube to finish "processing"

 

By the way, this is roughly what my regular 1080 SLI does on a much smaller loop as well. 

 

I should add that this was recorded with Nvidia Shadowplay with the same cards running the benchmark so the score has no real comparative value.

 

 

55C on a run of that, interestingly the CPU never went above 31C coolant is nice and cold

[done12many2]

1 minute ago, stealth80 said:

55C on a run of that, interestingly the CPU never went above 31C coolant is nice and cold

 

You lost me there bud.  What did you say hit 55c?

 

Yeah, when my fans are on max, my coolant sits perfectly at room temperature and doesn't move far from a degree or so higher.

[stealth80]

3 minutes ago, done12many2 said:

 

You lost me there bud.  What did you say hit 55c?

 

Yeah, when my fans are on max, my coolant sits perfectly at room temperature and doesn't move far from a degree or so higher.

card hit 55C, fans are ran by the CPU so they stay slow (ish). But like I said, the coolant is v.cool

[done12many2]

4 minutes ago, stealth80 said:

card hit 55C, fans are ran by the CPU so they stay slow (ish). But like I said, the coolant is v.cool

 

What are you talking about?  I think you need to watch the video again.  The cards never got above 36c.  lol

 

You must have been looking at CPU temp.  The GPU temps are both above that.  

[stealth80]

9 minutes ago, done12many2 said:

 

What are you talking about?  I think you need to watch the video again.  The cards never got above 36c.  lol

 

You must have been looking at CPU temp.  The GPU temps are both above that.  

I mean I ran Superposition in 4k on my rig, the card hit 55C and CPU 31C 

[done12many2]

18 hours ago, stealth80 said:

I mean I ran Superposition in 4k on my rig, the card hit 55C and CPU 31C 

 

Delta between your water temp and GPU temp should rarely exceed 10 to 15c.  15c being worst case scenario under some unrealistic load.  If the delta is higher, you need to see what's slowing the heat transfer from the GPU die to the water.  Either a small blockage in the block, especially in the small groves just above the GPU die, or the paste or thermal interface used.

 

If your water temps are high, your GPU temps will be 10 to 15c higher then the water.  Nothing you can do about that, but remove the heat from the water.  Only way to do that is to push more water through the rads themselves or increase your radiator surface area by adding more rads.

 

I'm going to guess that your water temps are not too high because your 31c CPU temp can't be lower then the water cooling it.  With that said, your delta between water temp and GPU temp is at least 24c, which is a bit high.

 

The new loop for my 5960x / 1080 SLI is only going to be a 480mm + 360mm so it'll be interesting to see how fast I need to spin the fans to keep the water cool.  It's just temporary until x299 drops.  Then the z270 / 7700k moves to the smaller 480mm + 360mm loop and x299 and the 1080 Ti SLI will be cooled by the larger 5 x 360mm loop.  I'll be a happy man then!  

[Mick Naughty]

6 hours ago, typographie said:

If your framerate is uncapped, your system is going to give you the highest framerate it can until something maxes out. Ideally you want your GPU to hit 100% first, but if your system is well balanced your CPU might be pretty high up there as well (and 120 FPS in BF1 is definitely going to give you high CPU usage). If you want to see a lower usage %, you'd just need to cap your framerate to a lower value.

 

The only problem that matters is if performance is not good enough. Poor performance might be diagnosed by seeing high usage of one component when you don't expect to, but high usage on its own is usually just an indication that your hardware is working as it should. It sounds to me like all is well.

If I could cap my frames and have a decent frametime, I would do that in every game. Only done in a few times in a few cod games but those always run better then bf games.

[Mick Naughty]

All dem cores. 

[Krillingt75961]

1 hour ago, Mick Naughty said:

All dem cores. 

So neither disk has much if any activity though your RAM has about half of it being used and your utilization is high. Got any other processes running? Maybe an SVChost being wonky and doing what it wants? I have had TS do some weird things in the past too.

[cfrozer]

13 hours ago, Mick Naughty said:

If you look at my gpu temp, it shows that water is indeed going through it. I have the pump on the lowest setting as I forget to turn it up.

 

ill bring it up next time to check all the cores ans h/t. 

 

 

Why would the temp affect usage? unless you are hitting the TJmax. Also I've see with older hardware the CPU usage tends to go up on BF1

 

[Mick Naughty]

7 hours ago, Krillingt75961 said:

So neither disk has much if any activity though your RAM has about half of it being used and your utilization is high. Got any other processes running? Maybe an SVChost being wonky and doing what it wants? I have had TS do some weird things in the past too.

The ssd's will never have usage. The game always uses 8gigs for me, noticed that on my 32gb rig. CPU drops down to 0-5% when only the game is closed. 

[Mick Naughty]

7 hours ago, That-Something_ said:

Why would the temp affect usage? unless you are hitting the TJmax. Also I've see with older hardware the CPU usage tends to go up on BF1

 

No idea, I asked the same thing. 

[Carclis]

17 hours ago, stealth80 said:

55C on a run of that, interestingly the CPU never went above 31C coolant is nice and cold

That sounds backwards going off my own experience. Running a single Fury + 7700k I always saw amazing temps on the GPU (usually 38/39C max) with the CPU reaching mid to upper 60's. I also have a similar radiator setup to you with 360mm (40mm thick) and 240mm (60mm thick) Ek radiators so it's a little strange to see such high temperatures in your setup on the GPU front. Could be a block mounting issue perhaps...?

[Carclis]

1 hour ago, Mick Naughty said:

No idea, I asked the same thing. 

It was unrelated but pretty important in that it's a pretty unconventional way to run your loop. The way your current setup works is by splitting up your flow into three with flow restriction dictating what component gets the most cooling. The problems that could arise in the future if you weren't aware of this could be a highly restrictive CPU block which forces most fluid to travel through your GPU blocks ie the path of least resistance, resulting in quite high CPU temperatures. To be clear there is nothing wrong with your setup, but you probably want a high pump speed to compensate for the poorer water flow with such a configuration.

[Mick Naughty]

4 minutes ago, Carclis said:

It was unrelated but pretty important in that it's a pretty unconventional way to run your loop. The way your current setup works is by splitting up your flow into three with flow restriction dictating what component gets the most cooling. The problems that could arise in the future if you weren't aware of this could be a highly restrictive CPU block which forces most fluid to travel through your GPU blocks ie the path of least resistance, resulting in quite high CPU temperatures. To be clear there is nothing wrong with your setup, but you probably want a high pump speed to compensate for the poorer water flow with such a configuration.

Well there's nothing wrong with it, plenty of high end systems do it. People wonna quote Linus when it comes to most computer bs. I'll do the same, singularity computers for one. Can't get high cpu temps either, the computer is well aware of what temps are safe. Pump speed will continue to be at the lowest speed possible. But like stated, completely unrelated.