[Guide] Improving Battery Life on Windows [+Enabling Deeper C States]

Just wanted to make a post about broken package c states. I kinda found out, that at least for my case, they seem to be 99% related to the Intel iGPU (driver). I had a good talk with @unclewebb about this, and he also had issues with his Intel iGPU on his desktop, and noticed, when playing around with the Intel GPU tool and toggle back and forth power settings, or disable the iGPU in device manager,the 0% package c states got resolved.

This actually let me think, that my random broken package c states were also related to the Intel iGPU driver. I have actually random igfx crashes, which is the Intel GPU driver on my laptop, maybe happening 2-5 times per months or more. I just noticed by accident, that when it happened last time, it caused a 0% package c state for me, until Windows reboot. So I have the theory now, whenever or at least with high chance (cant proof so far it happens always), Windows 10 restarts the GPU driver after it crashed, it causes the c states to break.

Either this is a bug with the Intel driver, or Windows 10 that it doesnt reset the driver 100% correctly when it reloads it.

This actually all got to me, after I reported a bug I found a months ago in Chrome and reported it to the Chromium team. Actually I knew this bug existed over 1-2 years now, but I was always too lazy so far to report it. Here is the bug ticket:

https://bugs.chromium.org/p/chromium/issues/detail?id=1107403

Turned out, there is a bug in mostly every Intel iGPU gen6(or so I think starting from Broadwell) > <=gen9.5 which seem to be related to using swap chain for video decoding. Just read the bug ticket if youre interested in it.

As I see it, it doesnt happen just in Chrome but every app which uses a swap chain for decoding, maybe even Windows 10 itself, and the driver doesnt seem to be stable when you push it with high memory change.

You can actually fix this with a simple workaround for Chrome by starting Chrome with "--disable-features=DirectCompositionUseNV12DecodeSwapChain" and then check in chrome://gpu that it is disabled . Everyone using a laptop and having a <=UHD630 and using Chrome or Edge should do this until the workaround of disabling the swap chain gets live.

I dont think that is true. What is reported is correct. As I see it, there is nearly no difference between c8 and c10 in idle when laptop is on, at least for me. Either 80% c8 or 30% c10 are kinda identical, for power drain. Also for me, C10 just works with laptop powered on, when I remove the power cable and am on battery.

@unclewebb implemented a handy new feature in latest TS to debug c8 and c10 usage during modern standby and it will show the time it was in c8 and c10 for the last modern standby sleep.

 

Hello I just wanted to make my first post here saying that I finally got package c states working on my MSI GL62M (7700HQ 1050 TI). Previously it would only reach package c state 3 and now it’ll reach all the way up to package c state 8! It had always bothered me before that even enabling deeper c states in the bios it never went deeper then 3 but today I found out what was wrong. I went through device manager disabling devices and figured out after disabling my WiFi and Ethernet card it reached deeper c states at first I thought it was a driver issue. But eventually I enabled ASPM to auto on pcie root port 1 and 4 in the bios which were WiFi and Ethernet respectively and turned on L1.1 and L1.2. (Note Native ASPM also has to be enabled in ACPI settings, DMI Link ASPM control set to L1 in system agent/dmi configuration). Finally with all of this enabled in windows power options PCI Express/Link State Power Management set to maximum power savings will allow the computer to reach lower then c3 package.

I haven’t done full battery benchmarks but the CPU hovers around .5W idling on battery with no devices plugged in and in Microsoft Word with a moderate brightness windows was estimating over 6 hours of battery which is fairly good as getting 4 hours previously used to be a good day.

Also I’m running Windows 10 2004 with latest nvidia/intel graphics drivers since I saw people talking about that.

 

Now there is: https://www.reddit.com/r/AMDLaptops/comments/imgx7j/lenovo_slim_7_tweaking_guidereview/

Looks like we can undervolt and do some other magic in the BIOS

 

If I'm going to buy a Evo 970, what is the best for battery life:

1. AHCI mode with Samsung NVMe drivers installed
2. Intel RST mode
3. AHCI mode with generic Microsoft drivers installed ?

