What's new
  • Please do not post any links until you have 3 posts as they will automatically be rejected to prevent SPAM. Many words are also blocked due to being used in SPAM Messages. Thanks!

(z170) computer crashs upon PCIe GPU load

Prickly007

Well-known member
Joined
Jul 8, 2014
Messages
809
Location
Metro-Vancouver
After installing my new CPU-MB-RAM and then Win 8.1 Pro on Wednesday, I finished softwaring the next night and then went to play a game, Rise of the Tomb Raider. Alas, I encountered the same problem as before, crashes that shut down the computer followed by a reboot. However, the crashes now occur immediately upon loading the main menu or when starting a benchmark, i.e. Heaven (and Kombuster, as well). As I discovered this afternoon, the render test in GPU-Z runs fine. System passed a quick run of Intel Burn Test (x10 @ v. high), with max temp of 72. GPU temps are not the issue.

After researching the issue well into the night, some of the things I tried include:

Cleared CMOS (not that I changed much)
Updated BIOS to latest version
Ran Winupdate
Turned XMP off, thereby setting RAM to 2133MHz
Under-clocked GPU, -50 on both core and memory
Set power plan to performance
Messed around with TDR settings, mainly TdrDelay
Reseated GPU and change power cable

All to no avail. One thread suggested, in System Agent Menu, changing Primary display from auto to PCIe and setting PCIe slot to Gen2/3 from auto... It still crashed, but Heaven actually ran long enough to complete the first scene. Could a BIOS setting really be the cause? Might the problem be the most recent BIOS (1601), board came with 1402. So few people have the Hero Alpha, Goggle isn't much help.

Enough was enough, so this afternoon I re-installed Windows again. But only installed the Nvdia driver (v.362.00) and Heaven, and it still crashed right away. I am at a loss! A hardware issue seems the most likely, but what? PSU?* Bad Motherboard?

* I had crashes on my old Phenom II x4 system, but increasing Vcore seemed to (temporarily) stop them; plus they were far more random. Since last December or so, I had to increase Vcore and/or CPU-NB half a dozen or so times, hence the reason I upgraded.
 
Last edited:

Prickly007

Well-known member
Joined
Jul 8, 2014
Messages
809
Location
Metro-Vancouver
Moved* to the x8 slot, same result. Thanks for the offer moocow!

No one thinks it might be a bum PSU? That always seems like a popular diagnosis across the interwebs.

* A time when, working in a case with a NH-D15, comically small Trump-hands are an asset! :haha:
 

Bond007

Well-known member
Joined
Jun 24, 2009
Messages
5,085
Location
Nova Scotia
Moved* to the x8 slot, same result. Thanks for the offer moocow!

No one thinks it might be a bum PSU? That always seems like a popular diagnosis across the interwebs.

* A time when, working in a case with a NH-D15, comically small Trump-hands are an asset! :haha:
Given this is reoccurring from a previous build as well, I would say likely the psu. Probably worth trying another gpu if you have access to one, but I would t have high hopes. I would think it's unlikely to be the motherboard, but also possible.
 

Prickly007

Well-known member
Joined
Jul 8, 2014
Messages
809
Location
Metro-Vancouver
Did a little more digging after I was able to play CS:GO for a few minutes.* Either the GPU can't handle high loads, CS:GO (using the settings from my old system) doesn't even hit 50% usage.

Also looked at the load (in Watts) on my UPS; not the best measurement I know, but it's all I have. At idle it's c. 100, Prime95 and Intel Burn Test hit 195/200-ish fairly quickly. CS:GO just now fell peaked at 210 Watts. I was able to find a few screen shots from my old system, when I was monitoring temps in more demanding games for adjusting Speedfan settings, and they drew anywhere from 350 to 425-ish Watts.

In fact, Rise of the Tomb Raider was highest, so I take this to mean it's time to order a new PSU and, if that fixes the issue, RMA my HX850. If not, RMA the GPU.

* Forgot to back up any other games. :doh:
 

Bond007

Well-known member
Joined
Jun 24, 2009
Messages
5,085
Location
Nova Scotia
no other pc you can try the GPU in, or other GPU to try in your system? That could save you some time and effort buying parts
 

Prickly007

Well-known member
Joined
Jul 8, 2014
Messages
809
Location
Metro-Vancouver
Only other PC would be the Phenom system that also had an issue with crashes, sadly I sold my old HD 6870; the first time I ever sold a decommissioned part. Oops! :haha:

I've started researching new models- damn insomnia- and very quickly came across a few people with Corsair HX models that also died at the 5+ year mark; like me, they left their system on 24/7. To be honest, when I was researching my upgrade options, I had been thinking that it might be time to pro-actively replace my PSU lest it die and possibly damage other, much newer, parts. But I opted not to be prudent and spent the money on a 6700k chip.
 
Error 403 Daily Limit Exceeded. The quota will be reset at midnight Pacific Time (PT). You may monitor your quota usage and adjust limits in the API Console: https://console.developers.google.com/apis/api/youtube.googleapis.com/quotas?project=345236864347 : dailyLimitExceeded

Latest posts

Twitter

Top