Question So the company who pre-built my PC left my PC under load and unsupervised for a week straight..

[Noobpunk]

Hey all,

So my PC had to be RMA'd due to a fault of the CPU. So I had sent the PC back to them to RMA right before Xmas last year.

One day I decided to purchase a game from Steam and I was asked if I would like to install the game to my PC remotely..and I thought ok, the PC is on and is being tested by the technician. However, this was during the xmas holidays and the company stated they would be off for a 4-5 days in the week of xmas.

Every single day they were closed, the PC was also on.

My friends also asked me how I got my hands on a game called Final Fantasy XIV: A Realm Reborn, which is an MMO that was temporarily closed for sale. My Xbox account was also online on the PC which showed to everyone on my friends list that I was playing this. I monitored for how long this was ongoing and it was throughout the 4-5 days that the company were off from work.

So now I have found out that my PC was on, in game for at least 4-5 days straight. I know this because every day I opened Steam on my phone to see if I still had the option to remotely install games onto my PC.

I have a 3090 too..and the memory junction temperatures on these cards are something else! They can go excessively high!

So I managed to actually buy the game for myself as Square Enix lifted the sales restrictions and I loaded up the game I noticed the memory junction temperatures going to the higher end of the card. Now I can only imagine what was going on during that unsupervised time off with the PC when it was on with this game the entire 4-5 days.

Just playing the game for less than an hour the card was pushing the GPU's memory junction temperature into the mid 80s(THIS IS WITH MY UNDERVOLT SETTINGS)..Now I am thinking how this would increase throughout the 4-5 days it was left on the whole time.


So I have two questions..

1 - Can there be damage done to a PC that is left on under load in an MMO game for a few days straight? Would I see the damaged now or is it a waiting game?
2 - Are there any tests out there that can help infer the health of my graphics card?

 

[Spacekeleton]

Damage is only caused when temps go higher than they should and each component has it's own level of that temperature, and GDDR6x memory the 3090 is using is hot as it is. If you have no idea what was going on with perfomance and temps there, you could just send GPU or any other part you want for diagnostics to manufacturer and demand compensation if it will be found out that any damage was caused during that time. But since it's MMO, I am sure a lot of people are grinding it for days straight with no problems, it's not some 4k rendering or something like that.
Still, you should ask who is responsible for leaving your PC in that state for that time.

 

[Cybernetictator]

As an avid FFXIV player it's fairly common to see players keep their PCs on for long periods of time. The game isn't really that demanding compared to other titles and it's fine as if PC overheats it will shut itself down to prevent any real damage. If you really feel like there's an issue I'd get the components checked but I believe you should be fine.

What the real issue is is who kept your PC on for that long/who was responsible because that person needs to get reprimanded..

 

[Karadjgne]

My old 3770k had a 4.9GHz OC and ran 24/7/365 for 7 straight years. The Only times it was ever shut down was for a couple of hors for cleanings, or when the power went out. Otherwise it was on. My current pc is the same way, never gets turned off.

Running a pc doesn't hurt it one bit. They do not require shutting down after playing for a few hours. There's ppl here who run servers and Nas that never get shutdown. You think the registers at 24 hr gas stations get shut down. Computers at TV stations? There's most likely more and larger computers in service that never get turned off than there are pc's that do.

What the real issue is is who kept your PC on for that long/who was responsible because that person needs to get reprimanded..

For what? The pc ran for a week straight with no issues, no cpu problems, no glitches, bluescreens or black screens, no loss of data or corruptions. I'd call that a successful repair! Most places, if it boots it's good! Gimme my money! This pc went a solid week. Bingo.

Computers are made to be run. They have limits, like a gpu or cpu will throttle if it reaches that temp, shutdown if it cannot throttle enough to stop the temp from going over that limit.

That pc passed with flying colors.

 

[Deleted member 14196]

Send it back And buy a PC from a reputable builder

 

[Unolocogringo]

A week is nothing.
My computers fold 24/7/365. They get shut down 3-4 times a year for windows updates and cleaning. Then back to work.