I read that Intel RST has some advanced power saving techniques not present in the AHCI mode. I have also heard the Samsung NVMe driver reduces battery life.

So I guess Option 2 would be the best?

 

Oops, didn't see this post. You need to try on your on hardware, because on all of my laptops different settings have different hardware configurations.



Also if anybody's bored or has the time, I'm interested in power consumption scaling of various processors over the years - Above is the AIDA 64 FPU stress test with CPUs set to various clock speeds (TPL > Speedshift Max, or Set Multiplier) and power consumption at that particular clock. Anybody else want to record and share info on their CPUs?

 

I did everything possible including updating Realtek SD card reader drivers and nothing helped until I updated Realtek Lan drivers. I hope this helps someone when they try to do troubleshooting on their end.

My laptop with i7-9750H is using 0.5 W at Idle for PKG Power with 65 - 75 C8%, 10 - 15 with C6%, 13.4 for c2% and rest is c3%.

 

I'll confirm the LAN drivers too.
On a clean install using the stock MS provided drivers my idle would never drop under 1W. After I installed updated drivers from Realtek (nothing else changed) my idle went to ~.5W - I've had it drop to as low as .3W (i7 6700HQ)

 

I was working on the OverPowered LP-2 (TongFang GK5CN5Z) model and the only way to bring it down from 1.8 W of power usage to 0.5 W was to install Intel Rapid Storage Technology (IRST or RST) drivers. I hope this helps someone when they try to troubleshoot their laptop.

 

Damnit.

One day I saw minimum cpu wattage as low as 0.195mw and discharge rate of -2.300ish mw.

Mustve been a fluke as its now at 0.225mw and -2.500mw discharge at lowest power.

Where o where did that came from.

Edit: Finally I think I managed to reduce everything to the absolute minimum.
-2.300~mw discharge with estimated 22+ hour on lowest brightness on a 13" gaming laptop

 

my gs63 is supposed to get 2 hours....now i get 6 thanks for this guide.....its really should be stickied

 

Hi Guys,

First of all, thank you so much for this thread. I can definitely see improvements on my Cyberpower Evo (basically an XMG clone from what I've heard) in terms of C0% (1-2% on idle) and PKG Power (1-2W) but after doing essentially everything in the guide and then some, I still have 2 big issues.

1 - The issue I originally was trying to solve when I found this was that sometimes when I turn on the laptop, the fan stayed on consistently, there was high CPU usage on one core (like 30% but still) and the laptop was constantly going from 40 - 50C. All of this on idle. I can't confirm if these steps have fixed that issue fully or not, but it still happened on one of my reboots where I had disabled the thunderbolt port in BIOS. I also ready somewhere in the thread that this might have some relation to waking from sleep instead of a dry boot? Has anyone found a fix for this issue as I might have to RMA the machine if I cannot consistently get rid of this high CPU bug. For reference, even before all the optimisations in this post, I get around 6-7 hours on balanced doing basic tasks when this bug doesn't happen.

2 - No matter what I do, I cannot get the CPU C-States out of C2 and C3. Right now as I type I have 95% C7% in Core, but my package states are around 20% C2 and 60% C3, "PKG" reads around 80. My C0% mostly sits between 1-2 but has random jumps to like 4 or 10 for a split second even on Idle. PKG Power is sitting around 1.5W.

Some of the things I have tried, other than the basic following of the guide (essentially all of which I've done):
- I have disabled my SD Card Reader slot in BIOS as I was getting a sleep error in powercfg energy report, the mentioned Registry Edits to fix this didnt seem to exist for the SD reader (I fixed it for another USB port).
- I've tried to disable thunderbolt, as I've noted some people mention it can be an issue here. Ditto for driver updates. Only seemed to make it worse disabling it as I mentioned earlier.
- Drivers installed via Windows update, Intel update tool, direct from sources like Realtek, and I even tried Driver Booster as a last resort.
- My CPU came from manufacturer with a -50 undervolt, I haven't modified this value but I did switch from "SpeedStep" to "Speed Shift" in the ThrottleStop interface @ 128.

