If you need help or have questions just call Alienware Phone support and point them to this message. They can help you "clean install" these drivers and also help with basic troubleshooting.
You can also ask questions here. We were all newbies once. However, you said you "weren't that techy" so phone support might be easier on you.
Should also mention I have installed NOTHING on this computer except WoW, LoL, GW2, D3, Minecraft, and spotify. And that I tried to use alienrespawn to return computer to a state prior to first crash. No luck.
Hey tesla I am having the exact problem on the exact same machine described by the original poster. After a few hours of trolling forums your name keeps popping up. The computer is 7 days old and can still be sent back for a return/replacement; however, if its a known issue with hd78xx series cards I don't see how that will help. I also purchased it with the full intention of keeping...soooo you know. Anyway, I followed your instructions to a "T" for the clean install of Catalyst 13.1 drivers and I continue to experience same problem. I have been in touch with a tech support rep for the past few days, and she installed GPU-Z which she has been having me use to log the state of the video card at the time of the crashes. temperatures all seem normal...there are occasional spikes that indicate the vid temp jumps over 127 degrees C for just 1 second, but I thought that must be an anomoly or at least innacurate (it seems nearly impossible for the temperature to just from 57 degrees to 127 and back to 57 in less than 3 seconds?). During the first 2 days following purchased this computer I was running games (diablo 3, Guild wars 2, WoW, Minecraft...all on max settings) for around 8 hours with no issues. It was during the 3rd day of my ownership that this issue began and has persisted. It has gotten to a point that I cannot play Diablo 3 for even 10 minutes without it crashing; however, the system will run for forever when doing normal computing tasks (surfing web, watching netflix, email, etc). The only time it crashes is when playing games; which leads me to lean towards the vid card as the issue. I don't know. I was hoping for an update, but out of the 2348975203945720394857 people experiencing this issue over the passed 2 years not one of them has explained what they did to solve the problem. If you know of any fixes for this or can point me in another direction or offer another solution besides the catalyst 13.1 drivers I would be incredibly thankful.
If you know of any fixes for this or can point me in another direction or offer another solution besides the catalyst 13.1 drivers I would be incredibly thankful.
Are you in USA?
Are you using a good UPS (like APC).
What is you hardware skill level? Are you able to work inside computer? If I asked you to "re-seat video card" or "check video card power plugs/numbers" ... could you do that without exact instructions and not blowing up computer?
Alienware™ 875 Watt Multi-GPU Approved Power Supply is the only info I have on UPS.
I'm pretty hardware savy even when I'm unfamiliar with product. I've been inside a few motorcycles and underneath a few cars in my day. If nothing else there is probably a "how to" video that shows and explains how to reseat the video card I can find online.
Last night I trolled through 2 years worth of posts on guru3d.com regarding the radeon hd7870 and found that there are about 10 gazillion people experiencing the same issue. It was suggested on this forum that the card ships at a higher clock than it was designed to run at? So I turned the core clock to 995 mhz and the memory clock to 1000mhz and allowed it to use up to 5% more power. Some of the people on the forum also suggested that the card shuts off due to a failsafe when it is being underpowered. After tweaking clockrate last night, I was able to play GW2 for aroud an hour with no crashes and D3 for 2 hours with no crashes. It didn't crash one time last night at all after the changes were made. I'll admit that it wasn't a long enough period of time to asses whether this change has fixed the problem; however, it was running exponentially longer than it has in the past few days. Still getting 60fps on D3 and 40-60 on GW2 according to fraps.
Also, I was looking through the logs from GPU-Z and even with this current setup the card is running between 95-99% of the GPU Memory. Is it common for a vid card to use almost all of its memory when gaming or watching movies?
If the machine is crashing, so hard that Windows doesn't have a chance to record the error ... that usually means a hardware problem. See my posts here for trouble-shooting:
Assuming the power supply wire colors indicate the different 12v rails (going to the video cards): Try with a power connector from each wire set at the same time. Meaning, use either P14 OR P15 and also either P16 OR P17
Last night I trolled through 2 years worth of posts on guru3d.com regarding the radeon hd7870 and found that there are about 10 gazillion people experiencing the same issue.
Well, I have personally been pushing Dell to get to the bottom of this. It seems to have started with AMD 7xxx cards and Aurora R4 (x79 chipset and/or PCIe3.0). Could be so many things (motherboard, power, drivers, OS, etc.) ... and according to your findings ... maybe a flaw that worked itself into AMD reference design. I don't have a suitable machine here to investigate further ... but Dell and AMD do.
