Overclock.net banner

1 - 10 of 10 Posts

·
Registered
Joined
·
3 Posts
Discussion Starter #1 (Edited)
Okay guys,

I'm having a few issues with this current build and would like some suggestions and user experience.

Here is the build;

Motherboard: Asus ROG Rampage VI extreme
CPU: Intel core i9 7920x
RAM: 2x8GB G.Skill Trident Z 3600Mhz (the other 2x8gb sticks seem to be DOA.... either that or i ruined them when removing the heat sinks to watercool).
GPU: GTX 1080
Storage: 1x 512GB 850 pro, 2x 512GB 970 pro
PSU: Thermaltake Toughpower iRGB Plus 850W PLATINUM
Chassis: Thermaltake Core P5 Open Frame


I have a full custom water-cooling loop.

EK-XTOP Revo D5 PWM Pump Plexi
EK RES X3 250 Reservoir EK RES X3 250 Reservoir
EK Coolstream XE 480 QUAD

4x 120mm Thermaltake Riing RGB
4x 120mm Corsair SP120 Fans
fans in a push/pull setup on the radiator.

EK FB ASUS ROG R6E RGB Monoblock Nickel
EK VGA full cover block (GTX 1080)
EK RAM Monarch X4 Nickel
EK RAM Monarch Module Water Block Black
EK M.2 NVMe Heatsink Black

Thermal Grizzly Conductonaut Thermal Compound used on cpu and gpu blocks.

As you can see this watercooling system isn't your average jo.

Currently I feel as though they system is severly underperforming and having some major issues. This system was built for running zbrush/keyshot primarily with also use of applications such as adobe premier pro, photoshop and illustrator. (Graphics card I took from my previous build).

When using zbrush, I am getting less performance than my surface book 2. My surface book 2 is not average either it is the top end $4000 model with an i7, GTX 1060 etc. As a comparison, i can only handle around half the file size before my system starts to lag as opposed to my surface book 2.

Opening photoshop causes the system to either freeze and never receover without me having to do a hard reset and it actually did cause a BSOD yesterday. This is whilst attempting to open a 15mb .psd file..

When using zbrush for extensive remeshing etc, zbrushes hard crashes without an error warning, it just directly closes off no warnings, no lagging no "this application has stopped working" error message.

I ran prime955 the other day.... Blend test with 24 torture test threads running and my cpu idles as 43 degrees celcius.

Attempting to run small FFT s test at 24 threads causing an imediatly shuts down the entire system within the second i click the start test button.


Running Small FFT s at 12 threads and my cpu idols at a whopping 93degress celcius.


Honestly in need of suggestions as this system definitely isn't running how it should be :s

If you own an i9 7920x can you please let me know what figures you're getting whilst running prime95. Are you able to small FFT s at 24 test threads without thermal throttling causing system failure?


I know there is definitely something wrong considering I can't open photoshop. Could this be a faulty cpu? Faulty mobo? Some other issue?


I should mention I'm currently only running the RAM at it's stock setting. I've not OC it to it's 3600 rating.
 

·
Overclocker in training
Joined
·
13,986 Posts
Hi,
I believe you have a couple issues maybe
PSU might be on the weaker side skylake-x will soak up the juice

But Thermal Grizzly Conductonaut on the blocks might be the ultimate issue if it got on anything might cause all sorts of weird stuff
Thermal Grizzly Kryonaut would of been better or simply noctua nt-h1.

I don't use prime95 just simple benchmarks and maybe realbench/...
 

·
Registered
Joined
·
3 Posts
Discussion Starter #3
I don't believe my issue would be with getting the thermal material on any other component. I'd have assumed this to have a more devistating consequence being essentially liquid metal and highly conductive.

I will need to pull the thing apart and when re assembling I will be using something such as kryonaut as I did find the coductonaught a little annoying to apply. It is suppose to have great results in lowering thermals though and that's why I chose it.

Waiting to here back from someone who owns a 7920x who can give a bit of insight into what sorts of temperatures they have seen on their chip. Surely there are some of you out there?

I'm still betting something is faulty as previously stated if it was the conductonaught being spilt into the CPU or surrounding components there would be a short circuit.

Could it potentially be the thermal compound being applied incorrectly? First time using the conductonaught it's a pain in the ass to use. Did follow the instructions though. Perhaps I applied too much?
 

·
Registered
Joined
·
3 Posts
Discussion Starter #4
I would have thought 850W to be more than enough to run this system all at stock clocks with still headroom to OC. I did run a heavily OC system with a 4770k on a 750W supply. Only things changing since than are obviously the mobo and the 7920x aswell as 2x m.2 SSD however I have added 100W onto the supply.

I don't think the supply is the issue?

However now that I am thinking about it......

I only have the 24pin power connected into the motherboard. From memory there is also an 8pin socket which can be used in conjunction with the 24pin is there not?
 

·
Registered
Joined
·
164 Posts
I would have thought 850W to be more than enough to run this system all at stock clocks with still headroom to OC. I did run a heavily OC system with a 4770k on a 750W supply. Only things changing since than are obviously the mobo and the 7920x aswell as 2x m.2 SSD however I have added 100W onto the supply.

I don't think the supply is the issue?

However now that I am thinking about it......

I only have the 24pin power connected into the motherboard. From memory there is also an 8pin socket which can be used in conjunction with the 24pin is there not?
There's (1) 8-pin and (1) 4-pin EPS connector on the motherboard you're using. At a bare minimum you need the 8-pin, which I thought was required to POST/boot. 4-pin may be considered optional if you're not overclocking and using much power. Still really recommend connecting both.

