Computer will crash/freeze while playing games far below my own Specs

[KingCrimsonTimeDelete]

Hi all,

A couple of months ago, I purchased a new rig from scratch and built it together with a friend of mine who builds custom PCs as a hobby. Everything usually seems to work pretty fine, as I'm able to use my computer for hours on end and usually play most games indefinitely. However, particularly with heavier games, my PC will crash or freeze at what is seemingly randomly. I initially suspected temperatures (as this initially seemed to happen after long gaming sessions), but after paying attention to temperatures through Hardware Monitor, this doesn't seem to be the case. My CPU rarely, if ever, passes 40 degrees Celsius and my GPU reaches a maximum of 70 at max load. Furthermore, games would crash after perhaps two minutes of play, so that can't be due to temperatures.

A second issue is that my rig seems to perform greatly beneath other, almost identical rigs. An example being Star Citizen, where a friend of mine with a near-identical rig (difference been an i7-6700k or something, I believe) would be able to play it at 1440p on max settings at 60fps, whereas I could barely reach 60 on 1080p lowest settings. Seemed a pretty big enough gap to warrant a scan for both software and hardware issues

Admittedly, I have no idea where to start checking for problems. Do my issues originate due to software failures? Hardware? How do I figure out where the core of my problem lies? Where do I start? I have no idea how to tackle this. I was hoping for some guidance.

Here are my PC specs. Nothing is overclocked. Latest GPU drivers downloaded, windows updates as well, Bios update done in the past.
http://uk.pcpartpicker.com/list/zkMkwV

(Note: I messed up and posted this in Tom's Guide sub-forum. Trying to find a mod to contact so I can get that one deleted. Ignore that please.)

 

[greens]

Hi all,

A couple of months ago, I purchased a new rig from scratch and built it together with a friend of mine who builds custom PCs as a hobby. Everything usually seems to work pretty fine, as I'm able to use my computer for hours on end and usually play most games indefinitely. However, particularly with heavier games, my PC will crash or freeze at what is seemingly randomly. I initially suspected temperatures (as this initially seemed to happen after long gaming sessions), but after paying attention to temperatures through Hardware Monitor, this doesn't seem to be the case. My CPU rarely, if ever, passes 40 degrees Celsius and my GPU reaches a maximum of 70 at max load. Furthermore, games would crash after perhaps two minutes of play, so that can't be due to temperatures.

A second issue is that my rig seems to perform greatly beneath other, almost identical rigs. An example being Star Citizen, where a friend of mine with a near-identical rig (difference been an i7-6700k or something, I believe) would be able to play it at 1440p on max settings at 60fps, whereas I could barely reach 60 on 1080p lowest settings. Seemed a pretty big enough gap to warrant a scan for both software and hardware issues

Admittedly, I have no idea where to start checking for problems. Do my issues originate due to software failures? Hardware? How do I figure out where the core of my problem lies? Where do I start? I have no idea how to tackle this. I was hoping for some guidance.

Here are my PC specs. Nothing is overclocked. Latest GPU drivers downloaded, windows updates as well, Bios update done in the past.
http://uk.pcpartpicker.com/list/zkMkwV

(Note: I messed up and posted this in Tom's Guide sub-forum. Trying to find a mod to contact so I can get that one deleted. Ignore that please.)

Power Supply
Corsair RMx 650W 80+ Gold Certified Fully-Modular ATX Power Supply

It isn't a bad Power supply, but given your descriptions - this is the component i'd like you to swap out first.
Even if you just go to a local brick and mortar to grab a comparable PSU just for testing - you can always return it and start an RMA on your corsair.

You don't have any typical errors or codes, no BSOD or crash dumps. Often what happens is the GPU gets a bit starved for power, and crashes. The rest of the system is still OK, some PCs are able to recover from this after a moment or continue to play sound, but regardless, the crashing is usually caused by a lack of amps on the 12v rail to the GPU.

 

[KingCrimsonTimeDelete]

Power Supply
Corsair RMx 650W 80+ Gold Certified Fully-Modular ATX Power Supply

It isn't a bad Power supply, but given your descriptions - this is the component i'd like you to swap out first.
Even if you just go to a local brick and mortar to grab a comparable PSU just for testing - you can always return it and start an RMA on your corsair.

You don't have any typical errors or codes, no BSOD or crash dumps. Often what happens is the GPU gets a bit starved for power, and crashes. The rest of the system is still OK, some PCs are able to recover from this after a moment or continue to play sound, but regardless, the crashing is usually caused by a lack of amps on the 12v rail to the GPU.

Oh, is 650W underpowered for the rig I got? I was told it was plenty, which is why I bought it. I do get BSODs and whatnot, but when I go to event viewer, all I can see is a notification that my computer closed down unexpectedly. I unfortunately did not have enough time to note down the BSOD error message that appeared. How can I recover crash logs?

I'd have to purchase a stronger PSU, reattach everything and hope that the crash happens during my testing time (happens a bit erratically and I'm starved for time, so it can go days without crashing). I was hoping I could somehow try and whittle down the possible sources before I start swapping hardware.

