enfinders

Hi there, so I’m a little uncomfortable with my current case temps and layout in general. I'm using an NZXT s340 elite. Under load my CPU reaches 70+ degrees, and i found my GPU gets to 80/82 degrees after a couple hours of gaming, which i don't like at all. My h115i is currently at the front intake, and unless my cpu gets over 65 degrees, only 2 fans are spinning (the two inner ones), so the system is quiet, but preventing the cpu from breathing properly. This is always the case, and unless i start to stress test both, my gpu in games always stays around 80 degrees. In a typical stress test scenario my cpu stays around 66/70 degrees, but my gpu drops to 75 max, sometimes even less, depending on the room temperature, but again, this is never the case, so.. pointless data ☹ I was thinking of getting the mastercase pro 5, as this will allow my gpu to breathe, by installing 2 140mm fans in the front and the h115i rad at the top. Question is: is this going to make a difference justifiable enough to spend 130 $ on a new case? My specs are: Ryzen 7 1700x 1.35v@3.8ghz EVGA GTX 1080 ti sc2 h115i 2 ml140s rear - radiator - two ml 140s front. Gigabyte x370 gaming 5

try booting into safe mode, if everything is fine then is probably a corrupted driver.

So, i was comparing the stock h115i fans with the ML140's (awsome fans, by the way), when i stumbled across some strange (and frankly, impossible) temperature and voltage readings. I have spikes close to 250 celsius, and some -20 celsius on some cores..??? Also i have some spikes on CPU vcore close to 3.5 v.. which i know is impossible.. at least i hope so. Also OCCT is reporting some strange spikes too... any advice on why it's happening? Below some graphs made from the raw data captured by Corsair link, coretemp, and SIV (gigabyte System Information viewer). What i did to record the data was using the "write to file" option in corsair link and coretemp, with the polling rate set to 1 sec. As for SIV, i had to use the record data option, with a 1h duration, and a 5 sec polling rate. I then used OCCT and programmed a 1h stress test, and started everything.

I agree. For gaming not really. productivity/ everything else: absolutely Just make sure you pair it with some fast ram

Why don't you use a monitoring program and log everything to file? Start it, make sure it's logging every single possible parameter (temps, voltages, usage) and start playing or start a stress test. this way you will have a more complete view of what's happening. Make sure it's logging to file at least twice a second. this way if it crashes at least you have a log saved to your hdd/ssd.

Appreciate the explanation yeah, in theory. Honestly, i didn't know a lot of things explained in the video. really informative, thank you. I'll do as you said. LLC to medium, at 1.35 volts. Again, appreciate it, thanks.

Ok so what's the difference between cpu vcore, cpu vid, and cpu vdd in hardware monitor? With llc set to extreme profile i'm touching 1.38, which is +0.02 from the original setting, at 1.35. I think i made a mistake, i did my test always referring to cpu vdd, on hw, and not the actual cpu vcore. But made me think: which is the voltage the CPU is actually using? And, shouldn't the voltage generated from the mosfets be the same as the one the cpu is running at? Below a screenshot of all the tests i did in an excel sheet: i guess everything is normal?

Thank you very much, that explains a lot. Still, i have a friend with a 7700k, and he experienced way less voltage drop. Could be a more powerful vrm? Also, do you think i should leave the load line calibration on auto or manually set it?

First update the bios, and try that way. Make sure everything is stock. Then check on your Motherboard manual which slots are preferable for use in dual channel config. As iHDCool said, try either one or another stick of ram and see how it goes. If it doesn't post try every single slot. Also be aware that Ryzen is particularly tricky with ram speeds. Try default 2133 first, then 2933, then 3200 if you can achieve that, if you didn't try it before. Also make sure the cpu fan is connected into the appropriate fan header (the cpu fan one). Other than that, you could try the single components on another system until you find out which one is the problem.

So this new ryzen build is giving me a hell of a lot of problems. So recently i was playing with it, and i noticed that my vcore voltage drops quiet a lot during stress test... Starring at an empty desktop, 1% cpu usage, no programs opened : vcore at 1.4 volts, 3.9 ghz stress testing in aida64: vcore drops to 1.332/1.335v, freq at 3.9 ghz stable, load line calibration on auto ( i noticed that is the most effective setting in reducing vcore voltage drop). Should i start worrying about my mobo, or is it something normal? to be said, vrm temps stay below 65 c, and cpu never exceeded 69 c

Here is a video showing the issue, if this can help.. https://1drv.ms/v/s!Aq17GnZDZJfi0FMU1XiAdcobh5Cc

Yeah it's depressing. So GTA V it's running from a HDD. OS it's booting up from an SSD. I should receive a 960 pro soon though. The cpu it's locked at 3.8 ghz, using the high performance profile on windows, so it never changes. It's 3.8 ghz at 1.35 v. Ram is running at 2667 Mhz cl 16 dual channel ( stock xmp profile), gpu is on full boost for the entire gameplay, sitting at around 1970 mhz, and after and hour or so it's at around 65/70 C.

It didn't work, i'm afraid The only two things that changed are 1 The 60 fps cap is gone now 2 i'm getting slightly higher fps and gpu usage.. but still between 50-80 %..

nah, I don't think a 1700x can be a bottleneck here... it's usage stays pretty low anyway.

Yep I did it. If that may help, cpu is at 3.8 ghz stable, the 1080 ti is stock freq. both stress tested and stable. Also i recorded a video, with all the metrics displayed using afterburner. if it helps i will upload it.