Cheers guys.
Boy, isn't this an ironic twist of faith. Some of you may remember my posts where I stated how glad and happy I was with the 7970m performance and (somewhat) decent and improved driver support.
Anyhow, here's the deal:
Since today afternoon, the computer would suddenly shut down whenever I was on windows and for no apparent reason. I could be browsing, opening a file, you name it. Instashut down.
My setup : the classic P150EM with a 3610 and a 7970m.
Now, this isn't my first rodeo:
- checked RAM (I got two sticks, switch them around, tried all 3 possible combinations for them)
- checked Discs, windows said they were OK (amazing after so many sudden shutdowns)
- checked WhoCrashed - nothing on it from today (or even this week)
- on one run, I was able to run Kombustor (on the 7970m) for 20 minutes with no issue. Temps were around mid 80's, and when I closed the window, it shut down -_-'.
- on another run, I ran SpeedFan, and was checking the temps. Everything was pretty much iddleish (around 50ºC on a computer lab) and then poof! Shut down! I'd rule out temp related issues by now (such as bad stock paste job)...
- rebooted into safe mode, uninstall Intel and AMD drivers, and installed the good ole stock ones. To no avail.
- pretty much tried any combination of installing Intel drivers, AMD drivers (all these from the Clevo site) rebooting and even driversweep. I would either enter windows 8 with no issues (and no 7970m support) or it would shutdown right before the windows 8 lockscreen
- Both battery and PSU seem to be fine. No issues on safe mode with either and both working.
Now, this pretty much seems to me that my card is the issue. I had heard (and even had once or twice) about BSODs with the 7970m drivers. But a laptop simply shutting down, seems to me that it is way more serious. It might be something shorting, or the card is pulling more power than what it can get. And no, I'm not overclocking anything.
Ill consider any suggestions you guys might have. But most likely, I'm going to take this lappy to the store I bought it from and activate my warranty. Since I installed my ownd drives and RAM, I hope i won't get screwed by the seller (as he said that I would not void my warranty by doing so). I had the laptop running with that hardware for almost 4 or 5 months now with no (major) issues.
Thanks in advance!
-
i remember someone got the same problem no long ago and conclusion is rma
-
There's no reason to suspect the 7970M with your troubleshooting. Most likely it's something with the Mobo power line, because any other component would cause some symptoms (overheating, artifacts, BSODs, Windows error message ..etc). Looks like it's RMA unfortunately. Seems strange that it would work fine for a while then shut down, a fried component on the mobo would usually brick it with no possibility of switch on. Must be some exotic/rare issue.
-
I don't think your amd 7970m is responsible for this. It may be having some hardware conflict aka circuit problem (resistor problem, something on motherboard causing short circuit , etc) . I suggest first you open your laptop cover , bring hair dryer(set it to cool) blow air on motherboard from every direction. Also check if there is any naked wires that may touch your motherboard or any hardware . If anything like that just move them a little bit.Furthermore, dis assemble hardisk ,cd drom and assemble it back again . I want you to open and see you gpu , cpu thermal paste and check how much there is on each and touch it by your hand . If your thermal paste still feels like liquid then its still opertional , if not then your guy used cheap thermal paste. The last thing I want you to do is grab linux :
Download Ubuntu | Ubuntu
load it on cd drom or USB flash drive(I prefer usb driver) . If you are going to use usb , format it to ntfs , download this program,
UNetbootin - Homepage and Downloads
follow this:
How to create a bootable USB Flash Drive with UNetBootIn and use it to install Mac OSX, Linux, or any other Operating System | NightLion.net | NightLion
boot your computer from usb disk by select usb boot from boot option. Then once you loaded linux , I want to used it for 1 hour (browse internet , play simple games flash games, watch youtube ). If the laptop did not have this issue then we can assume 98% its a software issue meaning your windows has an issue. If otherwise you this issue appeared again this we can say it definitely a hardware issue . If your not familiar with linux Ubuntu UI , just go check some you tube videos first and then boot linux. I hope I helped you as much as I can. -
Support.3@XOTIC PC Company Representative
An odd issue for sure. I had a desktop that would shut down often and instantly and reseating the CPU fixed the problem. If you have some thermal compound around to reapply give that a shot. Might as well reseat the GPU while you're at it.