Overall, I am happy with the power usage for now and can make further minor modifications in the future, but I need to get this random bug and the C states fixed first.

Does anyone have any suggestions of where to go from here? Only things I haven't really tried are a full Windows reinstall and an RMA, both of which I'd prefer to avoid if I can solve this myself.

Thanks in advance!

 

Note that sometimes disabling hardware in BIOS actually causes the CPU to not enter lower power states.

Basically most comes from either this, and appropriate drivers.

 

Yeah I've since reenabled both that and "Realtek USB 3.0 Card Reader". Does anyone know if there's a generic driver for this? With it reenabled the power report is saying "missing or misconfigured drivers". I think this could still be the issue.

On the bright side, I've managed to get mostly sub 1% C0 (as low as 0.5) with 1.3-1.4W idle. If I solve this power state issue i'll be in great shape (as long as the CPU bug doesn't come back).

 

You seemed to have tried everything. I'd be at wits end if I were you. Unfortunately this is one of the things that takes time to solve, through sheer experimentation and luck. If you love your machine, I recommend keeping it until a month before the warranty expires. If you haven't fixed the issue then, RMA and get a different laptop model.

 

Yeah a fresh windows install is looking like it might be the next thing, that or I just live with it. The CPU bug at least seems to be gone (knock on wood) and with basic tasks the battery indicator is estimating like 8.5 - 9 hours.

 

Quick update on this for those following. I did a full Windows Reset yesterday. I then installed all drivers direct from manufacturer, and a couple from Intel update manager and lastly windows update. I also did a handful of things from the guide (disabled some programs but not services yet, hipm+dipm, - 50 undervolt from manufacturer remained, switched from SpeedStep to Speed Shift).

Currently sitting with about 60% C8 residency on balanced, with about 1W PKG power and 2% C0, with plenty of room for improvements if I disable some more background. Battery indicator can go as high as 11.5 hours on idle, even more than the manufacturer "up to". Seems to have been a driver bug after all. Gonna keep an eye on it as I reinstall programs incase one of those was the culprit, but it's looking like it should be sorted now. I do also still have a powercfg energy error on some on the Card Reader but since it doesn't have the Selective Shutdown option anywhere in Registry I'll just have to let that one be!

Thanks again for all the help on this, and in putting this amazing guide together!

 

Working with a 2019 Alienware M17 that won't drop into anything lower than C3 (package). Powercfg reports an error with ASPM due to known driver incompatibility. I'm still getting 6 to 8 hours on battery, has anyone else been able to drop lower than C3 with this laptop (or the M15) and is this something I even need to concern myself with?

 

The ASPM issue is difficult to fix, and is almost always not the reason why your CPU can't get below C3.

On my Dell Latitude 3550 I had to switch from the Dell wireless to the AX200 Intel wireless which allowed it to reach deeper C states.

So might need change in hardware.

 

I recently upgraded to an AX200 from some killer wireless nonsense. I'm left somewhat clueless on this C-state issue, though I can't really complain about batter life even not being able to fully exploit the power-saving features of the chipset. I am always open to suggestions though...

 

You don't actually need to enable Speed Shift in TS, Windows 10 does it automatically as part of its power scheme policies. SpeedStep, on the other hand, should be kept enabled in TS.

 

On my old laptop - T480s, i5 8250u, undervolted 110mv - using speedshift at 180 I get about 0.5w package power on idle. Now I got a new laptop and I cannot get it below 1-1.8w fluctuating. Did everything the same, is it a settings problem, a problem of the laptop or is it a problem of tiger lake (i7-1165g7). Anyone else with tiger lake i7 here?

Also I noticed that the chosen frequencies in idle are much higher. I play around with park control but it only makes things worse.

 

Maybe ignore the park control. Your CPU should be idling under less than 0.5W package power. can you send a screenshot of your C state window? Over 80% of your package should be in C7/8/9/10 state.

