Jump to content

Genetics

Registered User
  • Posts

    122
  • Joined

  • Last visited

Posts posted by Genetics

  1. Hello everybody.

    I have an m15x with a 5850M GPU and am also interested in upgrading to an 7970M GPU.

    I have seen "upgrade kits" offers on eBay that go for 525$. They're for the m17x/m18x though. As I have read here, the bracket is too high. I would rather not add a cooper shim because it adds three more layers and needs paste on both sides. I would like to have the lowest temps for long-term reliability.

    Since I already have a 5850M GPU, could I just remove the X-bracket from my currenet GPU and swap it with that of the 7970M GPU? Is it easy to remove? Would there be no more use to grind the bracket or the screws down?

    Thanks in advance,

    Dakeryas.

    Hiya Dakeryas,

    1 : Afaik you cannot use X-bracket from 5850M for 7970M.

    2 : Default bracket for 7970M is good, I'm using it right now without grind it down.

    Thanks.

  2. I installed my 7970m a week ago and I'm having problems getting decent performance out of it. I'm running the 740Q CPU. With farcry3 I'm averaging around 30fps and dips down in the 20's at times. I've monitored my stock clocks they're fine. My temp's are fine too. I don't go far over 70.

    I just ran 3dmark11 and my combined score is low: Graphic score: 6293 Physics score: 2235 and Combined Score: 1972.

    Anyone know what might be the issue?

    What drivers are you using now? I also got something like this before. Then i switch to 12 Beta 11 by AMD and now seem pretty stable and good! Worth to give it a try bro!

    - - - Updated - - -

    RodEire: if you mean you have the internal display working but its the external that you cant get going. I have the same issue any time I connect to anything with older HDMI revisions or anything like an amplifier. If I go straight to my tv it works. I still haven't found a solution for this either but i have a media pc so it really doesn't matter to me.

    The one that's strange for me is my HDMI 1080p Samsung 23" monitor doesn't work through my DP adapter either. Same as when i connect to my denon AVR3806 that outputs to my 50" panasonic plasma. But connecting directly to my 50" panasonic 50PZ77 plasma works perfectly... Very strange. I bought my adapter off ebay and the company refunded me for the adapter as we assumed this the the problem. i dont have an original alienware one to test with unfortunately.

    Currently i am using this one --> http://accessories.us.dell.com/sna/productdetail.aspx?c=us&l=en&sku=330-1271

    Doesn't matter what TV or Monitor i attached to it, it works! Last time i tried with 39" inch LED TV and it works like Edward Cullen! :snicker:

  3. Glad its running better, my copper shim thickness is 1.5mm.

    If you need to test temperatures, try 3DMark11, Vantage, a game like BF3, Crysis 2, Metro2033, etc.

    I know you want to see how high the temperatures you can get but furmark is bad, even on desktop cards.

    Like mentioned above stay away from it. I haven't used furmark since the 400 series desktop GPUs.

    Your temperatures look better and look like you should be fine. I think 75C was the highest I saw which is fine.

    Yeah bro so far its running good. I will monitor and see how it goes in few days! Thanks again!!

  4. HWiNFO or MSI Afterburner have a feature which allows you to see the temps during games / benchmarks.

    @svl7 - Thanks for tips :Banane47:

    @ssj92 - I am quite happy now! After playing Far Cry 3 , seem like my system not shutting down anymore! Before my m15x shutting down on the introduction video on first mission! You can see my logs by TechPowerUp CPU-Z here http://goo.gl/ZLzDB

    Using 12.11 Beta 11 for the drivers!

    What do you think? It is stable now you think? :Banane36:

  5. Please guys... don't use Kombustor or Furmark. Use a game or a benchmark like 3dM11 for temperature testing. All that Kombustor/Furmark will do is stress the voltage supply circuits beyond usual levels. Not recommendable.

    @svl7 - Thank you for your reminder bro. It's just hard to monitor the temp by using the benchmark like 3dm11 or in gaming monitor temp. Will avoid using the Kombustor/Furmark in future! Thx again

  6. Okay let's back up for a minute. You're using the m17/18x backplate right? Did you grind down the screw posts or use a copper shim in between the gpu die and heatsink? If not, then your issues are from the thermal paste not contacting from the gpu die to the heatsink. This is what happened to me the first time and after placing a copper shim in between it was fine. My temps were 36-40C idle before the copper shim and would auto shutdown when I loaded the game. Now with the copper shim I get 28-32C idle and load can get up to 73C.