 

[Colif]

Sounds like they gave it extra long testing. I expect if it had failed during that time they would have fixed it for you anyway.

Sounds like PC hardware should be fine.

 

[Cybernetictator]

For what? The pc ran for a week straight with no issues, no cpu problems, no glitches, bluescreens or black screens, no loss of data or corruptions. I'd call that a successful repair! Most places, if it boots it's good! Gimme my money! This pc went a solid week. Bingo.

Not for that reason, but for security. If the PC was left on on a game someone could've got in and messed around with the OP's files. I wouldn't want my PC running in a shop, closed or not, just in case.

 

[USAFRet]

The problematic thing about this story is how the shop managed to obtain a game that apparently was not being sold at that time.
And then, why did you allow them to install it remotely?

The leaving on for a week is trivial.

 

[Noobpunk]

The problematic thing about this story is how the shop managed to obtain a game that apparently was not being sold at that time.
And then, why did you allow them to install it remotely?

The leaving on for a week is trivial.

So the game was released a couple of years back but around xmas they had to temporarily stop selling the game due to congestion in their servers. People were still able to play the game if you had purchased the game prior to the halt.

One thing I noticed once having the PC back was there was a Final Fantasy XIV Benchmark folder in my documents tab. So I am assuming it was to benchmark and test etc.

 

[USAFRet]

So the game was released a couple of years back but around xmas they had to temporarily stop selling the game due to congestion in their servers. People were still able to play the game if you had purchased the game prior to the halt.

One thing I noticed once having the PC back was there was a Final Fantasy XIV Benchmark folder in my documents tab. So I am assuming it was to benchmark and test etc.

If this were my system, I'd be doing a full wipe and reinstall once I got it home.
Not a 'reset'....all partitions, all data.....deleted.
A bare OS installed, from media I created.

 

[Noobpunk]

If this were my system, I'd be doing a full wipe and reinstall once I got it home.
Not a 'reset'....all partitions, all data.....deleted.
A bare OS installed, from media I created.

I have nothing on there but games so I am more worried on the hardware in case its been damaged because of this week stretch!

I have read online that Final Fantasy XIV has a 30 minute AFK timer too..

 

[USAFRet]

I have nothing on there but games so I am more worried on the hardware in case its been damaged because of this week stretch!

I have read online that Final Fantasy XIV has a 30 minute AFK timer too..

Many people leave their systems on 24/7. Doing all sorts of things.
Database server, Folding@Home, surveillance camera recording......
Unless I am on vacation, my house systems are on 24/7.

Its not what you have on there, but rather what else did they put on there.
The existence of just one application that I did not specifically request makes my spidey senses go nuts.

 

[Noobpunk]

Many people leave their systems on 24/7. Doing all sorts of things.
Database server, Folding@Home, surveillance camera recording......
Unless I am on vacation, my house systems are on 24/7.

Its not what you have on there, but rather what else did they put on there.
The existence of just one application that I did not specifically request makes my spidey senses go nuts.

Believe me I understand!

As I emphasised in my first post, the memory junction temperature on this game went into the high 80s with my undervolt..I can only imagine how high it must have gone when it was with the technicians pre undervolt in game!

I did a Furmark stress test on the 3090 for 5 minutes, the memory junction temp was at 94c by the 5th minute and I stopped the test. I dont know if 5 minutes is long enough to see if the GPU is healthy or not however I did not see anything out of the blue other than not having the smoothest frame rate during the test.

 

[Karadjgne]

As a professional repairing pc's, and someone brought in a pc with an issue, no extra software other than diagnostic (cpu-z, Prime95, etc) would be introduced and later deleted. If there was a game with a benchmark ability, I'd run that as purposed for stability, playability, diagnosis etc. So it's not surprising to see a benchmark result on your desktop. You had a cpu complaint, they ran the game to test the pc. Does it play, does it glitch, did the cpu cause data errors that give false negatives, is it fixed..

