I recently built a machine and I'm currently at my wits end with how to fix it at the moment. It shuts off at random, though the two times I've been able to notice it more often is when I'm playing games at 1080p (medium to ultra settings) or when I'm watching a video at full screen. It doesn't happen instantly, it will happen anywhere from 20 minutes to two hours in, without any consistency.
At first, I thought that the issue was heating, so I upgraded the heatsink/cooler to a CM 212+ Universal with one fan and my CPU stays in the 30C (idle) to 45C or 50C (high setting gaming) range. The video card stays between 50-60C when under load, or 33 idling. From what I understand, these temps are actually pretty solid and in a good range. No overclocking at all yet. I have temp logs I can post from AIDA64 if that would help.
I have a EVGA Superclocked GTX 580, Asus P8Z68 V-Pro mobo, 16gb G.Skill Sniper ram and a 750W Gold Corsair PSU (which I thought would be more than enough power).
Before I shell out another $200+ for a new, larger PSU, can anyone suggest if that is the right course of action, or if something else could be the problem?
Thanks for your help, I really appreciate your time and suggestions.
what cpu are you running and have you OCd anything at all?? one thing i would look at is ram. every asus ive worked on has puked with gskill ram. run memtest86 and let it run for a while (6+ hours with that much ram) but that power supply is fine. also take a look at the event log and post any errors or warnings that are listed
Running i7 2600k - no overclocking. Will run memtest tonight if I can.
Looking at my error log, the error ids I have are 10 (Source: WMI), and every once in awhile 4107 (Source: CAPI2). Nothing I can see that would point to a shutoff worthy error. Warning IDs are 100, 1008 and 1530, which all look search or registry related.
Would Corsair or Kingston be a good choice of RAM for an asus board?
Failsafe defaults puts your RAM at the default values set in the BIOS. For example, your RAM may be 1033 MHz, CL9, and 1.5V at default values. These are the values your RAM will be tested at. If these were the values you would always run your RAM at, then it's fine; however, if you're like all the other performance enthusiasts, you'd want to run your RAM as fast as your system will allow, OC or not.
Ideally, the best way to test your RAM is at the specs shown on the RAM's stickers, and one at a time. This is because in the event of failure, you'll be able to deduce which RAM module is faulty.
Thanks for that explanation. The RAM is 1600 MHz (I'll double check the rest before I run Memtest86 again when I crash for the night).
As far as testing one at a time, is there a way to do this in Memtest, or would I need to take out all but one stick and run them individually?
Today, I've had the computer on for hours doing a variety of low-intensity work — programming, building some HTML/CSS, creating e-books. Also doodling in Photoshop at high-res. No problems with that. However, when I tried streaming video even at lower res like 360P (screen was maximized though), the reboot would happen, often within 10-20 minutes or less. Would a RAM problem be related to this at all?
- Flashed 0706 bios last night (was 0606)
- Manually set the right speed on RAM instead of putting it on Auto
- Ran MemTest86 for 7 hours, no errors.
- Ran Prime 95 for about 45 minutes today. Max CPU temp of 64C, did not reboot at any point.
- Ran FurTest 1.91 for 30 minutes today, max GPU temp 61C, max CPU 50C, did not reboot at any point.
- Streamed a 2 hour film, no shut down.
- Been working in Maya all day, no shutdown yet (there had been last night while doing so).
Fingers crossed, this is working. Will try gaming later today, after work is done
Thanks everyone for the help so far. Hopefully I won't need to be back to this thread except to choose a best answer