Guest Posted November 21, 2015 Share Posted November 21, 2015 I think @Meaker did and it was 1.35v, but even 780M/880M where already 1.35v IIRC... Quote Link to comment Share on other sites More sharing options...
Khenglish Posted November 21, 2015 Share Posted November 21, 2015 I think @Meaker did and it was 1.35v, but even 780M/880M where already 1.35v IIRC...Thank you for the info. I will correct nvidia's mistake tonight. 1 Quote Link to comment Share on other sites More sharing options...
Khenglish Posted November 21, 2015 Share Posted November 21, 2015 So increasing memory from 1.347V to 1.48V yielded a stable memory OC from 6030MHz to 6425MHz. Not quite what I was hoping for, but still welcome.NVIDIA GeForce GTX 980M video card benchmark result - Intel Core i7-3920XM Processor Extreme Edition,Clevo P170EM powered by PremaMod.com980m 1476 core/ 645 mem, 4.5ghz 3920xm with 2000MHz 9-10-10 T1 memory (memory is very important for 3dm11 physics and combined scores).If I can get 8ghz rated memory chips for cheap I will replace them. If you have a dead 970+... send me a PM. 1 Quote Link to comment Share on other sites More sharing options...
chap Posted December 1, 2015 Author Share Posted December 1, 2015 wow khenglish good to see you have a 980m now too. very nice score u got there, way better than mine :/ i think ill get clocky again when it gets really cold my best scores always went with sub zero cool air ^^ Quote Link to comment Share on other sites More sharing options...
omega53 Posted December 31, 2015 Share Posted December 31, 2015 What would you say is the highest safe voltage on the core for the 980m? I'm currently running 1.1v and only hitting the upper 60's on temps. Im curious how much more the voltage regs can take before it becomes not worth it Quote Link to comment Share on other sites More sharing options...
Khenglish Posted December 31, 2015 Share Posted December 31, 2015 1 hour ago, omega53 said: What would you say is the highest safe voltage on the core for the 980m? I'm currently running 1.1v and only hitting the upper 60's on temps. Im curious how much more the voltage regs can take before it becomes not worth it You're never going to damage the core with the Prema vBIOS. 1.2V is not nearly enough voltage for that. What you do need to be careful with is blowing up the power FETs at the top of the card. Some people have added the missing 3 combo FETs to not worry about that, but with an unmodified card you probably want to stay under 1.15V, depending on your temperatures. Quote Link to comment Share on other sites More sharing options...
duracell071 Posted January 8, 2016 Share Posted January 8, 2016 If I push my voltage above 1.087v my card (single 980m/p370em - stock heatsink) crashes withing about a minute, my temps are reading about 83 degrees max when fans are on full, could it be the memory gets too hot? or might this be the max my card can handle? Thanks Quote Link to comment Share on other sites More sharing options...
Khenglish Posted January 8, 2016 Share Posted January 8, 2016 3 hours ago, duracell071 said: If I push my voltage above 1.087v my card (single 980m/p370em - stock heatsink) crashes withing about a minute, my temps are reading about 83 degrees max when fans are on full, could it be the memory gets too hot? or might this be the max my card can handle? Thanks It should go higher. When I would push my card too far when it was unmodified it would power off without a driver crash, and I would need to restart to turn it back on. What voltage does the card reach 83C? Quote Link to comment Share on other sites More sharing options...
duracell071 Posted January 8, 2016 Share Posted January 8, 2016 At 1.087v) (with overclock) but it's a bit less now since repast. At stocks it's 74 on medium fan. If I go 1.2 then i get immediate shutdown, 1.1 and above I get driver crash even when I use my stable 1.087v clocks. I think I need to get better thermal pads, what do you think? If anyone could tell me exactly the size pads they are using and where they place each one, would be a huge help, for now I use stick but the pressure does not seem very strong like on my 670 which left an indent Quote Link to comment Share on other sites More sharing options...
duracell071 Posted January 10, 2016 Share Posted January 10, 2016 (edited) Found the problem, turns out when i push things to far and drivers crash, they get corrupted, once o reinstall them I can go to higher voltages with no problems. I've got temps down a lot by opening up the vents on the bottom, and tightening one screw on the gpu heatsink, do you think laping it would be worthwhile? and how do i go about it when it has those two feet under it? thanks Edited January 10, 2016 by duracell071 Quote Link to comment Share on other sites More sharing options...
duracell071 Posted January 15, 2016 Share Posted January 15, 2016 (edited) So I went ahead and lapped the heatsink, there is a temp improvement but nothing special. I checked the 980m's gpu core to see if the metal was flat and it is far from it, it is almost a dome, visible being higher in the middle then the edges when i place a razor blade on it. I'm currently running card at 1.112v with core +222 and mem +165 and my temps are between 81-85 degrees on heaven benchmark when fan goes to the upper level. How are these temps? should i expect better? I've heard of lapping the actual gpu but this seems risky..... thanks Edited January 15, 2016 by duracell071 Quote Link to comment Share on other sites More sharing options...