Not for that reason, but for security. If the PC was left on on a game someone could've got in and messed around with the OP's files. I wouldn't want my PC running in a shop, closed or not, just in case.

If someone actually broke into the store, managed to bypass the alarms, the watch dog, the police drive bys etc, the absolute Last thing they will have done is stick around to play games and messed with the files on Op's pc. They'd have grabbed it complete and gone out the window with it. Buh bye, thanks for the expensive gpu!

 

[Noobpunk]

If somebody could please give a response to my message below!

As I emphasised in my first post, the memory junction temperature on this game went into the high 80s with my undervolt..I can only imagine how high it must have gone when it was with the technicians pre undervolt in game!

I did a Furmark stress test on the 3090 for 5 minutes, the memory junction temp was at 94c by the 5th minute and I stopped the test. I dont know if 5 minutes is long enough to see if the GPU is healthy or not however I did not see anything out of the blue other than not having the smoothest frame rate during the test.

 

[USAFRet]

It is impossible to state what, if any, "damage" resulted from whatever they did.

Generally, the component will throttle itself so as not to actually incur 'damage'.

 

[Karadjgne]

Furmark is well documented as being a gpu killer. If you actually read the articles, it states it should be used sparingly, if at all. It's not a stress test, it's an extreme torture test that'll drive a card like a 3090 to pull 500w when normal power limits will only pull @ 425w. With modded 1000w vbios, it's driven a 3090 to pull 630w.

That's not a wattage your standard VRM's and cooling can handle for long.

Running furmark seriously raises the chances of you doing far more damage to the gpu than leaving the game running, a 80°C memory junction or both combined. 80°C is chump change, the card will throttle at 83°C and can handle temps in the 90's.

 

[InvalidError]

As I emphasised in my first post, the memory junction temperature on this game went into the high 80s with my undervolt..I can only imagine how high it must have gone when it was with the technicians pre undervolt in game!

80C for memory is a non-concern since the PCB gets warmed up most of the way to that temperature by the GPU die's 200-300W and VRM losses. Most semiconductors, especially lower-power ones that have nearly constant power draw like VRAM, are still perfectly fine at 90C.

 

[Noobpunk]

Furmark is well documented as being a gpu killer. If you actually read the articles, it states it should be used sparingly, if at all. It's not a stress test, it's an extreme torture test that'll drive a card like a 3090 to pull 500w when normal power limits will only pull @ 425w. With modded 1000w vbios, it's driven a 3090 to pull 630w.

That's not a wattage your standard VRM's and cooling can handle for long.

Running furmark seriously raises the chances of you doing far more damage to the gpu than leaving the game running, a 80°C memory junction or both combined. 80°C is chump change, the card will throttle at 83°C and can handle temps in the 90's.

Ok,

I have only done the test for 5 minutes on the dot and all of my temperatures were in the 60s for the 3090, GPU Core & VRM etc..however the memory junction temperature was the only one that was above that in which it was in the 90s..

Just to confirm, my GPU was pulling a maximum of 390w.

I have only done this test once! So I hope I havent done anything to it..

 

[Karadjgne]

It survived. It should be fine. Cpus/gpus last well over 20+ years when not over-volted, so even if you lost half of the lifespan, you'll be dumping the card in less than 10 years anyway. It's really a non-issue

 

[Noobpunk]

It survived. It should be fine. Cpus/gpus last well over 20+ years when not over-volted, so even if you lost half of the lifespan, you'll be dumping the card in less than 10 years anyway. It's really a non-issue

I mean if it pulled 390w when the max wattage is 450 and the temperatures were within the specs, how have I lost a portion of the lifespan if so?

 

[Karadjgne]

Add up all possible damage, worst case scenario. Chances are good there's no damage at all, but that's impossible to predict. If you assume you could have lost half, you know where you stand if tempted to further abuse such as OC etc.

 

[DSzymborski]

I'm a little confused by some things, too. For example, how did the company know your User ID and login for Final Fantasy XIV? The client specifically doesn't save password information.

 

[Karadjgne]

Who says it was op's? Game is game, can use a different login and pass.