 

Sounds like you aren't getting deeper than C3. Probably need to check all of your drivers and such.

What is your speed shift and Windows performance slider set to?

Are you plugged into HDMI output to an external monitor or USB C Thunderbolt? Those will both prevent deeper C states in my experience.

 

The CPU is at 0.3 to 0.4w, but the package power is 1-1.8w....

I did something to get it down to 0.5-1, don't know anymore what but had to reset windows due to another problem and now again it won't drop below 1w and just consumes too much.

Maybe the GPU is the main problem? That's included in the package, right? I cannot activate panel refresh else my screen flashes black (will have to exchange the laptop due to this but guess that's not related to the high package power).

It is like 85 Percent in c8

 

average C7 and above is actually 95% - and that's all in C8. But the power use is 1.5w instead of 0.5 for this screenshot.
It"s going between 1 to 2 watt all the time in idle for package power. With 190 for speedshift the frequencies are a bit lower - but that doesn't much affect the too big package power. I would prefer not using speedshift over 140 as that limits single core turbo to 3.1ghz instead of 4.1 (for 4.7 I need to go around 100). But even with 255 it;s not dropping much power - but then even scrolling in browser gets sluggish.

See the second screenshot from quick CPU - cpu power is low, 6 out of 8 cores are parked. But package power is very high for idle


Edit: just noticed - it"s all about panel self refresh. If I enable it then I drop about 1watt of poer usage and it idles at 0.5-1watt. However my laptop gets unstable - to the point that the screen goes black for long time - and I cannot even login to windows anymore. Only way to get back is then to go safe mode and uninstall intel drivers... But I guess that is because there is some hardware bug on my laptop. It shipped with panel self refresh enabled from LG.

If its not idle the difference is even bigger. Surfing the web is like 3-3.5 watt with panel self refresh enabled but 6-7 watt with it disabled.

 

Looks like you are pulling 1.5W at 2% CPU usage at idle which seems about right to me given your voltages and comparing to my 7th gen Intel CPU. At 2% CPU usage, while plugged into my external monitor and an extended desktop, I pull around 5.25 W.

It could very well just be whatever programs you have installed or monitoring apps such as TS / Task Manager / Quick CPU / etc keeping your CPU spooled up just a bit. I would recommend against parking cores myself. I keep my SST EPP at 0 for maximum responsiveness. Here is what my idle looks like when not plugged into anything or running anything extra besides TS:

 

well it's all about panel self refresh - if I disable that my cpu gets way too much power. my i5-8250 is according to the numbers still idling with less power compared to the i7-1065g7. I will have to return the laptop and wait for an exchange as it often becomes unusable with panel self refresh enabled. With it disabled battery life gets cut in real work by about 1/3 - so that is really the most important thing in the guide for idle and low demanding tasks. Also with it disabled I lose about 10% in benchmark like Cinebench R20 multi core - as I run into thermal throttling earlier. its about 200 mhz less speed at 21 watts (which is about the max cooling capacity for my laptop at 90 prochot)

 

That is unfortunate you have that experience. It sounds like the OEM may have cut some corners in their design and figured nobody would notice or care. Just curious though, if you tried using older Intel drivers or anything like that? It may also be an Intel driver quirk more-so than the OEM cutting corners too though, since 10th gen is still somewhat new compared to your other 8th gen or my 7th gen CPU. Also curious, what was the OEM specified battery life advertised for your machine and how close to the advertised battery life are you coming? Hope it works out for you.

 

well it's an lg gram 16. I think there is some hardware deffect or maybe just badly connector monitor cable. Because sometimes it works okay with panel self refresh activated, and it shipped activated. The quoted runtime is 22 hours - and with panel self refresh 12-16 hours of surfing/office is no problem. with it deactivated its more 9-12 hours. Other people don't have that problem. If I knew it's the monitor cable badly connected I would just do it myself, but maybe there is a bigger problem or broken cable that is okay as long as self refresh is deactivated.


