PC shut down

Hello

 

I am experiencing a strange issue while playing Sins of the solar empire rebelion.

The PC shuts down with CPU debug light on motherboard lit up, I can't tell what is causing this, the event happens roughtly after 15 mins after I start the game just when larger battles start happening, I have a fleet composed of 1 capital ship and 10 light frigates attacking a nearby Ice planet to colonize, first I send my capital ship and then the frigates, as soon as they arrive the PC shuts down, I decrease my settings anti-alliasing and texture filtering and now a minute of actual battle and again shutdown.

 

I had a laptop before one about 8 years old, it's much weaker in terms of hardware, than my PC now, and I could play the game for hours, so I don't really think my PC is too weak for the game.

 

My PC is overclocked, so I thought, that might be the cause, so I ran Cinebench R20 and Unigine Heaven both for 1 hour to verify that everything is stable, those two hammer CPU and GPU harder than any game, so if 1 hour of that proves stable, I can safely exclude stability issues.

 

I also played Tropico 6, COD Modern Warfare and a bit older AC Black Flag, all on highest possible settings, and everything worked well, well except AC but that game has crashing issues since many other people reported them.

These games are much more demanding than Rebelion, so again that confirms my stability.

 

During tests with above torture test programs I monitored temps and voltages accross the board:

 

CPU:  85-90 C

GPU: N/A

VRM: 50 C

PCH: 40 C

Ambient: 20-25 C

Vcore: 1.4000 (no flactuations)

VSOC: 1.20-1.21 (BIOS setting: 1.2)

VRAM: (1.45-1.46) (BIOS: 1.45)

VDDP: 0.91

12V: 12.0000 (no flactuations)

5V: 5.00-5.043

3.3V: 3.306-3.31

 

According to above messurements, there is no overheating (I mean the CPU temps are not ideal but are safe (Tmax for my CPU is 95 C), and I have the stock cooler), however, these are the tests, they are designed to push components to their limits, I have RGB lihtning on my MOBO, which are set to imitate CPU temps, they change from Green to Blue to Red as temperature is climbing, during playing of Rebelion the lights were always blue, indicating that CPU temperature was not above 65 C. (the point when lights turn red), also my PSU is not strugulling to deliver necesary power, since all voltages are quite stable.

 

This leaves my with only one answer, maybe the game isn't compatible with my particullar config, bellow I will post the specs, so my question would be, does anyone else experience the same issue?

 

CPU: AMD Ryzen 5 3400G with VEGA Graphics (used)

         4.1 GHz overclocked @ 1.4 v on CPU all core

         1.7 GHz overclocked @ 1.2 v on SOC / GPU

         2 GB VRAM allocated to GPU

RAM: Corsair Vengeance RGB PRO 3000 MHz 2x8 GB

          Running at max supported 2933 MHz @ 13-15-16-28 1.45 v

MOBO: MSI B450 Tomahawk (7C02v1D)

PSU: Corsair CX550  550W 

Monitor: 1920x1080 @ 60 Hz

all connected to UPS to protect against Vdrops and surges.

 

If any more information is needed I will provide them, just want to play the game.

 

Thank you

Best Regards

 

          

20,665 views 6 replies
Reply #1 Top

As a general rule if your hardware abruptly shuts down then you have a hardware side issue.

A game itself cannot cause it directly as it does not have the ability to cut the power like that so it isn't the game is incompatible with your setup, it is your setup is unstable in specific circumstances.  As there is no BSOD we can probably rule out driver bugs and bad memory.

Sins is not a very threaded game so will be pushing your hardware in a slightly different way from other more modern engines which spread the load.

It could be the cpu temp spikes briefly causing a safety shutdown, or the motherboard detects it is drawing too much power and so shuts down for safety.  The light being on would possibly point to the motherboard making the decision to cut the power.

Personally I would back off the overclock slightly on the cpu & gpu (which are running rather warm in your other test) and see if that solves the problem.

You might also be able to use a realtime monitoring app with the game running (an game overlay one) and see what the general trend is for temperatures, power usage etc before the shutdown.

