41 replies to this topic

[ManDaisy]

Becareful , was in mechlab. Was idling in there when artifacts started popping up and my GPU died.

Radeon 7950.

Off to the ovens.

[ball0fire]

i srsly doubt mechlab had anything to do with it

other than the fact it uses yah gpu....gpu's just die sometimes

[NocturnalBeast]

ManDaisy, on 08 April 2014 - 06:38 PM, said:

Becareful , was in mechlab. Was idling in there when artifacts started popping up and my GPU died.

Radeon 7950.

Off to the ovens.

Most likely your video card and or system was inadequately cooled, or overclocked with the voltage cranked up.

[Catamount]

Games don't just melt GPUs. I only ever saw a case of that once, a case with Star Trek Online and Geforce 200 series models, but there was a highly specific reason for it.

MWO's mechlab runs at moderate framerates with moderate GPU load. There is nothing that gives it a special proclivity to kill GPUs

[Summon3r]

ManDaisy, on 08 April 2014 - 06:38 PM, said:

Becareful , was in mechlab. Was idling in there when artifacts started popping up and my GPU died.

Radeon 7950.

Off to the ovens.

lol?

[Odins Fist]

ManDaisy, on 08 April 2014 - 06:38 PM, said:

Becareful , was in mechlab. Was idling in there when artifacts started popping up and my GPU died.

Radeon 7950.

Off to the ovens.

Sorry to hear your GPU took a nosedive into oblivion, that stinks.

The Mechlab didn't nuke your card though, sounds like it suffered a catastrophic failure.
Any number of reasons could have done it, overheat, overvoltage, and the worst of all, liquid cooling leak.

I had a friend that swore up and down that his coolant leak form his CPU block didn't kill his video card and Mobo, in his own words --- "There's no way, I was using non conductive liquid in my loop..!!!"

To which I said "It's only non conductive if your case is absolutelty spotless with no trace of dust, especially on the component it leaks onto".. He said, "Nope", and then I explained that once non conductive liquid has any sort of contaminant like dust, then all bets are off, and the liquid now has a good chance to be conductive.

His card a small layer of dust on it, Game Over..

[Grendel408]

Before you toss your GPU out... check this first dude... I actually witnessed these, thought **** my GPU is done for cause I overclock it (but don't leave oc'd while not gaming) and it's probably about time it died... nope. It's MWO... with a fix coming next patch according to Dev post.

But... from personal experience, regularly clean out your PC... especially if you're a pet owner. Dust and debris build-up in the GPU fan assembly will sure as hell fry your component due to improper cooling and bad ownership... see spoiler:

Spoiler

All you need is a duster can w/no CFMs (so you can do this outdoors with no complaints lol), a stool (my drum stool as show in spoiler), and about 20mins of time. If your duster can starts to lose pressure, warm it back up in a sink so it's back to room temp and continue until dust free. It helps to remove components too at time

Edited by Grendel408, 08 April 2014 - 09:22 PM.

[Thorqemada]

My GPU had a cooler not working right - that way the Mechlab may hurt yours but since i got the replacement GPU with a working cooler it runs very cool.
Probably you had a hardware defect that came accidentaly to daylight when you played MWO by random bad luck.

[nightsniper]

It should be no surprise that the Crytek 3 engine utilizes much more a system then most others today. It flat out pushes CPU's and GPU's. For a longtime the manufacturers of these components have made bold claims about the equipment knowing full well that few if any would ever reach the limits. Early closed beta I had an issue with heat on my Mobo running a early AMD quad core. System was not up to stress. I have seen some interesting things while testing and trying to get various systems to run MWO and this engine. Most recently working with AMD, Crucial and Microsoft on the game exit crashing that started last March but has been mostly corrected with the 4/1 patch. AMD tech support had me download a program to monitor CPU and GPU performance and Microsoft had me down load a Windows 7 patch for AMD multi-cores.

Playing the other day I crashed entering a matches twice. The file showed after the connecting screen went black that the GPU and video Memory was running at 100% seems strange on a black screen. This again was a hard freeze and I had to use the power button to exit and restart. My point is that this game does at points stress a system and if your not monitoring and running a well tuned PC you are likely to see failures due to the stress. I do believe we will see more of this from other games as well many engines that compete with Cryengine certainly try to utilize more of the computing power.

If you are concerned you may want to go to someplace like CNET and look for utilities that can monitor your heat and performance. I installed a thermaltake system into my case that has sensors on the video, memory, and CPU even though they are water cooled. Just to be on the safe side. I can say that so far heat has not been an issue on this PC.