To be specific in regards to crashes, games themselves would never crash. Rather, my PC would freeze or get a BSOD.

 

[greens]

Power Supply
Corsair RMx 650W 80+ Gold Certified Fully-Modular ATX Power Supply

It isn't a bad Power supply, but given your descriptions - this is the component i'd like you to swap out first.
Even if you just go to a local brick and mortar to grab a comparable PSU just for testing - you can always return it and start an RMA on your corsair.

You don't have any typical errors or codes, no BSOD or crash dumps. Often what happens is the GPU gets a bit starved for power, and crashes. The rest of the system is still OK, some PCs are able to recover from this after a moment or continue to play sound, but regardless, the crashing is usually caused by a lack of amps on the 12v rail to the GPU.

Oh, is 650W underpowered for the rig I got? I was told it was plenty, which is why I bought it. I do get BSODs and whatnot, but when I go to event viewer, all I can see is a notification that my computer closed down unexpectedly. I unfortunately did not have enough time to note down the BSOD error message that appeared. How can I recover crash logs?

I'd have to purchase a stronger PSU, reattach everything and hope that the crash happens during my testing time (happens a bit erratically and I'm starved for time, so it can go days without crashing). I was hoping I could somehow try and whittle down the possible sources before I start swapping hardware.

Well like i said, that PSU SHOULD be fine, but your problems make me think that it might not be working 100%.

If you're getting BSOD, have you messed with your settings?
You may need to supply the CPU more voltage if you're getting BSOD. Did you enable XMP profile for your RAM?
Have you messed with the BIOS at all? I find that with my custom builds lately, I have to manually switch a number of settings before things run at 100% and do so smoothly. Honestly now that i think about it, my Skylake builds have all required settings beyond default to run well.

 

[KingCrimsonTimeDelete]

Well like i said, that PSU SHOULD be fine, but your problems make me think that it might not be working 100%.

If you're getting BSOD, have you messed with your settings?
You may need to supply the CPU more voltage if you're getting BSOD. Did you enable XMP profile for your RAM?
Have you messed with the BIOS at all? I find that with my custom builds lately, I have to manually switch a number of settings before things run at 100% and do so smoothly. Honestly now that i think about it, my Skylake builds have all required settings beyond default to run well.

Only BIOS settings I've messed with were those for the case fans. I haven't changed anything else. I'll see how I can enable this XMP profile, I don't really know what most of those things do, so I was too afraid to change anything

 

[greens]

Well like i said, that PSU SHOULD be fine, but your problems make me think that it might not be working 100%.

If you're getting BSOD, have you messed with your settings?
You may need to supply the CPU more voltage if you're getting BSOD. Did you enable XMP profile for your RAM?
Have you messed with the BIOS at all? I find that with my custom builds lately, I have to manually switch a number of settings before things run at 100% and do so smoothly. Honestly now that i think about it, my Skylake builds have all required settings beyond default to run well.

Only BIOS settings I've messed with were those for the case fans. I haven't changed anything else. I'll see how I can enable this XMP profile, I don't really know what most of those things do, so I was too afraid to change anything

No Worries. With the way UEFIs are built these days, it would be pretty much impossible to cause any kind of permanent damage.

Your system uses very nice components by the way, looks like a high quality build!

Enable XMP 2.0 profile. That loads up all the OC RAM settings automatically for you.
I'd like you to increase your CPU voltage. There are a number of ways to accomplish this, either manually configuring or using an offset. What we really want to do is make sure it doesn't drop below 1.35 for testing purposes. We can back it down from there, but for testing purposes 1.35 should clear up any low voltage issues and presents zero risk to your system.

 

[KingCrimsonTimeDelete]

I've enabled the profile. Do I change the cpu voltage at the "Cpu Core/Cache voltage"? It's at 1.136V right now.

 

[greens]

Yeah up that to 1.35 and push your PC as hard as you can to see if you can't get a crash.

 

[KingCrimsonTimeDelete]

All right, will do (I'll change that and only that). CPU stress test of 15 minutes enough? Or do you mean playing a heavy game?

 

[greens]

Either way, the best thing would be something you know was giving you grief before. If there was a specific test that always failed or a game that always crashed, go with that.

 

[KingCrimsonTimeDelete]

Either way, the best thing would be something you know was giving you grief before. If there was a specific test that always failed or a game that always crashed, go with that.

Yeah but that's kind of the problem I have. The crashes are extremely, extremely random. I've had moments where everything would work fine for multiple days with a certain game, but the same game would crash several times on some other day.

I'll spend some time in the game and see if it crashes, but unfortunately I can't really confirm anything through it until it actually crashes.

Also, since I've enabled the XMP 2.0 profile, I've been hearing occasional "popping" sounds while watching youtube videos. This only started happening once I changed this specific setting.

 

[KingCrimsonTimeDelete]

Well, all I can say is that it hasn't crashed so far, so unless that changes in the future, I guess that fixed it. What does that imply, though?