Reply #2 Top

Thank you for your reply.

 

Honestlly, I don't think the motherboard shut it down, I can tell since I have exactly the same motherboard with Ryzen 3600 overclocked all-core to same speed, mining monero 24/7, never did it shutdown, it is now wattercooled and aproaches temps close to 80 C, before however, that CPU was also on stock cooler, and on "factory" settings (with voltages on auto approaching dangerously high 1.47 V, (siriouslly, do motherboard manufacturers want to destroy CPUs, or their motherboard's VRMs??), Anyway that CPU was approaching 115 C way above max temps., and motherboard didn't pull the plug until the temps were nearing 130-135 C (the temps when I tried overclocking), that's why I got wattercolling, to overclok it without such high temps., Also, that CPU draws much more power, than it's little brother 3400G, so with Monero mining 24/7, I can safely exclude motherboard pulling the plug.

 

However, I did try to go back all the way to factory settings, resetting the BIOS, I didn't even apply XMP, just left it on Auto, the PC shut down again albeit, about after 90 minutes, I was playing on 5v5 map, 60 or so planets, can't remember the name, and the shutdown happen right after my fleet now containing about 20 more illuminators, and a Radiance Battleship jumped to a nearby solar system, it did, go throug all smaller battles, but after 90 minutes, just when a very large scale battle was in place in enemy star system, there were about 50 ships fighting, some of my allies, some enemy and my fleet, another shutdown occured.

 

well i thought maybe the map has just too many things happening at the same time, before I was playing on a large 5v5 map with over 150 planets, the crash happened after 15 minutes, now I chosen a smaller map with same ammount of players (AI), and I could play for 90 minutes (it's estimated, I didn't have a stopwatch, to see exactlly how long it takes), but really if I have to minimize the map, and ammount of AI players, to make the game run on my system, I will rather not play, I kind of like large scale battles, so this solution won't do.

 

but still I'm puzzled, how is that heavy duty stuff like Cinebench R20 or Heaven won't make my PC shut down, and prove it unstable, but such non-demanding game (for today's standards) will? 

Reply #3 Top

Quoting wellahudson97, reply 3

for best solution click here.......

 

I don't understand, how will configuring my network router solve the problem?, I was on AI game not multiplayer, so I don't even use network.

Besides, even so, configuring the router won't solve issues on CPU or motherboard in my PC.

 

Can you explain what is the solution actually?

Reply #4 Top

KIKI443 .....You can ignore that comment....he was a spammer and has been 'removed'....;)

Reply #5 Top

Quoting Mohitvermaji51, reply 5

An overheating power supply, due to a malfunctioning fan, can cause a computer to shut off unexpectedly. Continuing to use the faulty power supply can result in damage to the computer and should be replaced immediately. ... Software utilities, such as SpeedFan, can also be used to help monitor fans in your computer.

 

I can confirm the PSU fan is working and there is no sign of overheating, the PSU is Corsair CX or VS not sure exactlly, and was at the time of error less than six months old, the PSU is 500 W (so there is no overload either). Let me say again that I play more demanding games than Sins of Solar Empire Rebelion, and for much longer seasons, and I didn't experience simillar problems, except AC Black Flag, but like I said in original post, that game has it's own issues, as reported on many forums.

 

In addition I stressed the CPU with Prime 95 for twelve hours and everything was stable, also Cinebench R20 several times, plus I stressed the GPU for 30 minutes with MSI Furmark (that one is verry extreme test, and as far as I know running this for longer will damage the GPU, at least that's what I heared from other people, that work with OCing), and also Heaven Benchmark with repeating loop for 2 hours, no artifacts, no shutdowns, everything was fine, I already said what the temperatures were so TL;DR return to original post for more details.

 

Haven't played for a whille because of all this, but I intend to swap the 3600 CPU from another PC, and RX 560 GPU to see if that makes any difference.

 

Thank you for suggesting the PSU fault though, however, I don't think that is the case here.