2012 14" Dell Latitude E6430 [3rd,4C,Q] + GTX 1050 Ti @ 4Gbps-EC2 (EXP GDC 8.4) + Win8.1 [artnetx] // Optimus internal LCD acceleration
- Disabled wake on LAN, NIC onboard sequence and Integrated NIC in bios
- Uninstalled nvidia drivers with DDU
- Disabled in device manager "Xeon processor E3 - 1200 v2/3rd Gen Core processor PCI Express Root Port - 0151" (nvidia nvs 5200)
- Connected GTX 1050 TI
- System detected an unknown videocard
- Installed driver version 375.63
- GTX 1050TI was detected and working without problems, no errors
Check your CPU t° while gaming. I'm still surprise about the performances of the old true quads, with modern games the CPU goes on rampage and heat a lot.
To give you an example, i'm on mafia III atm with my QM2670@R9270, the cpu t° are 85°/90° already ingame, if i oc a bit the cpu/ram to 3.2ghz max turbo, the t° climb to 90°/95° on 4 cores on average. The limit since the t° max is 100°. Note that i unmounted, cleaned, and repasted the laptop 1 month ago!
So at the end, depending the cpu optimization and cpu load of a game, the cpu will heat more or less. In some games that are cpu intensive, it just shut down the laptop cause cpu or chipset t° too high.
Even a bad noname 450watt is enough to power decently a 1050, since it need only 10A on 12v rail. I don't think is a power issue, at least it should not.
Thx for feedback. Initially I thought it was because of the very old psu, but now it seems more of a software issue (possibly because of the driver), because both Skyrim and Fallout NV are using the same engine.
Yes it could be. I proposed you the one that involve the cpu t°, it is never bad to check the whole averages t° inagme. The operating system and drivers matter too like you said, maybe an issue with he engine with your actual os setup?
As i said, check the psu current specs under the 12v rail. If it is not a really bad one with like 5A@12v, 30A@5v and support 10A minimum on 12v it should be fine. With the dell psu it should be also more than fine for this card, power side!
I would exclude high cpu temperatures. With NVidia nvs5200 (dgpu) I saw temperatures climbing easily to 98*C under heavy loads. Last night I played Crysis 2 (very high settings at native 1600x900) for about 3 hrs and the temps never exceeded 80*C (measured with Core Temp). This is a remarkable achievement for my old laptop. I repasted cpu using Arctic Cooling MX-4 few weeks ago. Also the maximum cpu performance state is set to 98% to prevent it entering turbo boost mode.
Also my mistake, Skyrim and FalloutNV are using different engines, but the existing problem is related only to this two games.
Thx for the psu advice.
@wimzilla, the 22nm 3rd gen quad-core runs a lot cooler than the 32nm 2nd gen quad-core CPUs. My own testing finding a 45W 2nd gen would often hit TDP limits way on approach or just barely into the start of their turbo multipliers limits. The 3rd gen wasn't nearly as affected with a 45W i7-3720QM hitting turbo multiplier limits before TDP limits.
The 3rd gen i-cores are a great CPU. The are still performance competitive against their 7th gen equivalents. See source.
@artnetx So repasting with MX-4 made a significant difference? The 3740QM in my E6430 gets insanely hot under load, it's hit the 105'C mark at times. I've disabled TurboBoost in the bios and limited it to 2.4GHz in Windows, it sits between 80 and 90'C under load now, but I'd prefer getting full performance from it if possible.
I can test out Skyrim on my setup for you if you want, it's almost identical to yours, just has an EVGA 1050 instead, Win 10, and a 96W power brick. Fallout 4 runs fine on it though.
(Also, where can I get the A20 bios for the E6430? I've only been able to find A18. Thanks)
@nando4: Yep completely agree, i would prefer invest in a true quad old generations instead of the the actual dual i5/i7. The latest true quad i5/i7 are out of price atm, unfortunately. Performance per dollar side the Ivy bridge is awesome. Was just pointing on the t° issue due to the fact that some games demand much more cpu side.
I just repasted and changed the pad on the chipset this morning. I was quite dumb the last time i cleaned everything, i put an old 1mm thermal pad for the chipset instead the new arctic 0.5mm one freshly arrived. This morning i swapped the pads, repasted, t° drooped even further, mid 80° for the cores and 55°/60° for the chipset, under stress test with 100% fan! I use throttlestop to fix the turbo clocks, happy.
@artnetx: Usually when you start a game, can play a couple of minutes depending the load you apply and then everything shut down, mean an t° or power issue. So if the t° of the cpu/egpu are not in cause, then the psu could not be so reliable even with more than 10A@12v or your wiring setup not sustain heavy current under heavy load. Aka feed the egpu with only one connection, so 10A@12v through a couple of wires.
Or just a random game crash with your os, drivers, updates config.
@ybodhanya, MX-4 did not made any difference regarding minimal temps, i.e. 55-65 idle as always, but I observed a lot better thermal management under load. The cpu was way more likely to stay under 90*C in light to medium gaming. I re-enabled turbo boost in bios, but set maximum performance to 98%. Bios A20 is available starting 03/03/2017. http://www.dell.com/support/home/us/en/04/Drivers/DriversDetails?driverId=1DWW4
After that I refreshed few other drivers and ended up with additional few degrees on average on temps, but I cannot say if it's because of bios update or something else.
Yeah, regarding skyrim, would be glad to see how it performed. btw what drivers are you using? did you got all kind of errors (12, 43)? are you using setup 1.3x?