• Happy holidays, folks! Thanks to each and every one of you for being part of the Tom's Hardware community!

[SOLVED] random Pc restarts RTX 2060

Feb 22, 2022
8
0
10
Hello, I just got a new gpu a Asus dual RTX2060 - 06G EVO
the heaviest thing I have ran on it was ARK survival evoled on high settings, and I got some temperature from 69c to 84c
and 40c to 54c in idle.

so here is the current situation, I installed it, install the latest driver and then I tried a couple of games, like Sekiro on ultra, Metal gear revengeace, divinity 2 and the previously mentioned Ark. I usually play for about 2 to 4 hours at most.
while Running games I don't have any problem whatsoever.
However, while wathching youtube specifically when I just leave a youtube video and I just don't interact with the PC I just leave the videos playing on the background. Never while gaming, the pc just restarts itself. No Blue screens, no artifacts, nothing weird happens it just restart as if someone pull the plug.
I usually leave my pc on for about 18 hours a day, this has happened after about 2 hours, or sometimes after 10 hours.

my specs
win 10 64bit
Intel Core i5 8400 2.80Ghz 6Cpus
24576 MB Ram
GPU -> Asus dual RTX2060 - 06G EVO
Mother board -> gygabyte b360 Auros gaming 3 wifi
PSU -> thermaltake TR2 600W atx 12v 2.3
Model: TR2-600NL2NC

my previous GPU was a GTX 760.
I used a PSU calculator for all my components and it said I needed 358 W, the PSU is 4 years or older I don't really remember.
I have read similar problems on this site, and it would seem that the PSU is usally the issue but I'm an amateur so, I ask for any feedback you can give.

thanks.
 
Solution
1) sorry I really don't understand what the numbers mean.

2) so the values are bad?

3) can I improve them somehow?

4) or does it all goes donw to the PSU again?
1) The nrs speak for themselves, but the example, your +3.3V is at 3.542V in that software, the max allowed in that stat is 3.465V which is lower, so you see that current value is above max value allowed.

2)Yes.

3) No.

4) Yes.


Have to say that this is measured by software through the sensors which can be inaccurate, best way to measure is by multimeter, but measuring the 3.3V is only possible by using the right pins at the 24 pin connector.

▶ Manually Test a (PSU) Power Supply With a Multimeter by Britec - YouTube

still i would be replacing the psu...
Looking at what happens and that TR2 from Thermaltake which just aren't very good quality do i think it's the psu. Especially if it's already 4 years old, would replace that first.


Other thing i noticed is the ram configuration, with 24gb you either have 3x8gb or 2x8+2x4gb, either way you likely mix ram which can give problems too, but first bet is psu.

You have the latest bios?
 
Thanks for the answer,
first, yes my RAM is mixed, 2 sticks of 4 and 2 sticks of 8. I've had this configuration since more than 8 months I think. I never knew that such a thing could cause trouble. should I take out the sticks of 4?
as for the BIOS, more than likely NO. unless there is some kind of automatic process for that and I did not realize.
 
Well i would first look at that psu as said.

Mixing ram can give problems, but if it is running stable with those four sticks for 8 months might it well be alright, you could run Memtest86 overnight (at least 4 passes) and see if it gives problems or let Aida64 run for a while and only check ram.

What bios do you run? Can look via "dxdiag" , first page should say what bios you run. Rightclick Windows start button and choose search or run to make it work.
 
Sorry for asking so much, I ran the Aida64 stability test for 20 minutes and I think everything is fine, this are the results of that, and also a screenshot of the BIOS information
https://ibb.co/NNt0kRt
https://ibb.co/ys132kj
lastly I tried to run a gpu test and I got this message, I really did not understood it so I refrained from trying.
https://ibb.co/5Y1NgCv

once again thanks for everything and sorry for the trouble
 
Sorry for asking so much, I ran the Aida64 stability test for 20 minutes and I think everything is fine, this are the results of that, and also a screenshot of the BIOS information
https://ibb.co/NNt0kRt
If those values are good would you be out of ATX spec on the 3.3V rail, The pc should probably shut down because of those, but as said focus on the psu first and get something better.

3tviM1R.png




Looking at that do you run a very old bios and would i really update to the latest.


lastly I tried to run a gpu test and I got this message, I really did not understood it so I refrained from trying.
https://ibb.co/5Y1NgCv

once again thanks for everything and sorry for the trouble
Reading some about that and seems that AIDA64 can up a certain value in the register that might help with video driver recovery. Think i would let it try that, you can also do it yourself,
https://superuser.com/questions/123...timeout-detection-and-recovery-delay-tdrdelay
For a registry back up,
.
might be good to first do a back up of the registry before adding values or letting AIDA64 make such changes.
 
Last edited:
Ok BIOS are update now are F15 the latets.
I've also change the registry to include the tdrDelay parameters.
so for now I'll keep monitoring my pc beahaviour. it might take a while before I can get a new PSU but maybe I can borrow one from a friend, just to test.
Also, regarding this

If those values are good would you be out of ATX spec on the 3.3V rail, The pc should probably shut down because of those, but as said focus on the psu first and get something better.

sorry I really don't understand what the numbers mean.
so the values are bad?
can I improve them somehow?
or does it all goes donw to the PSU again?

once again thanks a lot.
 
1) sorry I really don't understand what the numbers mean.

2) so the values are bad?

3) can I improve them somehow?