-
Although it sounds like it's not the 7970m and is likely a mobo issue, I would say go ahead and do the RMA or exchange anyway.
Those kinds of issues where it's difficult to determine a source (or if the source is as fundamental as the mobo), it's safer to RMA/exchange. The last thing you want is to bandage the problem until the warranty expires at which point you'll have to pay. -
-
agree with my precedessors here for it likely being a cpu/mobo/psu related issue. those three components are the only ones being able to cause a machine to shut down instantly without any bsods or warning signs like artifacts or error messages whatsoever.
as for the 7970m: unfortunately, many people seem to have adopted the strategy of blaming their 7970m whenever ANYthing goes wrong with their systems. its not meant to be an offense!its kinda understandable actually, considering all the reports about driver and gaming issues here. but that doesnt mean that this gpu can be held as the scapegoat for world hunger and waru get my point
Sent from my Galaxy Nexus using Tapatalk 2 -
Cheers guys! Thanks for all the answers.
I think on one part I may have not made my idea clear. I don't think the issue is the 7970m per se, but it is being triggered by it. It becomes evident from what I have been experimenting. Also, a small update on what I did since yesterday:
- My can of compressed hair is empty
. Also, my hairdryer has no "cool" function. Still, from what I can see, just the usual small ammount of fine dust on the fans, and nothing loose on the board.
- Decided to remove and reinsert my drives (I got an SSD on msata 3 and a HDD on my caddy). Oddly enough this allowed me to boot into windows with my 7970m working
. Although after resetting, the screen would just flicker (turning black and then on, just like when your updating your graphics driver). The next reboot, I was stuck at the same point.
- I'm able of using Windows 8 with all services normally IF I disable the 7970m from the device manager. Obviously I'm just running my HD4000, but I stressed both it and my CPU and got no shutdowns.
- I also checked the paste and the heatsink. They used IC Diamond compound, and applied a thin layer (not too thin). Seemed like a good paste job (I've done worse with good results). The paste wasn't dry yet.
- Haven't tried booting to Ubuntu yet. I'm running UEFI now, and since I have I haven't tried dual booting to linux. I just keep my linux related tools on a VM, and that serves me nicely.
Now for what I think I'm gonna do:
Like many of you suggested, an RMA seems to be in order. Moreso since the store I bought it from (where it was assembled and where they repair them) is like a 20 min walk from where I live. Hopefully, this will be all sorted soon enough (got a lot of coding to do and stuff to write). Maybe I'll finally get a decent pc for my work station at the uni cg lab.
One final note. If the issue, eventually, altough improbable, happens to be with the 7970m, I'll be very happy to have a new one. If they deem that I am not covered by my warranty, I'll probably get a 7970m as well. I really like this card, and would like to keep using it.
I'll keep you guys posted. Once again, thanks for all the insight, and for taking your time to help me! Much appreciated! - My can of compressed hair is empty
-
The plot thickens...
So, remember that store so-close-to-home that I posted earlier? It closed. And apparently posting a notice about a new address or if they have gone out of business altogether isn't something deemed important here in Portugal -_-'. The store has a "call-us-we'll-pick-the-laptop-at-your-place" service, but I didn't even tried it, cause I finally have some good news.
Good news is, I finally found the culprit:
Remember when I said I was running two Ram modules? Well one is an 8GB stick (bought it from clevo) the other is 2GB that I had lying around from an Asus G60 that I'm no longer using (It has no GPU, the Asus 260m GTX are reaaaaally hard to find, and pricey). I switched them around as I said previously and I still had problems. Even with just one of them I had a BSOD.
So I ran memtest (totally forgot about it earlier, it's really a must).
Memtest results mean that basically, one of the sticks is pretty much dead. Guess which one: the newly bought shiny 8GB ram stick, or the 3 year old 2GB stick that came from a busted laptop.
Yep, the 8GB ram stick is the issue -_-'. I'm now running only on the 2GB stick and it is fine (with the 7970m active, up until now). I think I have another 2GB stick lying around I'll probably fit in.
