First of, here is my whole configuration. I will try to provide as much details as I can.
PSU : EVGA 600 BQ 80+ Bronze (600w)
MB : Asus Z170-A (LG1151 socket)
CPU : Intel Core i5-6600k
GPU : MSI GeForce GTX1060 6GB Gaming X Plus (9GHz memory speed)
COOLER : Cooler Master Hyper EVO 212 (single fan, push config)
RAM : Crucial CT8G4DFS8213.M8FA 1x8GB
SSD : Samsung 850 EVO 250GB (AHCI)
HDD : Some WB blue just for file storage
Monitor 1 : Asus 24" IPS panel 5ms connected via HDMI
Monitor 2 : LG 19" TFT panel 5ms connect via VGA (GPU > Display Port to VGA active adapter > VGA > Monitor)
Mouse : Genius Deathtaker
Keyboard : Genius K9 Gaming
PC CASE : Cooler Master MasterBox 511 RGB
Fans : 3 pressure fans pushing air in from front of case, 2 airflow fans pushing air out of case (one at end of case, and one on top)
Fans are DC, not PWM, all properly configured in BIOS, 3 fans are on separate headers, and 2 are on splitter (splitter is default noctua splitter that you get from buying their fans)
Hardware temperatures (all temperature are measured by HWiNFO64 while ambient temperature is 21-23c) :
CPU idle : 24-26c
CPU heavy load (prime95 26.6) : 50-54c
GPU idle : ~35c (fan off by default)
GPU heavy load (heaven, furmark, superposition) : not a SINGLE time reached more than 64c, and that 64c was one peak, usually it sits between 62-63c (fans start spinning at 61c)
Sorry for long post, but want to make sure you have all informations, issue is following:
I had Gigabyte GTX950, sold it, and bought used GTX 1060 from above. Before installing new GPU, I went into safe mode and used DDU to wipe all GPU drivers from system, then installed new GPU and all drivers for it from GeForce Experiance.
New GPU is here, everything is great, played Apex Legends on everything ULTRA (except textures, because ultra textures requires 8GB VRAM, and my GPU is 6GB VRAM), pushing between 80-144 FPS (Played for 2 hours). Then I played bit of warframe, also everything on ultra but with vsync ON to 60Hz(fps).. not a single issue, after that I start CSGO and my PC freezes completely with audio glitch. By that I mean, the last sound I heard milisecond before freezing is the sound that is repeating constantly. Nothing works, keyboard, mouse, even restart button on my pc case is not working. Only way is to hold power button to force shutdown or switch PSU switch to off.
Friend told me that CSGO is built on ancient engine, and that MAY be some driver conflict... okay, I did FULL Windows 10 wipe, by that i mean full reinstall from USB, and formatting ssd from scratch. First thing after new windows install, installed Asus chipset drivers, then everything else for motherboard (just drivers, not a single software, with exception of Realtek HD Audio Manager), then GeForce Experience, then latest NVIDIA drivers from GFE. After that Steam and Origin (Warframe, CSGO, Apex Legends).. that was left overnight to download... when I woke up, i started first CSGO since that's the one that kept my PC freezing, played for and hour.. no freeze. Great, I thought my problems were solved... Then later that day I start Warframe (when you install warframe on new system it launches at 800x600 windowed mode) entered email and password, submit, PC FREEZE. At that point I'm already blaming GPU, since with previous GPU didn't had ANY errors nor freezes. Took my GPU to friends PC, we stress tested it for 1h, not a single issue. Came back home, put GPU in my system, load up windows, warframe, PC FREEZE. That's not good enough... friend brings his Zotac GTX 1080 mini, we put it in my system, start up warframe, PC FREEZE, restart pc, start CSGO, PC FREEZE.
I really don't think it's faulty GPU unit, because it's simply nearly impossible that 2 different GPU's from 2 different vendors, freezes system in exactly same way. If it was ANY other type of crash, I would return GPU instantly.. but like I said... 2 different GPU's on my system, exactly same freeze.