I run and AMD FX 6300 6 core
Gigabyte 990FX UD3 version 4 Mobo
16 gig of ram
1 Radeon HD 7770 2 gig DDR5 ( took out the second card till crossfire is supported)

I can run on High and very high with minimal FPS loss but like to run on high with custom settings turning some down to medium. I have two monitors a 32 inch wide screen Dell monitor and a smaller 20 inch viewsonic I run the game on the 32 and the monitoring on the 20. Since last patch 2nd monitor is a little tough to keep from going black though you have to keep alt tabing.

Older systems and weakened components will meet an early grave as engines utilize more of the resources on a consistent basis.

[VXJaeger]

I think that we have a winner here, GJ me[c]hlab.
http://mwomercs.com/...ost__p__3282951

Edited by Egomane, 10 April 2014 - 12:34 AM.
CoC violation

[Kuga ZA]

There is an issue with the Mechlab running cards at 100% GPU load. I've noticed this with my 2 x 770's. I have a G19 and Aida64 with custom monitoring etc, will post a video tonight and show this.

[Graves24]

Catamount, on 08 April 2014 - 07:25 PM, said:

Games don't just melt GPUs. I only ever saw a case of that once, a case with Star Trek Online and Geforce 200 series models, but there was a highly specific reason for it.

MWO's mechlab runs at moderate framerates with moderate GPU load. There is nothing that gives it a special proclivity to kill GPUs

Starcraft 2 used to melt Nvidia cards when idling in the campaign screen. They fixed it when hundreds of people complained about it.

[VXJaeger]

I did some testing and here some rough results:
Card: Gigabyte HD6990 2Gb

Test results on GPU temps, DX9:

GPU basic temperature: 64°C
Mechlab basic screen right after startup: 74°C
Mechlab on conf screen after rotating mech: 92°C
On match and hellofa fighting going on: 82°C

So you might say there's some optimizing to on on mechlab's code.
And I have Cooler Master HAF X-case with 3 fans in and 2 out, so heat build-up is not the problem here.

[Catamount]

Graves24, on 10 April 2014 - 08:30 AM, said:

Starcraft 2 used to melt Nvidia cards when idling in the campaign screen. They fixed it when hundreds of people complained about it.

Yeah Nvidia GPUs in particular had a problem with this in STO, because Nvidia treated TDP like a very loose suggestion, back when Nvidia had nothing even remotely comparable to AMD's Powertune. In STO's case it was the loading screens, so it was probably a similar issue.

[VXJaeger]

I found some sorta solution to mechlab problems on my rig, probably works on all highend AMD machines.
My machine AMD FX-8350, Gigabyte HD6990, Asus Sabretooth, 32Gb

1. Clean all old AMD drivers from your system with AMD Clean Uninstall Utility.
2. Install latest Beta-drivers.
3. Very important: DON'T TOUCH ANY SETTINGS ON CONTROL CENTER!!!
4. Change DX9 -> DX11 and VSync to "off".
5. Profit.

Now mechlab gives 40fps without flashing&heating and in match with "Very High"-settings steady 35fps, which is tolerable. Temperature holds on steady 72-75°C.

There seem to be somesorta mismatch with Radeon's drivers and MWO videosettings, and they struggle which one rules. If Control Center is left untouched after installing, MWO's own settings get upperhand and Works As Intented™.

Edited by VXJaeger, 12 April 2014 - 08:57 AM.

[ManDaisy]

Happy end everyone. The bread baking oven method really does work! Back to normal now

[Iceolator]

Well, it's radeon. Honestly, what were you expecting when you decided to save a few bucks buying AMD instead of Nvidia?

[ThatBum42]

My understanding is that AMD cards are technologically superior to Nvidia cards, but Nvidia has better software support/drivers for their hardware. The OP's card is just a bad example. This is also why people say AMD cards are more suited for GPGPU or bitcoin applications.

[Catamount]

Myaff, on 17 April 2014 - 06:24 PM, said:

Well, it's radeon. Honestly, what were you expecting when you decided to save a few bucks buying AMD instead of Nvidia?

Yes, because obvious fanboy trollbait definitely makes for the most productive of discussion

[Sug]

Myaff, on 17 April 2014 - 06:24 PM, said:

Well, it's radeon. Honestly, what were you expecting when you decided to save a few bucks buying AMD instead of Nvidia?

Wasn't Nvidia the one with that bad batch of cards in like 2006?

Me, friend, roommate, and sister each went through 3-4 Nvidia cards in the course of a year. All with bad artifacting/glitching.
Swore off Nvidia and have been buying AMD ever since : /