It was suggested on this forum that the card ships at a higher clock than it was designed to run at? So I turned the core clock to 995 mhz and the memory clock to 1000mhz and allowed it to use up to 5% more power. Some of the people on the forum also suggested that the card shuts off due to a failsafe when it is being underpowered. After tweaking clockrate last night, I was able to play GW2 for aroud an hour with no crashes and D3 for 2 hours with no crashes. It didn't crash one time last night at all after the changes were made. I'll admit that it wasn't a long enough period of time to asses whether this change has fixed the problem; however, it was running exponentially longer than it has in the past few days. Still getting 60fps on D3 and 40-60 on GW2 according to fraps.
Good work.
Not sure what you did (under-clocked it uses 5% more power ?) ... but if it becomes stable it's a start in the right direction.
This is what the project needs. Someone to start with the basics and work their way up to the more complicated things. Keeping notes and testing as they go. Methodically going though the system ... not jumping around as they see fit looking for an easy fix.
When not testing with Guild Wars 2 or Diablo 3 ... check out OCCT Power Supply test. With that running and maybe a long virus scan going at the same time ... if machine can survive that, it should be good and stable.
Alienware™ 875 Watt Multi-GPU Approved Power Supply is the only info I have on UPS.
I'm pretty hardware savy even when I'm unfamiliar with product. I've been inside a few motorcycles and underneath a few cars in my day. If nothing else there is probably a "how to" video that shows and explains how to reseat the video card I can find online.
I just never know what people's skill level is (or motivation level). Sometimes I recommend stuff and people feel "put-out" to have to troubleshoot anything when they can just return it.
No, not internal power supply. I meant UPS Battery (like APC unit) to make sure machine is getting good, clean, uninterrupted, high amps power (into power supply).
no it doesn't use 5% more power underclocked; it is now set up to be allowed to use 5% more power than it was if needed. I'm going to give it about a week with current setting and no changes and if I'm still not having issues, I'll consider this current setup a fix. Afterwards, I'm going to begin tinkering to see if I can do some of the things you've suggested to get the card to function at its orginally shipped settings. I am not using a UPS battery. I can assume from your response that an APC unit is a good way to go as far as buying one?
1. no it doesn't use 5% more power underclocked; it is now set up to be allowed to use 5% more power than it was if needed.
2. I'm going to give it about a week with current setting and no changes and if I'm still not having issues, I'll consider this current setup a fix.
3. Afterwards, I'm going to begin tinkering to see if I can do some of the things you've suggested to get the card to function at its orginally shipped settings.
1. Is that in Catalyst v13.1 ... AMD Overdrive? I don't use it (I run my 5870 at "stock clocks"). I see where you could "underclock" GPU and RAM if you wanted to, but I don't see a setting to modulate the power (amps) to the card. Maybe some pics would be helpful?
2. Great plan. Under-clocking was an idea I had a while back but it has been hard to find someone who is cooperative to try it. With the machine just "turning off" like that, that could be the video card just shutting down (thinking it need to try to save itself).
3. Good. Yes, I would like to know the "stock clocks" for that card, how far you had to turn it down to be stable, and if you can get it back to normal/stock. It almost sounds like the TJmax for the GPU is in-correctly set in card's vBIOS?
stock settings are 1000mhz to core and 1200mhz to memory. currently running 995 to core and 1000 to memory. Spoke too soon though it seems. I just crashed like 10 mins ago while playing GW2 on max settings (v-sync on). GPU-Z logs aren't indicating anything strange happening as far as I can tell. If I sent you the logs from the crashes would you mind looking over them to see if there is anything significantly changing prior to the crash? Trying not to get frustrated, but its a pretty big bummer that it was working fine for a few days and now possibly mis-behaving again. I'm going to try to play with some more settings and tweak the clocks a little more later on tonight.
APC is owned by schneider electric..which means I can get a pretty good deal on these through my work; however, I'm not sure which one I need? I was looking through their website and couldn't find anything to help me determine which once I need.
APC is owned by schneider electric..which means I can get a pretty good deal on these through my work; however, I'm not sure which one I need? I was looking through their website and couldn't find anything to help me determine which once I need.
Ya, they have their hands in so many markets over at www.apc.com, that sometimes it's hard to find pre-sales info on the consumer stuff. I recommend the high-end consumer models (for high-end desktops).
I like the 1500 (I connect my main PC, monitors, and other stuff on my side of the office to it) , but they also make the 1300 and 1000 in this nice LCD model.
1. stock settings are 1000mhz to core and 1200mhz to memory. currently running 995 to core and 1000 to memory.
2. Spoke too soon though it seems. I just crashed like 10 mins ago while playing GW2 on max settings (v-sync on). GPU-Z logs aren't indicating anything strange happening as far as I can tell.