4) or does it all goes donw to the PSU again?
1) The nrs speak for themselves, but the example, your +3.3V is at 3.542V in that software, the max allowed in that stat is 3.465V which is lower, so you see that current value is above max value allowed.

2)Yes.

3) No.

4) Yes.


Have to say that this is measured by software through the sensors which can be inaccurate, best way to measure is by multimeter, but measuring the 3.3V is only possible by using the right pins at the 24 pin connector.

▶ Manually Test a (PSU) Power Supply With a Multimeter by Britec - YouTube

still i would be replacing the psu, than again i cannot look into your wallet.
 
  • Like
Reactions: peace123456
Solution
Thanks for everything, at least now I understand more about my pc.
One last question. My ableviable PSU options are these:
Coolermaster Fuente Poder 500W
MODEL: MPW-5001-ACAAN1
ATX VERSION: ATX 12V V2.31
PFC: Active PFC (>0.9 typical)
INPUT VOLTAGE: 100-240V
INPUT CURRENT: 10-5A
INPUT FREQUENCY: 50-60Hz
DIMENSIONS (L X W X H): 150 x 140 x 86 mm
FAN SIZE: 120mm
FAN BEARING: Sleeve Fan
POWER GOOD SIGNAL: 100 – 500 ms
HOLD UP TIME: >10ms at 100% Full Load
EFFICIENCY: 75% @ Typical Load
MTBF: >100,000 Hours
OPERATING TEMPERATURE: 10-40°C
PROTECTIONS: OVP, OPP, OTP, SCP
REGULATORY: TUV, cTUVus, CE, FCC, BSMI
MB 24-PIN CONNECTORS: 1
CPU 12V 4+4 PIN CONNECTORS: 1
SATA CONNECTORS: 5
PERIPHERAL 4-PIN CONNECTORS: 3
PCI-E 6+2 PIN CONNECTORS: 2
SKU: MPW-5001-ACAAN1-US
Coolermaster Fuente Poder 650W
ATX VERSION: ATX 12V V2.52
PFC: Active PFC
INPUT VOLTAGE: 100-240V
INPUT CURRENT: 10-5A
INPUT FREQUENCY: 50-60Hz
DIMENSIONS (L X W X H): 140 x 150 x 86 mm
FAN SIZE: 120mm
FAN BEARING: HDB
FAN SPEED: 2400 RPM
NOISE LEVEL @ 20%: 12.82 dBA
NOISE LEVEL @ 50%: 20.4 dBA
NOISE LEVEL @ 100%: 41.2 dBA
EFFICIENCY: 80% Typically Load
80 PLUS RATING: 80 PLUS Standard
ERP 2014 LOT 3: Yes
OPERATING TEMPERATURE: 40°C
POWER GOOD SIGNAL: 100 – 500 ms
HOLD UP TIME: >14ms at 100% Full Load@230Vac
MTBF: >100,000 Hours
PROTECTIONS: OVP, OPP, OTP, SCP, UVP
REGULATORY: CB, TUV, cTUVus, CE, FCC, CCC, EAC, RCM, BSMI, KC
ATX 24-PIN CONNECTORS: 1
EPS 4+4 PIN CONNECTORS: 1
EPS 8 PIN CONNECTOR: 1
SATA CONNECTORS: 6
PERIPHERAL 4-PIN CONNECTORS: 3
PCI-E 6+2 PIN CONNECTORS: 4
80 PLUS: Standard
MODULAR: Non Modular
WATTAGE: 500 to 750W
SKU: MPE-6501-ACAAW-US
Corsair Fuente Poder 650W CV65
Weight: 1.9
Adjustable: Single/Multi 12V Rail No
ATX Connector: 1
ATX12V Version: v2.31
Continuous power W: 650 Watts
Fan bearing technology: Sleeve
Fan size: mm 120mm
MTBF hours: 100,000 hours
Multi-GPU ready: No
Warranty: 3 Year
PSU Form Factor: ATX
Zero RPM Mode: No
EPS12V Connector: 2
Floppy Connector: 1
Modular: No
PCIe Connector: 2
SATA Connector: 7
SKU: CP-9020236-NA
EVGA Fuente de Poder 500W
ATX Cable 1x 550mm
EPS Cable 1x 600mm
PCIE Cable 1x 550mm, 670mm
SATA Cable 2x 450mm, 570mm, 690mm
Cable Periférico de Cuatro-Pin 1x 450mm, 570mm, 690mm, 810mm (f)
Adaptador Flexible 1x
AC Power 1x 1200mm (m)
Tipo de Cable Modular No

24 Pin ATX 1x 24pin (20+4)
EPS (CPU) 1x 8pin (4+4)
PCIE 2x 8pin (6+2)
SATA 6x
Periférico de Cuatro-Pin 3x
Floppy 1x
SKU: EVP-100-W1-0500-KR
which one would you get?
 
yeah unfortunately I there aren't many tech stores around, and getting something from amazon might take up to like a month so I'm trying locally for now.
outside of those what would you recommend?
 
Tried to keep price as low as possible with good quality, a comparison of three units,
The RMx is the best and 650watts is more than you need, but the 750watt Bitfenix might give you a bit higher wattage for a higher end gpu. The Bitfenix on the other hand is not modular while the RMx is fully modular.
All things considered is a 650watt enough and is the RMx the best of these three. Think the Bitfenix 650watt is something i could have left out, but if you feel that 650watt enough and want cheapest is it an option.
See now that the RMx costs more than the link shows, it is worth the money.