Problem if I open up the laptop I void warranty - so I cannot check that but will have to return it to US (living in Taiwan right now - but amazon taiwan ships from US without free returns). It's 11 generation, they skipped 10. for notebooks.[/QUOTE]

 

After looking at it a bit more, well it's not only the panel self refresh. I configured my old laptop and the new one exactly the same when it comes to power option, all the cpu states and so on. But on watching a video via Chrome the package power is simply twice! Both have about the same resolution 2560x1600 vs old 2560x1400. The tiger lake is really much less efficient on doing such things. Even though I increased speedstep to 255 on the tiger lake it would be using about 4.5-5w package power - and use about 1300-1600mhz. Yes at the same time parking cores, and C0% being only about 30-40% on 4 threads. The old i5-8250 at any speedstepp epp value over 210 gets stuck in 800mhz, C0 = 100% but not dropping frames and using about 70% CPU, using about 2.5w core package. Yes it's undervolted too so its 0.5v core voltage against about 0.7v used by the 1165g7.

Both CPU pkg power is understated, however the correlation is correct. Idle at the same screen brightness would use about 5w on the T480s, now about 8-9w watching video (package power is in reality more 3-4w). The tiger lake laptop idles at 4-5w (so less power), now while watching video however it uses more power than the T480s and consumes 9-10w (so real package power is more like 6w). Both doing the same work. Really crazy, but the tiger lake in effect on low power tasks uses about 30-50% more energy than the i5-8250... This reverses on high power demand, when the i7-1165g can use about 15w for what the i5/8250u would need its full 28W package to get the same work done.

So tiger lake is a big disappointment when it comes to doing low power tasks! Maybe the i5-1135g7 is better? Well I hope to get a new lg gram in 2 weeks or so that time the i5-1135g7 instead. Let's see then. I know the i5 gives a bit better efficiency, maybe 10%. But that would still be quite far away from the older 8xxxU series.


A huge difference comes if i switch from Chrome to Edge. With edge the tiger lake actually only uses around 2.5w of package power (in reality maybe 3). Maybe I really should get used to edge instead of Chrome (and I already added some power saving stuff in chrome concerning interrupts over default - but that difference is simply huge). The 8000U was fine with Chrome, edge needs a bit less energy, but as it's on a different level to start with the difference isn't too big. I guess panel self refresh is just 0.5w per hour, about no matter the power level. Tiger lake just sucks on light stuff (while in general the newer lg gram has the less power hungry components on general, just gpu/cpu sucks in comparison. Not sure if this is mainly the graphics or the processor. I just feel intel tried too hard to make up the performance vs AMD and didn't care about battery life. They just did everthing no matter what settings to make tiger lake perform better in benchmarks to catch up. In real idle it's okay, but light workload became a battery hog...


Edit: well it seems to depend on the video and site. So edge is not always better than chrome. For Youtube and AV1 - it seems chrome is the same or marginally better. In general youtube seems to run well on Chrome and Edge, and with tiger lake AV1 clearly beats VP9 for power efficiency. So that's of course a point where tiger lake will trump the old processors, as they cannot decode av1, except by CPU which is crazy ineffecient. The horrible chrome engergy consumption was on viki.com...

 

Is it possible to lock the power package to 0.5 watt so it doesn't spike when doing stuff like browsing or office work? That should allow for crazy long battery life in theory, no?

Of course responsiveness would be much lower, but I still think the machine would be useable?

 

If you are on a recent gen Intel CPU (6th or 7th gen onward iirc), just set your EPP value from within ThrottleStop to 255 for that effect.

Pretty sure the newest versions of TS even allow individual EPP values per profile now. Of course if you want to put a hard cap on power consumption peaks, you can certainly set PL throttle points at a very low wattage also from within TS. However, keep in mind if you artificially constrain CPU power delivery to a very low wattage value, the CPU will fall to slowwwww clock speeds under any sort of moderate load.