Also, is your CPU delidded? You mention using conductonaut, but nothing about a delid. Pretty much all Skylake-X chips need a delid to avoid seeing 90+ deg C across cores even without overclocking. This also depends on the version of Prime 95 you're using to test. Anything after version 26.6 defaults to having AVX workloads on, which will see higher core temps and greater power draw vs non-AVX workloads.

There's a dedicated Skylake-X / Kaby Lake-X thread in this subforum with a lot of good info. Recommend going through that as a starting point.
 

·
Premium Member
Joined
·
9,494 Posts
Doubt power is an issue, your proc hits maybe 200-250W max under torture testing? What was the BSOD code? To me it sounds like either you have a bad driver, or possibly bad ram (or ram config). Also, I stopped using asus boards after X99, their stability for me went to ****e, had a 5820k with an asus workstation board that would just not run 24/7 stable, replaced it with an asrock and bam it was a night and day difference. Have you run ram stability tests outside the OS?
 

·
Registered
Joined
·
2,948 Posts
You need to mention if you are delided or not. Also that monoblock is a cancer. Mine didnt even mount correctly the vrm with the crap pads provided and even after that, the temps sucked. I have a few x299 monos and that one was the one that gave me issues. Also are you running it overclocked. If you ocing it then you need 140% and increases your vccin. This is how you should cooling your vrm and chip not with that junk mono.
 

Attachments

·
beginner
Joined
·
2 Posts
hey guys, im probably wrong here, but im a noob(feel free to suggest me somewhere pc noobs can go for questions, and have a general question about my pc.

so recently i build my own pc while literally having no knowledge at all really and just looked at what is expensive and put it into cart. iam a gamer, so i wanted the best possible rig i could get for 4.000€ which i had ready.
I ended up getting the 7920X, Gigabyte Aorus Gaming 9 board, Corsair Vengeance LPX DDR4-3333 DIMM CL16-18-18-36 Dual Kit (32GB) and a KFA2 hall of fame 1080TI.

Now the thing is, games run actually like ****, framerates i get in most games are lower then on my prevoius pc which had a 3770k and a 780gtx.
Is this normal that this CPU (7920x) is totally bull**** for PC gaming and my fps in games are more then halfed to my previous 3770k system. For example old games like diablo i drop under 30 fps while my old system could maintain 150+ FPS even in 4 player mode. Same with newer games like GTA andd so on, my new PC cant even touch any of the results.

So my question:

Could it be that my BIOS is just total messed up, or is this CPU really the worst for gaming i could get ?

thanks for your time in advance !
 

·
Registered
Joined
·
545 Posts
@kiamstylez

Well your choice of X299 and i9-7920X was not necessarily the best value for money way of getting a good gaming PC. But regardless it should be capable of being a great gaming and productivity PC... just unnecessarily expensive for just gaming.

Never mind about that for now. Let's look at what you can do to improve performance. First thing is that X299 i9-7920X works best with quad channel memory... with only two memory sticks you are only using half of the system bandwidth. Also the memory timings could be better. But regardless, now you have two sticks if you get another two you need to get exactly the same thing. (Plan B would be sell and replace with something like this F4-3200C14Q-32GTZR). Also in the BIOS you need to set the memory up to use one of the faster XMP settings.

The Gigabyte board does not have the best VRM power delivery, and the VRMs will get hot powering a HCC CPU like that i9. You need to make sure you have airflow over the VRM heatsink to help it as much as you can. Alternatively get a MB with stronger VRMs and better VRM cooling like the EVGA X299 Dark or Supermicro C9X299-PG300.

The i9 probably will require delidding. Do a Youtube or forum search for details. There are people on eBay that can do it for you if you don't want to do it yourself.

Overall the problem is likely to be 90% thermal throttling problems. What kind of temperatures are your CPU, VRMs and GPU running at? Make sure you have plenty of cool air going into your case and fans pushing warm air out and away from the cool air input vents. Also maybe some other people will have some helpful input.
 

·
beginner
Joined
·
2 Posts
hey thanks for your response and sorry for deleting my post earlier, i put it back up.

For cooling on CPU i use a dark rock pro 3 and its all in a corsair 570x rgb case with plenty of additional cooling installed, i used all spots in the case i could for cooling fans.

While doing nothing in Windows my CPU hovers around 30-38 c° and during gameing its around 60-65 with gpu at 60 at its highest. However if i use gigabyte tools like easytune which sets my cpu to 4,4ghz and i do the cpu-z bench it will exceed 100c° for a short time and during gaming i will be around 70-85 degrees or even higher in the 90s. When i use BIOS at standard my CPU frequency seems like to never get higher to improve performance,i also tried this turbo boost from intel which makes single core applications perform better, but this is all quiete difficult, and most times i use this my PC will simply crash to bluescreen, also while doing geekbench for example, my pc will shut down instantly when i click "Start Bench".

From reading a bit into my problems i read the problem is that most games rely on high single core rates where my cpu is simply not suited, but at least, i thought, it should somehow perform better then my old rig which had the pretty old 3770k.

Is there any data i can read out with a tool that would help understand my problems ?

This all makes me actually quiete sad since i spend so much money, and i was thinking about changing my motherboard and cpu to get the new intels, which will perform way better in highend gaming. I also read that in future games where multiple cores are supported my cpu then would be awesome.

Edit: Also my board shows the Errorcode "E3" which i looked up in the handbook it means " OS S3 wake vector call " Is this something to be worried about ?

thank you everybody
 
1 - 10 of 10 Posts
Top