generic 500W power supply problem or gpu 1080 Asus 8gb Turbo problem

mkoijng

Prominent
Jan 29, 2018
4
0
510
Just did a recent build.

g4560
Asrock b250m pro 4
8gb (4gbx2) 2133 Kingston Hyper fury X
Sandisk PLUS 120gb SSD
Asus 1080 8GB Turbo (over kill but it was free refurb).
windows 10 Pro
Thermaltake Versa h22 case (stock 500W psu)

I installed all the default drivers and everything worked. Started playing World of warcraft, 4 hours in, the computer switched off. Display settings were set at 7/10 (getting that 100FPS sweetspot =] ) Then multiple times on different occasions, get screen freezes and random reboots. Often while playing wow, other times was general browsing with music on and the music distorts then reboots or freezes.

Ran system diagnostics, no errors and temperatures were as follows from immediately booting back up:
cpu: 58 deg C
Mobo: 48 deg C

Further diagnostics: https://imgur.com/a/Vyx5J

does anything seem out of the usual as to why this might happen? Could it be a heat issue or would a psu upgrade do the trick?

Seeking to purchase Antec EA-650G EarthWatts 80Plus Gold 650Watt Power Supply Unit. $89 AUD

would this be sufficient or can you recommend another on this list which is in the $100 AUD area.

http://www.msy.com.au/Parts/PARTS.pdf

I did not check the GPU temperature under load but will stress test tomorrow and see if the system fails. will be running FurMark_1.18.2.0 unless you can recommend another program.

also another anomoly is that the bnet folder corrupted after a fresh install of the game, had issues but after reinstalling a few times, started working again. I doubt it is an SSD failure but thought i'd mention it anyway since everything, save for the 1080, was el budgeto.



 
Solution
No, the generic PSU should be thrown in the trash. The Earthwatts unit would be a FAR better solution than the generic one but honestly that is not one of the better units out there and you might be back here asking this same question again in two years. I'd spend the extra money, if there is even a difference in price, and go with the Truepower classic. That is likely the BEST unit available in your market, with sufficient capacity, below 100AUD.

If you must use the Earthwatts unit, at least it's better, by far, than what you have now.
This is plenty big enough for your configuration, regardless of what CPU or other components you use with that graphics card, unless, perhaps, you are doing some major overclocking of both the CPU and graphics card. Probably, even then. It's also a MUCH better unit than that Earthwatts unit.


PCPartPicker part list / Price breakdown by merchant

Power Supply: Antec - TruePower Classic 550W 80+ Gold Certified ATX Power Supply ($99.00 @ Umart)
Total: $99.00
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2018-01-30 01:27 AEDT+1100
 
thanks DarkBreeze for the hasty reply. just to clarify, are you indicating that the Antec EA-650G EarthWatts 80Plus Gold 650Watt Power Supply Unit is the go ahead or were you stating that the Generic 500W psu (that came stock from the Thermaltake Versa H22 case) is sufficient? There is a $40 AUD difference.

EDIT: Ah i get it, you recommended the Antec 550W truepower. although not outside of my budget, would the antec 650 gold be sufficient?

If the generic psu is sufficient, would it be another peripheral that might be the problem for reboots and screen freezes that you could identify form the information provided.

Again much appreciated.

 
No, the generic PSU should be thrown in the trash. The Earthwatts unit would be a FAR better solution than the generic one but honestly that is not one of the better units out there and you might be back here asking this same question again in two years. I'd spend the extra money, if there is even a difference in price, and go with the Truepower classic. That is likely the BEST unit available in your market, with sufficient capacity, below 100AUD.

If you must use the Earthwatts unit, at least it's better, by far, than what you have now.
 
Solution
Exactly. Or even a GT 1050 for that matter. If you have any kind of discreet gaming card, or even if you don't, you should always have a quality power supply of some kind if you value your hardware and don't want to be buying a new motherboard every two years because the ripple on your generic power supply boiled the caps into oblivion long before it's time should have come.
 
i love you guys, will pick up the Truepower and report if i have any more problems. Hopefully it was just the shitty psu causing that problems. 100% sure the 550W TP is sufficent? The difference betweens the 550 and 650 Truepower is minimal.
 
The 650w Truepower WOULD be likely to place the load demand closer to a level that would keep the unit in it's cool zone, and run more efficiently. It would be quieter and cooler than the 550w, but the 550w is still plenty, some might even say overkill, for that configuration. If you think you may overclock later at some point with a different CPU OR manually overclock your graphics card to any significant degree, then it couldn't hurt to have the extra 100w so long as the extra 100w is on an equally good unit.

An extra 100w from an inferior unit might actually give you LESS capability than a lower "rated" capacity that was simply a superior unit. Many units, when they are very good, can easily do far beyond what they are actually rated for on the label while most mediocre or garbage units can usually not even do their rated capacity.
 
Just a follow up. I bought the Antec 650 Truepower which did not alleviate the problem, however, I agree it was a necessary purchase.

I stress tested the computer GPU and it would only ever hit 50% and restart or freeze the computer. Rather than test all other hardware, I transferred the 1080 GPU to another identical computer which was running fine with an R9 gpu. After install ing the 1080, ran the stress test and again only hit around 50% and froze and reset.

I placed the R9 gpu and a 970 Strix in the new computer. installed drivers and ran the stress test. got to 100% and worked fine.

I think I've isolated the problem and can say that it is a faulty GPU.

Is it strange that I could play with the 1080 for long periods of time where it will reset/freeze? Also is it strange that only when pushed past a certain point it dies? Wouldn't a broken GPU just not run? I liken this to driving a car up to 60km is fine but as soon as I go past 60km+ it stalls.
 
Naw, graphics cards can exhibit anything from only minor glitches to completely not working, and everything in between. Unfortunately, it's hard to pin down without sometimes eliminating everything else first OR trying it in a different machine like you did. Sucks that the card is bad, but hopefully it should still be under warranty and I'd start an RMA right away.