    First, thanks for the reply!

    1 : The temp improved with the HWINFO64! But still can reach to 84c after 2 mins running on Kombustor. Could you tell me what is the max RPM for the stock GPU fans M15x? Currently i set 5200 RPM when temp reach 80c.

    2 : I don't have the copper shim, it's hard to get here, but i do make one custom by myself, could you tell me what is the thickness (prefer by you) for the copper shim? I attached the custom copper shim for your review (It's not easy to do the sanding bro!)

    3 : Yes i am using m17x backplate, I don't grind down the the screw posts, but make the copper shim.

    post-6998-14494994033287_thumb.jpg

    post-6998-14494994032998_thumb.jpg

  7. If it stresses or sets clock yes. I don't use kombustor except on a desktop with gtx470 to auto load overclock. It's been fire and forget the last year.

    Curiously what are your idle clock temps like on the 7970m

    Sent from my DROID BIONIC using Tapatalk 2

    Idle clock temps very good i guess, its around 38-45c only. But if i use the Kombustor, the temp increasing like hell till 85c then turn off :72: Same goes if i'm on gaming, more load more faster the temp increase!

  8. Check gpu fan speeds when over 71 or 75c. Override with hwinfo64 to 5000rpm and recheck temps. If still high, repaste

    Sent from my DROID BIONIC using Tapatalk 2

    Tried this before but still failed, did the re-paste using different thermal including Arctic Silver 5 . Is it normal if i use MSI Kombustor the card will max to 850/1200 all the way? Thanks :)

  9. I'd say if you're not having any problems running at stock voltage, just run at stock then. If you must undervolt then try slowly by going one step below what your stock voltage is.

    Ex: my stock voltage is 1.050v so I would try a 1.025v voltage.

    UV to 0.950 will resolve your HEAT issues for sure bro, but the performance for the card will drop like shit :/ Running at stock voltage will turn off my laptop once the temp reach 85c, I am trying to hack the default copper with copper shim later and see how.

  10. Its probably because your card is unstable on full load at 1.0v. That's how mine was when I undervlted it from 1.050 to 1.00, it froze after awhile and after reverting back to stock its fine now but throttles again.

    Indeed. I am still thinking the long term solution to tackle down this problem. Currently i have no idea what else to do. Undervolt to 0.795 will drop the performance of the card. Any idea bro?

    Thanks

  11. @hitokiri_jaguar

    I got same issues with you. Using the default 7970M backplate (For Alienware M17x) for my m15x. Undervolt my card to 1.0V actually reduce the throttle and heat issues (At least my laptop now don't turn off if the temp reach 85c). Tried with many drivers 12.10,12.11 and some Beta version gives me BSOD. Now i am using A01 by Dell, seem the card is quite stable, but i still got 'freezing' issues when gaming (Medal of Honor Warfighter).

  12. @hitokiri_jaguar

    Welcome to the club! I am also using I7-720QM with AMD 7970M, i got overheat and throttling issues before. Your thermal pad is good. Its been 2 days and so far it is good (NO bsod, and the gpu load is good while i am playing COD Black Ops II and Medal of Honor Warfighter), the temp is around 70-75c during the gaming.

    You can try :

    1: Disable the SpeedStep (I think so) in your bios.

    2: I am using latest Beta drivers by AMD now.

    So far it is good..hopefully it will good until forever!

    Cheers bro!

  13. svl7,

    I am using the drivers from dell which is Gfx_AMD_A01_8.973_N3GM1_setup_ZPE, but RECENTLY my m15x caused BSOD (A clock interrupt was not received on a secondary processor).

    From your first post, i can see that you mention about the drivers A00. I am wondering whether A01 is the reason why m15x encounter BSOD at the moment. What do you say bro?

    Thanks.

  14. thanks for quick reply :)

    as i said, not only furmark cause high temp, but in heavy task games like sleeping dogs or MoH warfighter the system still shutdown midgame

    i tried repaste using Coolaboratory's Liquid Ultra and Laird's thermal pad on Memory module and VRM... but still no use, idle around 45-50, load increase slowly to 85-86* then shutdown immediately!

    I have even tried lift the battery side much higher.

    any suggest? i'm really frustrated now :(

    I got the same issue on m15x when first upgraded to 7970m. I solved the issue by re-fix the bracket for the card. Make sure it's properly tight to the GPU.

    Here is my GPU logs when playing MOHW on Ultra settings --> 7970m m15x - Pastebin.com

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use. We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.