3. If I sent you the logs from the crashes would you mind looking over them to see if there is anything significantly changing prior to the crash?
4. Trying not to get frustrated, but its a pretty big bummer that it was working fine for a few days and now possibly mis-behaving again. I'm going to try to play with some more settings and tweak the clocks a little more later on tonight.
1. Try 900 Core, 1000 memory
2. Well, still, it sounds like it was better
3. If you want
4. So, that's why I suggested using a Stress Test program like OCCT (instead of trying to use games as a stress test). If it's going to fail, you want to know right away so you can try something else. If it passes OCCT Power Supply Test for 15 minutes, then you can try games.
Tesla1856
8 Wizard
•
17.4K Posts
0
January 24th, 2013 14:00
Try the newly released AMD Catalyst v13.1 video card drivers.
http://support.amd.com/us/gpudownload/windows/Pages/radeonaiw_vista64.aspx
For you AMD 78xx/79xx users ... I saw these issues Resolved in the Release Notes:
•A sporadic system hang encountered with a single AMD Radeon HD 7000 Series GPU seen on X58 and X79 chipsets.
•An intermittent hang encountered with AMD Radeon HD 7000 Series GPUs in an AMD CrossFire + Eyefinity setup.
http://support.amd.com/us/kbarticles/Pages/AMDCatalystSoftwareSuiteVersion131.aspx
Be sure to "Clean Install" (see my 01-23-2013 Post)
http://en.community.dell.com/owners-club/alienware/f/3746/p/19340381/20283836.aspx#20283836
If you need help or have questions just call Alienware Phone support and point them to this message. They can help you "clean install" these drivers and also help with basic troubleshooting.
You can also ask questions here. We were all newbies once. However, you said you "weren't that techy" so phone support might be easier on you.
goforthandconqu
9 Posts
0
March 26th, 2013 18:00
Should also mention I have installed NOTHING on this computer except WoW, LoL, GW2, D3, Minecraft, and spotify. And that I tried to use alienrespawn to return computer to a state prior to first crash. No luck.
goforthandconqu
9 Posts
0
March 26th, 2013 18:00
Hey tesla I am having the exact problem on the exact same machine described by the original poster. After a few hours of trolling forums your name keeps popping up. The computer is 7 days old and can still be sent back for a return/replacement; however, if its a known issue with hd78xx series cards I don't see how that will help. I also purchased it with the full intention of keeping...soooo you know. Anyway, I followed your instructions to a "T" for the clean install of Catalyst 13.1 drivers and I continue to experience same problem. I have been in touch with a tech support rep for the past few days, and she installed GPU-Z which she has been having me use to log the state of the video card at the time of the crashes. temperatures all seem normal...there are occasional spikes that indicate the vid temp jumps over 127 degrees C for just 1 second, but I thought that must be an anomoly or at least innacurate (it seems nearly impossible for the temperature to just from 57 degrees to 127 and back to 57 in less than 3 seconds?). During the first 2 days following purchased this computer I was running games (diablo 3, Guild wars 2, WoW, Minecraft...all on max settings) for around 8 hours with no issues. It was during the 3rd day of my ownership that this issue began and has persisted. It has gotten to a point that I cannot play Diablo 3 for even 10 minutes without it crashing; however, the system will run for forever when doing normal computing tasks (surfing web, watching netflix, email, etc). The only time it crashes is when playing games; which leads me to lean towards the vid card as the issue. I don't know. I was hoping for an update, but out of the 2348975203945720394857 people experiencing this issue over the passed 2 years not one of them has explained what they did to solve the problem. If you know of any fixes for this or can point me in another direction or offer another solution besides the catalyst 13.1 drivers I would be incredibly thankful.
Tesla1856
8 Wizard
•
17.4K Posts
0
March 27th, 2013 13:00
Are you in USA?
Are you using a good UPS (like APC).
What is you hardware skill level? Are you able to work inside computer? If I asked you to "re-seat video card" or "check video card power plugs/numbers" ... could you do that without exact instructions and not blowing up computer?
goforthandconqu
9 Posts
0
March 27th, 2013 18:00
HaHa. I am in the USA.
Alienware™ 875 Watt Multi-GPU Approved Power Supply is the only info I have on UPS.
I'm pretty hardware savy even when I'm unfamiliar with product. I've been inside a few motorcycles and underneath a few cars in my day. If nothing else there is probably a "how to" video that shows and explains how to reseat the video card I can find online.
