Launching Steam games causes hard reboot!

[Dr.EM]

Hi all!

I am getting less sure this is a hardware issue but nonetheless, as per the title any steam game seems to cause my PC to reboot when launched, usually after the developer splashscreens etc.

I first saw the issue maybe a couple of months ago when I tried to launch Bioshock Infinite, a game I hadn't ran for some time, PC rebooted and it worked after. But the issue has reappeared more and more often, seemingly starting with the most graphically intensive titles but now happening on everything!

I have had a Google of the issue and certainly I'm not alone in this but I haven't found anything at all definitive on the issue. So far I have downloaded the latest graphics drivers for my card but this didn't help, and tried verifying game data on Steam which passed without faults. I ran SpeedFan and HWMonitor to check temps, SpeedFan reported 128C for CPU eek but only 46C on HWMonitor so I think SpeedFan read an unused sensor or something, physically the CPU felt cool. All temps seemed normal otherwise.

The PC did freeze with BSOD outside of a game yesterday and stalled during startup today but I'm not sure if this is even related.

Anyhow, it was built with parts mostly from here a couple of years ago, here is the spec:

• Sapphire Radeon R9 270X Boost OC 2048MB GDDR5 PCI-Express Card
• Intel Core i5-4430 3.00GHz (Haswell) Socket LGA1150 Processor - Retail
• MSI H87-G43 Intel H87 (Socket 1150) DDR3 ATX Motherboard
• Corsair Carbide 330R Silent Mid Tower Case - Black
• Corsair XMS3 8GB (2x4GB) DDR3 PC3-12800C9 1600MHz Dual Channel Kit (CMX8GX3M2A1600C9)
• Samsung 840 EVO 250GB
• Arctic Cooling Freezer 7 Pro Rev 2 CPU Cooler
• LG GH24NSB0 24x DVD±RW SATA ReWriter (Black) - OEM

Hope someone can help, the issue is clearly worsening and it's a shame because this PC was rock stable and quick for most the time I've had it!

 

[CEUOTC]

What PSU do you have?

It could be the GPU, PSU or the SSD at a guess.

However a quick dirty way would be to remove the GPU post driver removal (DDU lets you do this), connect the monitor to the motherboard HDMI/DP etc then boot up, install intel drivers for iGPU then try and launch a game via steam.

Others might have a more indepth reply soon(tm)

 

[Dr.EM]

Oops missed it off, it's Antec 750W High Current Pro PSU. Should be overkill in a system like this I think.

The Evo SSD has a known issue where access speeds on infrequently used files degrades with time and there is a firmware fix for it, which I've yet to apply, but I doubt it's linked with this? I need to do a system backup before I can do the update really.

 

[CEUOTC]

Aye your PSU is more than enough I would lean more to the GPU being the culprit.

 

[Dr.EM]

It was my initial thought because games will put the greatest stress on it, temps are low and I'm using latest drivers, what should I look at next? To be fair I haven't tried to watch temps during a game bootup, but it might not be possible unless I can find one stable for a least a bit!

 

[Dr.EM]

I tried to run the system up with Intel graphics as suggested but am not having any luck.

I used the Display Driver Utility to remove the Radeon drivers. I then installed intel HD graphics drivers for my CPU. I found no control panel or anything so used the updater tool to do it again and get the latest. This has placed a shortcut to the Intel graphics control panel on my desktop but it never opens. Apparently this might be because there is no DVI connected to the motherboard but I'm doubtful.

In my BIOS under integrated graphics I had to set from 'PEG' to 'IEG' (peripheral to internal graphics?) to get any output from the motherboard DVI. It will boot as far as the login screen where the windows startup sound loops/glitches shortly followed by BSOD. In safe mode it will start but still I can't open the intel graphics controller.

Don't really know what's going on here

 

[CEUOTC]

You have ruled out your GPU, so no harm in refitting it, l would now look to remove the ram and try one at a time in different slots.

 

[Dr.EM]

I suppose so, assuming that it wasn't the Intel installation itself causing a new issue. I'm surprised it wouldn't even boot to Windows using it.

I've just tried each RAM stick in the first slot and games still crash at bootup as before with either. I find it wierd that Bioshock Infinite run 'first time setup' every time, including apparently installing DirectX. I would think if the system had the latest DirectX (11) that it wouldn't ever do that. Mafia 2 also crashed though.

So is it looking like the SSD, or maybe just a software conflict somewhere since everything used to be fine with this hardware? Or potentially the firmware for SSD really does need doing.

 

[OracIe]

Is it the 840 listed above and how old is it? Is it running most recent firmware? Have you tried running Samsung Magician on it?

 

[Dr.EM]

It is Samsung 840 EVO 250GB 2.5 inch Basic SATA Solid State Drive bought at the start of November 2013. I have not yet applied the new firmware or ran Magician.

 

[ED209]

check your ram, run memtest and allow this to run through a number of passes, will display any errors

update the SSD to the latest version as well.

 

[FloppyPoppy]

Seing as you have ruled out several potential causes it may be worth 1, trying to re install steam and seing if this still happens if its only steam games.

2, hooge pain in the bum but might also be worth trying a fresh copy of windows ( maybe on a spare drive? )

 

[olivier renault]

Sound like RAM rot to me.

But the in any case :

- first and foremost, check event log, and try force windows to BSODs and generate minidumps.

And in no particular order...

- run dxdiag.

- if you have any overclock, reduce or run stock everything.

- if it looks like memory, run Memtest86. Bad RAM can trigger all sorts of behaviour, but most of the time, it would just BSOD. Also, simply bumping the memory voltage can bring back stability. Or, reduce memory speed, if only as a test.

- Also, check your SATA cables and PSU cables are not becoming loose. And, if the GPU hasn't come loose in it PCIE slot, and its power connectors.

Try with unplugging everything not important, and run stress tests.

- Go to BIOS and check CPU temperatures. If the cooler is loose, CPU load can trigger the emergency reboot (CPU temp > 100°).

- Stress test components (I use OCCT, Unigine, ATTO benchmark).

If all looks OK, make sure the firmwares are up to date, reinstall windows (I never bother with repair, but you can try).

 

[Dr.EM]

Only have a short time with the PC today but I ran memtest. It completed a pass in about 45 mins with no errors found.

Might be of note that the memory detected in the BIOS as only 1333MHz and when I built the machine I just set it up to 1600MHz myself and left it there since the memory should work at this speed. Maybe a voltage bump was required to support this? Memtest should have ruled it out though, if one pass is enough?

Thanks for all your suggestions, I clearly have a few things still to try out

 

[Mei]

i hope its not raptr lol
that would suck

 

[5960xNvidiaGamer]

Sounds like a software/driver problem to me, I would do a fresh install of windows and see if that sorts it.

 

[Monkeynut]

Definitely fresh copy of Windows.

 

[Pulseammo]

His PSU might be sufficient wattage but it doesn't mean that it isn't faulty. Perhaps when it's under higher load the system is rebooting.

Have you tried something else that'll put load on it such as a benchmark run outside of steam. Try running unigen heaven or valley on extreme settings to see if it reboots. You could even try something like Intel Burn Test or Prime95 to see if it'll reboot just under CPU load.