Things I tried :
-Full Windows 10 reinstall
-Power Options set to "High Performance"
-Reinstalling GPU drivers
-Disabling 2nd monitor (19" one) that had adapter, and used only main one with straight HDMI cable
-Tried with another monitor with native DVI-D with straight DVI-D to DVI-D cable
-Removing ALL and ANY types of overclocks (I do my CPU overclock via XTU software)
-BIOS load optimized defaults
-BIOS battery remove
-Reseating RAM memory to all 4 available slots
-Unplugging HDD and using SSD only
-Switched SSD from SATA Express 10Gbps lane to default SATA 6Gbps lane
-Disabling iGPU in BIOS
-Ran Prime95 26.6 for 30 minutes, no errors
-Ran Heaven for 1h, no errors
-Ran MemTest86 for full 4 cycles (13 test, 4 times ~ 2h30m), no errors
-Setting DRAM Frequency from AUTO to DDR4-2133MHz (that's my RAM speed)
-Cleaning GPU and RAM connectors on motherboard (canned air cleaning)
-Powering my PC separately from monitors, other power outlet
-Powering my PC from different source (not on same fuse as my TV and monitors, fuse that doesn't have anything drawing power from it)
-Powering my PC directly to power outlet (without any extension cords)
-Unplugging all USB devices, just leaving ethernet, keyboard, mouse and power cable
-Scanned entire PC for viruses (Licenced ESET Internet Security 18.104.22.168 [latest version with all updates] that I've been using for more than 2 years)
-Every Windows 10 update
-Disabling Virtualization in BIOS
-Inspecting all capacitors on motherboard (all are fine)
-Plugging keyboard and mouse to USB2.0
All of that and freezes still happens. But the thing is, it's full freeze-lockdown, there is no BSOD, no errors, so I can't even read what's causing the issue because it's just freezing, it doesn't give any error, I tried to leave it frozen for 30 seconds to see if BSOD will show up, but it didn't. When PC freezes, on my other monitor I have GPU-Z and HWiNFO64 opened, when freeze happens (minimum-maximum stats by HWiNFO64) :
CPU temp 30c-44c
CPU Vcore : 1.184v-1.248v
CPU fan : 730RPM-893RPM
CPU usage : 0.07%-100%
GPU temp : 55c-61c
GPU Vcore : 0.625v-1.063v
GPU fan : 0RPM - 844RPM (0%- 30%)
GPU Core Load : 0%-72%
5 other chassis fans from ~760-920RPM
SSD temperature 28-34c
All cores thermal throttling : NO - NO
All cores critical temperature : NO - NO
All cores Power Limit : NO - NO
Package/Ring Thermal throttling : NO - NO
Package/Ring critical temperature : NO - NO
Package/Ring Power Limit Exceeded : NO - NO
CPU Package Power : 18.488W - 50.488W (TDP 91W)
GPU Power : 8.837W - 128.555W (TDP 125W)
My motherboard doesn't have VRM temperature sensors, in HWiNFO64 they introduced that meter (VR VCC) for Hasswel CPUs, and it's showing for me as 79c-84c, but when I touch VRMs with my finger, they are MAXIMUM 40c, because I can hold my finger on them and it won't burn me, it will not cause that heat, that I have to move my finger (home heating system is at 60c, so I know how it feels when you touch 60c radiator).
Latest thing that I tried was removing 1x8GB Crucial RAM and putting in 4x4GB Kingston HyperX DDR4-2666MHz (I had to borrow it from friend, it's not mine) and since then, I've been actively playing Warframe for about 1h, and CSGO for 30 minutes, no freezes. I left my Warframe running in background (single player instance with heavy open world RAM usage) AND started CSGO just in main menu... Warframe was in background for 4 hours WITH CSGO for about 1 hour... No freezes. Now, I don't want to jynx it, but I highly think it's RAM memory fault, because since I put Kingston, didn't had any freeze (PC uptime with games 5h20m), that's very little time to test if it's RAM fault or not because even with Crucial RAM it happend that I play games for 2,3 hours and no freezes, then I start game and PC FREEZE.
1. Is it possible that Crucial RAM was faulty, but GTX 950 wasn't able to utilize those sectors because of being lower end GPU?
2. Is it possible that GTX 1060 is more aggressive and needs dual channel?
3. How is possible that only thing I changed was GPU and those issues start to happen, and it's not GPU fault?
4. Is it possible that MemTest86 needs more than 2 hours to find RAM issue?
5. I've read somewhere that MemTest86 is there to report hardcore error, it's not reliability of RAM test, true?
6. I left my PC running overnight, just on desktop mode, and when I wake up, it's working, not frozen, IF RAM is culprit wouldn't that happen anyway?
I will continue to test throughout the day, since it's 5am currently, and will report if anything strange happens. I know I made reaaaly long post, but I wanted this to be in details as much as possible in order to narrow down the issue and get as much input from you guys, your thoughts, opinions, advises, suggestions...
Thank you for reading this until the end
PS : I actually had to type and save in text editor all text in order to save what I typed if I get freeze, but I didn't. Also important note
: freezes happens only during launching the game, or playing the game, it did not happen in desktop/internet browsing/youtube, etc...