Setting EPP to 255 has a very similar effect, but does not require a wattage cap to be explicitly set and will allow the CPU to run a little more consistently under various types of loads without as serious of risk of severely downclocking on moderate to demanding loads.

 

Sure, I already tried that. But it doesn't work, even when the EPP value is set to 255, it will draw over 1 watts when browsing. Same when I set PL limits. It says POWER THROTTELING, but it will draw more than 1 watt anyway.

 

I suspect it is using enough power to allow it to continue functioning. It has to use enough power to work and expecting to go lower might be outside of its functional limits. If you could set it too low it might stop working. I know a few years ago some laptops that had voltage and power adjustments accessible in the BIOS allowed manually setting the values too low and that actually bricked them because NVRAM could not be cleared and they would no longer power on.

 

PSA - Corsair's iCUE software will straight up kill your battery.

 

Epp 255 makes scrolling in browser sluggish as ****, and still uses 3-5watt package power doing light stuff. Actually Epp over 200 increases real world power consumption on 1165g7. That processor sadly is an utmost fail on battery except for watching movies where it excells. Or with low screen resolution. On WQHD, even worth UHD it just fails for office work. Set resolution to 1280*800 and I get nearly 40 percent less CPU power use.

180 seems a good compromise. Anything over 210 increases power use and is sluggish.

It's crazy when surfing, writing in notepad++ or working on a Windows remote desktop - so all things a mobile phone does at ease, CPU consumes about 60 percent of overall power, while nvme, 16 inch wqhd screen at around 230 nits and the rest of your laptop just need 40 percent...

But yes that laptop can show YouTube 4k av1 videos for 20 hours on a single charge with 230 nits screen brightness, but just gets 8-9 hours of work done on a 80 wh battery. Really everything but that Intel CPU is super efficient.

And on tiger lake no way to hard limit CPU power package. 5w cpu power package for simply scrolling up and down this forum in Edge. 7w in Chrome. 4w. for quickly moving the mouse. That is simply just insane. I'm sure apple M1 does this with 1w or so (apple M1 is actually using less power for webbrowsing than watching videos. That's like it should be. So crazy how intel botched this on ice lake and tiger lake. Comet lake was the most efficient from intel for light load...

 

Here's a CPU, undervolted -80mV, locked to 400MHz. Even then, max power consumption is 3.5W, for the CPU Package under a LinPack benchmark
When web browsing, you may need to consider GPU power for scrolling. the Iris Xe integrated graphics might not be as optimised as the UHD series

 

Yes, that is well possible. It's the GPU and CPU together that use a lot more power for tiger lake Vs intel uhd graphics.

The power package consumption for low power activities is higher both at idle and at low load.

On kaby lake refresh I had 0.3 to 0.5w idle, and surfing possible under 3w package power.

 

Well I'm really not sure it's the GPU. I think 2d stuff like scrolling is handled by the CPU. With GPU-Z, GPU power is showing at 0.1w. And somehow since the newest windows update I can activate Panel Self Refresh in the intel Graphics command center, however the power never goes below 1w on my 1165G7 (those times it went down to 0.5 were only a couple of hours since I own this laptop).

On low power I'm just consistently 1-2w above Kaby lake on package power. Seems the newest windows preview update improved it a bit.

 

Hello, this guide really helped me but I think that I am not at the lowest power possible yet. I started at 9w package power on idle, and now I get around 2w idle package power and 9w total power. My laptop is a Dell e5440 with a i5 4310u, and according to this i can get it down to 0.5w idle. I already disabled all the startup services I can, visual effects is set to lowest, I have a samsung 850 ssd so I enabled HIPM+DIPM+DEVSLEEP, I undervolted 80.1mv, most of my cpu c state is C7, intel igpu is set to max battery, and there are no extra devices connected to the laptop. Is there anything else I can do to decrease power draw? Would the DVD drive being present make a difference?

 

Hello, does anyone here with an alienware 17 r5 reach c states beyond c2? I have the i9, gtx1080, g-sync, tobii eye model but only reach c2 whether on battery or plugged in, I wonder if something is bugged and can it be fixed?