Earlier when I ran with the 2GB stick solo,I had a BSOD and reboot. I thought it was related with the issue at hand, but it was probably due to me messing around with the drivers in the first place. Also, I never had RAM issues shutting down a pc completly, so I overlooked that. Usually a BSOD, screwed 3D geometry, textures or even artifacts (when a GPU uses RAM as VRAM). I admit I never saw it comming.
I don't think it is worth RMAing my whole pc for a silly 8GB ram stick it came with. Considering the price of an 8GB stick (about 50€, and the time I'd lose sending it back, I'll bear with the limited RAM and will buy a couple of 8GB sticks later on. Perhaps if another issue arises, I'll just bundle the ram with it and RMA it. Right now I'm quite glad this is mainly over. Sorry for getting you guys worked up for nothing.
I still find it curious that the issue arised when switching to the 7970m. Like, always. Never noticed any issue when running solo with the HD4000. Even moreso, considering that the HD4000 uses RAM as VRAM. Should have tried Texture or Geometry heavy games, and not flash and simple OpenGL games. It is also possible that this isn't the only problem, but I'll find that out in due time, won't I?
Again, thank you for bearing with me, my ramblings and my issue. You have been most understanding guys.
Also, if a mod could change the thread title to [SOLVED] or something, it would be helpful!
Cheers! -
Glad to hear
you solved the problem . Anything else ? :thumbsup:
-
Nope, all good really!
Been playing Bioshock Infinite and GW2 for a few hours. The only downside is that I'm now using some pesky 2x2GB DDR3 1333 instead of the nicer 8 GB stick. But as long as I don't have to spawn my usual Ubuntu VM and Android VM at the same time I should be good ^^. Thanks for your thorough resolution process king601. Would have been my next step (considering I didn't RMA it).
Regarding my store, in their website WSI Bytes & Gadgets - Portáteis à sua medida they still keep the very same address (the one that is closed) and they seem to still be accepting orders. Think I'll call them on their mobile's to check if everything is OK with them.
Best regards. -
No problem friend
. Since your using temporary ram and you probably looking top but new rams. I suggest you get 32 GB of ram with 1866. Mghz speed. It will make you laptop night and day difference, especially you are running VM machine which eats lot of ram. I am surprised you are running VM machine you need mum l minimum 16 GB of ram .I myself run multiple VM machine for coding puprsoses sand play at the same time
Which takes at least 16 GB of ram
Sent from my Galaxy Nexus using Tapatalk 2 -
My VMs (except for the Android one) only use console. No window manager. They mostly run mongodb and apache with a few other services here and there. Although I'll admit sometimes things get messy (each PHP connection eats about 2 MB off of my RAM), I would get through. Never ran more than 3 VMs each with either 1 or 2GB for them (and usually one or two cores). Performance-wise it wasn't too bad, and it proves to be a really good and portable way to test my pre-production code on the go. If I need to work on another pc, I'll just take my VMs with me ^^.
I do confess I'm yearning for some of that 1866 goodness xD . A full 32 GB set will set me back at around 200€, around here. I will get there, I will, but for now, I'll be aiming closer to a 16 GB set (and throw in the crappy 2x2GB ones for a full 20GB). The difference is really noticeable, with programs opening and loading resources quite slower due to slower RAM... Still, better than no pc at all!
-
I'd still look into opening an RMA, any decent reseller wouldn't have a problem shipping you a new stick of RAM in return for your old one if it's dead.
-
-
Ah yes. The 7970m. We have a saying in my country which translates roughly into:
Having bad reputation is worse than actually being bad.
not blaming you though.
I would probably have thought the same had I been in your case -
.
So far so good. Apart from the excruciatingly slow RAM. That and the fact that I can barely have one VM running. I think next week I'll try opening an RMA for the stick. Could really use those extra 8 gigs...
-
The last time I got a random shut down, I tightened the CPU heatsink screws as they were very loose. My situation was probably different since I couldn't even start up my laptop, but see if it works for you.
-
. It would have been more hassle-free if it were the CPU heatsink screws not being tightened enough...
AMD 7970m failing [No overheat]
Discussion in 'Sager and Clevo' started by Jaycob, Apr 11, 2013.