Last night I trolled through 2 years worth of posts on guru3d.com regarding the radeon hd7870 and found that there are about 10 gazillion people experiencing the same issue. It was suggested on this forum that the card ships at a higher clock than it was designed to run at? So I turned the core clock to 995 mhz and the memory clock to 1000mhz and allowed it to use up to 5% more power. Some of the people on the forum also suggested that the card shuts off due to a failsafe when it is being underpowered. After tweaking clockrate last night, I was able to play GW2 for aroud an hour with no crashes and D3 for 2 hours with no crashes. It didn't crash one time last night at all after the changes were made. I'll admit that it wasn't a long enough period of time to asses whether this change has fixed the problem; however, it was running exponentially longer than it has in the past few days. Still getting 60fps on D3 and 40-60 on GW2 according to fraps.
Also, I was looking through the logs from GPU-Z and even with this current setup the card is running between 95-99% of the GPU Memory. Is it common for a vid card to use almost all of its memory when gaming or watching movies?
Tesla1856
8 Wizard
•
17.4K Posts
0
March 27th, 2013 19:00
If the machine is crashing, so hard that Windows doesn't have a chance to record the error ... that usually means a hardware problem. See my posts here for trouble-shooting:
http://en.community.dell.com/owners-club/alienware/f/3746/t/19481860.aspx?PageIndex=1
Also, check power supply video card power plugs:
Assuming the power supply wire colors indicate the different 12v rails (going to the video cards):
Try with a power connector from each wire set at the same time.
Meaning, use either P14 OR P15
and also
either P16 OR P17
(read that carefully)
Tesla1856
8 Wizard
•
17.4K Posts
0
March 27th, 2013 19:00
Well, I have personally been pushing Dell to get to the bottom of this. It seems to have started with AMD 7xxx cards and Aurora R4 (x79 chipset and/or PCIe3.0). Could be so many things (motherboard, power, drivers, OS, etc.) ... and according to your findings ... maybe a flaw that worked itself into AMD reference design. I don't have a suitable machine here to investigate further ... but Dell and AMD do.
Tesla1856
8 Wizard
•
17.4K Posts
0
March 27th, 2013 19:00
Good work.
Not sure what you did (under-clocked it uses 5% more power ?) ... but if it becomes stable it's a start in the right direction.
This is what the project needs. Someone to start with the basics and work their way up to the more complicated things. Keeping notes and testing as they go. Methodically going though the system ... not jumping around as they see fit looking for an easy fix.
When not testing with Guild Wars 2 or Diablo 3 ... check out OCCT Power Supply test. With that running and maybe a long virus scan going at the same time ... if machine can survive that, it should be good and stable.
Tesla1856
8 Wizard
•
17.4K Posts
0
March 27th, 2013 19:00
I just never know what people's skill level is (or motivation level). Sometimes I recommend stuff and people feel "put-out" to have to troubleshoot anything when they can just return it.
No, not internal power supply. I meant UPS Battery (like APC unit) to make sure machine is getting good, clean, uninterrupted, high amps power (into power supply).
goforthandconqu
9 Posts
0
March 28th, 2013 16:00
no it doesn't use 5% more power underclocked; it is now set up to be allowed to use 5% more power than it was if needed. I'm going to give it about a week with current setting and no changes and if I'm still not having issues, I'll consider this current setup a fix. Afterwards, I'm going to begin tinkering to see if I can do some of the things you've suggested to get the card to function at its orginally shipped settings. I am not using a UPS battery. I can assume from your response that an APC unit is a good way to go as far as buying one?
Tesla1856
8 Wizard
•
17.4K Posts
0
March 28th, 2013 16:00
goforthandconqu
9 Posts
0
March 28th, 2013 16:00
stock settings are 1000mhz to core and 1200mhz to memory. currently running 995 to core and 1000 to memory. Spoke too soon though it seems. I just crashed like 10 mins ago while playing GW2 on max settings (v-sync on). GPU-Z logs aren't indicating anything strange happening as far as I can tell. If I sent you the logs from the crashes would you mind looking over them to see if there is anything significantly changing prior to the crash? Trying not to get frustrated, but its a pretty big bummer that it was working fine for a few days and now possibly mis-behaving again. I'm going to try to play with some more settings and tweak the clocks a little more later on tonight.
goforthandconqu
9 Posts
0
March 28th, 2013 16:00
APC is owned by schneider electric..which means I can get a pretty good deal on these through my work; however, I'm not sure which one I need? I was looking through their website and couldn't find anything to help me determine which once I need.
Tesla1856
8 Wizard
•
17.4K Posts
0
March 28th, 2013 16:00
Tesla1856
8 Wizard
•
17.4K Posts
0
March 28th, 2013 19:00