Unreall Posted March 4, 2017 Share Posted March 4, 2017 hi i have installed your mod prema nice work for that i can easy reach 1380mhz on core and 5500 on mem at 1,075v;P (62c) with liquid metal asic is 74,6% Buuut if i go further like 1,1 volt wothout any clock change after 5-10 mins it kicks my graphicscard out of the device manager and i have to shutdown windows and wait like 1 min to see it again in device manager whats wrong here? any one have any idea? i have the 240w on my alienware 17r1 (2014 Quote Link to comment Share on other sites More sharing options...
octiceps Posted March 4, 2017 Share Posted March 4, 2017 19 hours ago, Unreall said: hi i have installed your mod prema nice work for that i can easy reach 1380mhz on core and 5500 on mem at 1,075v;P (62c) with liquid metal asic is 74,6% Buuut if i go further like 1,1 volt wothout any clock change after 5-10 mins it kicks my graphicscard out of the device manager and i have to shutdown windows and wait like 1 min to see it again in device manager whats wrong here? any one have any idea? i have the 240w on my alienware 17r1 (2014 Sounds like the voltage is too high, causing instability and crashing. Quote Link to comment Share on other sites More sharing options...
Unreall Posted March 4, 2017 Share Posted March 4, 2017 or maybe the power suply is to weak ? i have 129.9 wattage on my gpu in hwinfo looks loike the gtx 980m lvl so i will need the 330watt powersuply? Quote Link to comment Share on other sites More sharing options...
octiceps Posted March 4, 2017 Share Posted March 4, 2017 3 minutes ago, Unreall said: or maybe the power suply is to weak ? i have 129.9 wattage on my gpu in hwinfo looks loike the gtx 980m lvl so i will need the 330watt powersuply? You can check system consumption at the wall with a Kill A Watt just to be sure (and factoring in PSU efficiency), but I doubt the rest of the system is using over 110W unless you're running Prime95 at the same time. I have seen my 980M use up to 150W when overclocked and it didn't exhaust my 230W PSU. But it's pretty well-known that 980M's power delivery is poorly built and can't handle much overvoltage without instability, and can black screen even at stock if VRMs overheat. Quote Link to comment Share on other sites More sharing options...
Unreall Posted March 5, 2017 Share Posted March 5, 2017 so my max voltage is 1.075v? i hope not ^^ is see many cards running 1.2 volt so if my voltage are to high it kicks the card out of the device manager? this is the crash dump : Workload Single init returned error message: File: device.cpp Line: 85 Function: void __cdecl eva::d3d11::device::pimpl::initialize(class eva::d3d11::device &,const struct eva::d3d11::device_creation_parameters &) Expression: D3D11CreateDevice( adapter, parameters.driver_type, software, flags, feature_levels, feature_level_count, D3D11_SDK_VERSION, m_native.receive(), &m_feature_level, m_native_immediate_context.receive()): DX11 call failed. Quote Link to comment Share on other sites More sharing options...
octiceps Posted March 7, 2017 Share Posted March 7, 2017 (edited) On 3/4/2017 at 6:03 PM, Unreall said: so my max voltage is 1.075v? i hope not ^^ is see many cards running 1.2 volt so if my voltage are to high it kicks the card out of the device manager? this is the crash dump : Workload Single init returned error message: File: device.cpp Line: 85 Function: void __cdecl eva::d3d11::device::pimpl::initialize(class eva::d3d11::device &,const struct eva::d3d11::device_creation_parameters &) Expression: D3D11CreateDevice( adapter, parameters.driver_type, software, flags, feature_levels, feature_level_count, D3D11_SDK_VERSION, m_native.receive(), &m_feature_level, m_native_immediate_context.receive()): DX11 call failed. Many 980M cards can't go much past 1.1v without black screening due to overheating VRMs, so I doubt 1.2v stable is even possible on an unmodded card. I believe @Mr. Fox actually added components to the PCB of his 980M so he could max out the voltage, so maybe he can chime in here. Edited March 7, 2017 by octiceps Quote Link to comment Share on other sites More sharing options...
senso Posted March 7, 2017 Share Posted March 7, 2017 I think it was @Khenglish that did the modded cards for @Mr. Fox. If I recall he as already said he added the missing mosfets in the VRM zone of the card, what mosfets did you use Kenglish(if you want to disclose)? Quote Link to comment Share on other sites More sharing options...
Khenglish Posted March 10, 2017 Share Posted March 10, 2017 (edited) On 3/7/2017 at 11:57 AM, senso said: I think it was @Khenglish that did the modded cards for @Mr. Fox. If I recall he as already said he added the missing mosfets in the VRM zone of the card, what mosfets did you use Kenglish(if you want to disclose)? 3x TI 87350d. This increases the current rating on the core from 120A to 240A. A 980m at 1.2V can pull up to around 240W. Edited March 10, 2017 by Khenglish 1 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.