Guild Wars 2

So I have this M6500 Precision with a FireGL8720 (HD5870) that I have been using to play GW2 on the go. And its been fine since i rebuilt it (3 weeks ago~).

Well, tonight was the first time playing since the patch and at the Wurm World boss, as soon as the wurm popped out of the ground and the display zoomed out my Fans went to 100% and smoke came out of the GPU (and it smelled like electrical burning).

When the screen zoomed back in, the laptop settled down and things returned to normal.

But really, What in the hell was that about?!

what does your laptop being bad have to do with the game? If you want to play good looking video games buy a PC

Nurgle.6597:

what does your laptop being bad have to do with the game? If you want to play good looking video games buy a PC

Bro, re-read the first sentence then read his signature.

sounds like the fan voltage controller or perhaps the fan starter capacitor is degrading.

Your laptop is defective (likely blew a capacitor). GW2 stressed the system, but NOTHING any piece of code running on a computer can do will CAUSE a hardware failure on a properly designed and functioning computer system….PERIOD.

Not sure what Squishy is trying to pull here..He knows enough about PCs to know this is pretty much BS.

Ive never seen a newer laptop/GPU/CPU that doesnt have thermal limits waaaay below what it takes to make smoke.

Maybe he’s just trying to get Ashley’s attention.

MrFayth.3546:

Ive never seen a newer laptop/GPU/CPU that doesnt have thermal limits waaaay below what it takes to make smoke.

Yeah, it really should throttle before burning if the CPU or GPU was getting crazy hot.

I’d guess a bad capacitor, voltage regulator or something hit its limits and leaked some magic smoke. If the board was at least slightly overengineered (as it should be) it could still work with a small component or two burned. Not that I’d trust it to, but anyways.

Brother Grimm.5176:

Your laptop is defective (likely blew a capacitor). GW2 stressed the system, but NOTHING any piece of code running on a computer can do will CAUSE a hardware failure on a properly designed and functioning computer system….PERIOD.

Tell that to Nvidia and one of their old drivers being a cause to some of the Geforce cards to catch fire (albeit this was an issue with thermal control if I remember correctly) while running certain games…

It was weird to say the least.

Before the Wurm event I was messing around on an alt in Queens Dale, then jumped over for shatterer then Modnir, with no issues. And I have been doing Teq and other ‘annoying’ performance based events on this laptop too over the last couple weeks.

After the burn out, I rebooted the system after letting it cool down and it was working normally still, so continued to play GW2 (Hell why not, if the GPU is dying/Fried might as well let it go out in a blaze of Glory). So did Tequalt and Karka afterwards. And the issue did not resurface. Also to note, Performance was the same as before (60FPS until Battle, then 32FPS-48FPS)

But what I do NOT understand, why did only the wurm event cause this? And why was it WHEN the wurm poped out of the ground with all that smoke and crap? Yet, the rest of the time no issues? I have been to way worse events with similar effects (when Ele pops and When MAW boss is finally attackable, extreme lag until the 30 second slow down) but this didnt happen then.

sure faulty hardware does come to mind (the 8720 was bought from Ebay, so who knows what condition it was really in), but I burned in the GPU with all sorts of test (even Scrypt mining) to verify a stable operating temperature. And even when that white cloud of smoke rolled out, the GPU’s temp wasn’t above 63c.

Oh and btw I am not blaming GW2, not directly, but rather what in the wurm event would cause this much stress on a GPU. Texture Fill rate pushing the Core to the limit perhaps?

I plan on tearing the laptop apart this weekend and looking for physical signs of damage. As I still have 2 weeks left to do a RMA through my Ebay Seller :-)

Details;
AMD drivers 14.2 version 2 (Built for FireGL, but the Card detects as a HD5870M)
M6500 + FireGL7820
Ingame settings all high except, Shadows -Low, Post Processing – Low, Character Limit/Quality – Low/Medium, FXAA – Disabled, Reflections – Disabled, High Res options – Disabled, Affect LOD – Enabled, vSync – Enabled (Locked at 60hz)

and no, I am not trolling. Just a little perplexed by this is all. It’s been a while since I had a flaming Ball of smoke sitting on my lap :-)

So fire, smoke, and electrical burning smell… and the laptop still “works”? O_O???

Maybe it was just some dust in the fan that burned and got the components hot enough to smell like burning. Can’t imagine the GPU should have any business working if itself was the source lol?!?!?

Sir Kaboomski.1508:

So fire, smoke, and electrical burning smell… and the laptop still “works”? O_O???

Maybe it was just some dust in the fan that burned and got the components hot enough to smell like burning. Can’t imagine the GPU should have any business working if itself was the source lol?!?!?

And welcome to my confusion :-)

It could have been some thermal paste, but the fans and such were completely clean and free of dust. I am not ruling out a Cap, or like wise. But the fact the darn thing still works, and like before, is what is making me question this.

edit Also, why did this happen on the Wurm event, but not at any other events. I kinda wonder if there is an option is that broken in the game, such as the high res textures, that might have helped with the issue.

Does the fan spin properly at 100% when benching? with say prime 95 and uniheaven? Or is it gw2 only? If its only gw2 it sounds like a problem with the main group voltage controller on the integrated psu circuit.

If its not the main voltage controller and the temps are that low and still you got smoke and high fan speeds, It definitely sounds like the fan voltage controller. Either the circuit itself failed is burning the board or it is pushing too much power into the fan from a failing cap (lower resistance than spec) and slightly melting the motor housing.

dodgycookies.4562:

Does the fan spin properly at 100% when benching? with say prime 95 and uniheaven? Or is it gw2 only? If its only gw2 it sounds like a problem with the main group voltage controller on the integrated psu circuit.