 

Could be a hardware issue - if this laptop does not have a MUX switch and the dedicated GPU is constantly on, this could cause the CPU to limit itself to the C2 state. I am not sure - can you check if your dGPU is constantly active?

 

Update after 1 year: Battery life has fallen to just over half. The total battery cycle count is just 53 over 3 years. Could the fact that the laptop was almost always plugged in and at 100% charge (Acer doesn't provide anyway to limit battery charge level) be a factor? I don't think heat is much of a factor as it's placed at the end opposite to the CPU and GPU and the laptop runs cool. I only play lightweight Indie games anyway and CPU and GPU temps. rarely exceed early eighties, usually in the 60-75°C range.
Right now the battery stops charging randomly at 80-85% and falls to zero straightaway from ~25% causing the laptop to shutdown. According to Acer battery is just fine. They choose to ignore the battery charging issues.

 

That doesn't sound normal. My parents keep their acer laptop plugged in year round, and after a few years their battery is still at 80% health.

Can you download AcerBattery.exe www2.acer.com.au/tools/AcerBattery.exe and share a screenshot? Also share BatteryInfoView screenshot

If you're battery is suddenly shutting down, you may consider a battery recalibration, where you drain the battery completely (Keep running it after it automatically goes to hibernate, and keep turning it on until you get the "BatteryCriticallyLow" BIOS message or something similar) and then charge all the way to 100%, then repeat a couple of times.

 

Thanks for the Acer battery monitoring tool. I wasn't aware of it. Discharging current, battery temp. and cycle count are nice extras.
I have observed battery health status as low as 52.6% in BatteryInfoView.
For the past few months battery never charges to 100%. Randomly stops charging at anywhere between 70-90% but mostly around the 85% mark. Yesterday the laptop shut down when battery fell immediately from 16% to 0%. That makes battery calibration quite difficult.I guess one of the cells is degraded. Only 5 days left of the extended warranty.

 

GET RID OF THE LAPTOP WHILST YOU CAN!!! Keep ringing up acer and tell them this is not normal. Don't settle for the "this is usual behavior crap." You need to return the laptop and get a refund where possible

Otherwise, you'll need to replace the battery yourself - usually about $50-100 AUD.

The behavior is very abnormal, so that might be an issue with the cells as you guessed or the BMS. Both unfixable via software unfortuantely

 

Not possible to return it as it's just short of the 3 year mark. I had bought it from Amazon India. The replacement battery costs around USD70-80. But considering how poorly the original battery fared I am hesitant to fork up the money for a new one. Acer confirmed that the extended warranty doesn't cover the battery. It's only warranted for the default 1 year. Unfortunately not many options for third-party batteries here and most of them don't inspire any confidence. If hope the battery doesn't degrade too much in the next year or so. I am looking forward to buying a Ryzen laptop.

 

I'm seeing a bit of weirdness on my Clevo 957HR from the new NVIDIA drivers, ie. anything newer than 457.20.
With 457.20 (or older), the GPU gets disabled if nothing is using it, and the GPU fans go quiet. This is as it should be.
With a newer driver, the GPU stays on all the time, GPU fans spinning (not full speed, because nothing is actually using the GPU, but at a constant medium-slow speed that is audible). This is annoying, but also very bad for battery life. Any tips/ suggestions appreciated. Hardware details in my sig.

I have tried using nvslimmer to get rid of the unnecessary stuff built into the driver. Hasn't helped.

 

Try uninstalling current one using DDU and let Windows update get the nvidia driver. I had issue with 980M with latest Geforce drivers where GPU stays ON all the time with optimus.Windows update driver somehow fixed the issue and boost clocks are better as well.

 

Windows Update gave me the 457.20 that I had already mentioned is the last one that works. Many newer drivers have come up since. The current driver released is 471.11 (there's a hotfix driver 471.22). Really frustrating not being able to use the latest drivers.