If its not the main voltage controller and the temps are that low and still you got smoke and high fan speeds, It definitely sounds like the fan voltage controller. Either the circuit itself failed is burning the board or it is pushing too much power into the fan from a failing cap (lower resistance than spec) and slightly melting the motor housing.

And this was more or less the kinda response I was looking for :-)

The Fan’s spin at 100% normally and 50% idle correctly even now. The temps fluctuate between 50c and 63c depending on GPU load. Right now things are still working like they were before the issue last night.

I am concerned about the FanIO controller though. I am going to tear my laptop apart tonight and see if I can find physical signs of damage and trace that circuit out. if its bad then Ill RMA the board back to my seller. If its not, then I gotta dig deeper into the laptop. At least the GPU’s in the M6500 are modular, so if it is just a bad/dying GPU thats an easy rip and replace fix.

I main reason why I asked about the Wurm event, where the issue occurred, is I needed to find out what in that event could cause this kinda load. I have tested Fill rate, Script/Sha-256 Mining, and simple 3dmark testing from the day i got the card for a 3-4 day stretch with no issues (Temps never went above 65C’s during the mining testing).

But then all the sudden a wurm busts out of the ground and my laptop is smoking. Makes no sense LOL.

Can’t speak for the wurm event but Bloodtide coast does give me the char matrix crash recently, never get that anyother map. My guess it could be some bug they made when they created the wurms though I’ve never crashed during the actual events. Another thing is that annoying map they use that zooms in and out has crashed my game many times especially when there ISN’T much activity around me so it’s not overheating or the system being pushed too far.

well, I didnt crash just had a huge thermal issue.

So, took my laptop apart and found some burn marks around a diode on the GPU. Looks like its tied to the power management as the GPU wont get out of power savings mode with out using MSI afterburner now.

So, RMA’d the GPU with my seller and the Seller said to keep it so im going to run it till the replacement part gets here. Verified voltage and the Diode IS working correctly for now. But I have to use MSI afterburner to get it up to core 700 memory 1000.

getting 80-100 fps still, and the temps are 63c in a heavily populated zone and 58c in a normal zone (like an open area in queensdale)

I really do think GW2 has some hidden options enabled since the patch that pushed my GPU harder then it ever has and the GPU’s hardware didnt like it and well…poped. I have been playing for 3 weeks on this laptop and never had any issues or warnings (like a hot laptop…you know what i mean) that would suggest eminent hardware failure.

I dont run high res textures, but it seems there are issues going on with some players that have them enabled, maybe its semi-related?

anyway, we can argue the point that software can’t damage hardware. But thats only true if all the safety’s are in place (driver level) and thermal protection is correctly working (vBIOS/systemBIOS). Faulty software will trump hardware any day. And its possible that something in the patch triggered something dormant that has been vulnerable in my system since the beginning. If the same thing happens again, after the GPU replacement, I will then take this up with ANET directly.

I run on a laptop as well, prior to patch I had constant 50 – 60 fps (have it limited to 60 fps in options), gpu temperature never got above 65 degrees C in WvW 3 way battles in SMC. Since patch I am getting FPS drops to 10 or so, massive lag, skill delay, and GPU temps up to 79 degrees C. They did something in the update, just not sure what yet. Also, if you have High Resolution Textures checked, uncheck it. This may help a bit.

Edit: just read your last post about textures, so ignore that part.

I’d take a crash over smoke and fire any day ^^. Sadly don’t think Anet will feel too obligated to accept any responsibility for such a thing, you’ll get the apologetic platitudes only.

KillSwitch.8192:

Brother Grimm.5176:

Your laptop is defective (likely blew a capacitor). GW2 stressed the system, but NOTHING any piece of code running on a computer can do will CAUSE a hardware failure on a properly designed and functioning computer system….PERIOD.

Tell that to Nvidia and one of their old drivers being a cause to some of the Geforce cards to catch fire (albeit this was an issue with thermal control if I remember correctly) while running certain games…

How is that NOT a design issue (either a driver design/code error or the hardware configuration was wrong to begin with)? Note the “properly designed and functioning”, part of my statement.

Note that most modern electronics will have MANY “filter” capacitors in the power plane circuit (designed to keep the power voltages highly stable and noise free across the PC board) that are NOT absolutely necessary for the proper operation of the electronics themselves. These can “blow” without effecting the overall operation of the circuits themselves. As long as they don’t SHORT the power plane or cause any surge damage at the time they fail, you will get a puff of smoke and burning smell while nothing else will appear to be wrong and the circuit will continue to work without it just fine (abet, the supply voltage to nearby components may be slightly effected).

Note that these usually blow because they are contain some impurity defects from the factory (potential low quality component) that finally “fail” after some period of time under stress.

It’s hard to speculate what actually happened, but my guess would be that there was a loose foreign object (maybe a little solder ball) that managed to short out some power connection near the diode.

The diode probably got hot enough to smoke, but may still be good. Diodes can take some pretty hot temperatures without failing. They actually make good temperature sensors for soldering irons because their voltage drop varies linearly with temperature.

This is probably just a one time occurrence, but I would make sure to look around for loose objects inside the laptop case (maybe shake out it a bit).

Edited to add:

Brother Grimm.5176 is correct. I’d like to add to his post, in that static damage is often the cause for these failures. Static can actually damage a component without it failing right away. In fact, it can take months in some cases.

I cringe when I see people handle computer component circuit boards and touch the boards and connectors – especially when they aren’t even wearing a ground strap. This is something you should never do.