Since I switched from LM18 to LM19 on my laptop, I have noticed a considerable drop of battery performance. Fully charged, I could run my laptop for ~ 3 hours under Linux Mint 18.3. Ever since I did a clean install of LM 19, that duration has dropped to ~ 1.25 hours.
TLP is installed and active with its defaults. As far as I can remember, that was what I did for LM18.
System = ASUSTeK Computer Inc. 1.0 K53SV
BIOS = K53SV.324
Release = Linux Mint 19 Tara
Kernel = 4.15.0-29-generic #31-Ubuntu SMP Tue Jul 17 15:39:52 UTC 2018 x86_64
Have others also noticed this behaviour?
Higher power consumption LM19 compared to LM18?
Forum rules
Before you post read how to get help. Topics in this forum are automatically closed 6 months after creation.
Before you post read how to get help. Topics in this forum are automatically closed 6 months after creation.
-
caolislay
Higher power consumption LM19 compared to LM18?
Last edited by LockBot on Wed Dec 28, 2022 7:16 am, edited 1 time in total.
Reason: Topic automatically closed 6 months after creation. New replies are no longer allowed.
Reason: Topic automatically closed 6 months after creation. New replies are no longer allowed.
-
caolislay
Re: Higher power consumption LM19 compared to LM18?
Interesting. Apparently no-one else experienced this kind of problem. Maybe I posted it in the wrong sub-forum, but then again no-one corrected me.
Three days ago, I did a clean install of LM 18.3 Cinnamon. Battery lasts ~ 3 hours again, running on a kernel 4.10.x. I updated to 4.15 after two days to check whether that made any difference. It didn't. Laptop runs for 3 hours even on latest kernel.
System = ASUSTeK Computer Inc. 1.0 K53SV
BIOS = K53SV.324
Release = Linux Mint 18.3 Sylvia
Kernel = 4.15.0-30-generic #32~16.04.1-Ubuntu SMP Thu Jul 26 20:25:39 UTC 2018 x86_64
Apparently the higher battery usage comes from the changes within the code of LM 19.
Three days ago, I did a clean install of LM 18.3 Cinnamon. Battery lasts ~ 3 hours again, running on a kernel 4.10.x. I updated to 4.15 after two days to check whether that made any difference. It didn't. Laptop runs for 3 hours even on latest kernel.
System = ASUSTeK Computer Inc. 1.0 K53SV
BIOS = K53SV.324
Release = Linux Mint 18.3 Sylvia
Kernel = 4.15.0-30-generic #32~16.04.1-Ubuntu SMP Thu Jul 26 20:25:39 UTC 2018 x86_64
Apparently the higher battery usage comes from the changes within the code of LM 19.
-
Ascaris
Re: Higher power consumption LM19 compared to LM18?
What desktop environment are you using?
I recently bought an Acer Swift 1 laptop, which comes with Windows 10. Naturally, I immediately installed Linux Mint on it (Cinnamon x64, 18.3). When Mint 19 came out, I performed the upgrade.
I hadn't really used the battery much before I upgraded to 19, but I did after that. I decided to test it... I installed Laptop mode tools (and verified the many power settings were on using Powertop) and i965-va-driver. I downloaded the "Big Buck Bunny" movie (which I had seen used as a laptop battery tester elsewhere) in 1080p, in h.264/.mov format, and had VLC play it on an endless loop until the battery ran out. I had it set to never dim or turn off the screen. I had one external mouse plugged in, wifi on, bluetooth on, LCD panel at 40% brightness.
I don't remember the exact figure, but it was very close to four hours.
I then booted into (ick) Windows 10 and tried the same there. With the same setup as in Mint Cinnamon, it went 7 hours on the nose before the battery ran out.
I know Windows tends to have better battery run times, but that's way too much.
I noticed that in the System Monitor, VLC was only using a tiny bit of CPU... but Cinnamon was consistently using more. What was Cinnamon doing? No idea.
I tried enabling the setting to disable composition in fullscreen and repeated the test in fullscreen, but there was no difference.
I installed Mint 19 Xfce x64 and set it up in the same way as Cinnamon had been. I repeated the test, and this time the looped video went 5 hours, 50 minutes. Not as good as Windows, but still far better than Cinnamon.
It is possible this additional power use I saw in Cinnamon is what you're seeing too, if you are using Cinnamon.
I recently bought an Acer Swift 1 laptop, which comes with Windows 10. Naturally, I immediately installed Linux Mint on it (Cinnamon x64, 18.3). When Mint 19 came out, I performed the upgrade.
I hadn't really used the battery much before I upgraded to 19, but I did after that. I decided to test it... I installed Laptop mode tools (and verified the many power settings were on using Powertop) and i965-va-driver. I downloaded the "Big Buck Bunny" movie (which I had seen used as a laptop battery tester elsewhere) in 1080p, in h.264/.mov format, and had VLC play it on an endless loop until the battery ran out. I had it set to never dim or turn off the screen. I had one external mouse plugged in, wifi on, bluetooth on, LCD panel at 40% brightness.
I don't remember the exact figure, but it was very close to four hours.
I then booted into (ick) Windows 10 and tried the same there. With the same setup as in Mint Cinnamon, it went 7 hours on the nose before the battery ran out.
I know Windows tends to have better battery run times, but that's way too much.
I noticed that in the System Monitor, VLC was only using a tiny bit of CPU... but Cinnamon was consistently using more. What was Cinnamon doing? No idea.
I tried enabling the setting to disable composition in fullscreen and repeated the test in fullscreen, but there was no difference.
I installed Mint 19 Xfce x64 and set it up in the same way as Cinnamon had been. I repeated the test, and this time the looped video went 5 hours, 50 minutes. Not as good as Windows, but still far better than Cinnamon.
It is possible this additional power use I saw in Cinnamon is what you're seeing too, if you are using Cinnamon.
-
lazarus
Re: Higher power consumption LM19 compared to LM18?
TLP or not, it may even be something as simple as default values have changed. eg. screen brightness.
I noticed similar with one of my laptops, running LM19 Cinnamon. I reduced screen brightness, idle times for screensaver, etc, etc. and even went so far as to shut off all system sounds. 'Til I noticed in pavucontrol that the volume slider was 150% As I had basically disabled sound everywhere else, I zeroed this too... and battery life went up! Not by much, but enough to notice. Apparently there is a difference between not playing sounds at 150% amplification to not playing sounds at 0% amp.
This made no difference whatsoever on another LM19 Cinnamon laptop though. So may or may not help you.
Also... have you checked Timeshift's configuration? I'm pretty sure that under 18.3 it doesn't default to enable snapshots on boot, but on the one LM19 Lappy I know I haven't altered beyond updating a fresh install it does do a boot snapshot. That's a significant few minutes of HDD access every time it boots, with corresponding power use...
Another possibility is wifi configuration. I'm unsure whether there's a difference here between LM18 & LM19, but if there is, more frequent polling would certainly shorten battery life. (I'd configured my NAS to prefer samba on a Win based LAN, then forgot all about it. As my current set up doesn't use smb, there was a period of confused handshaking before it'd connect. My bad.)
Perhaps monitoring with
I noticed similar with one of my laptops, running LM19 Cinnamon. I reduced screen brightness, idle times for screensaver, etc, etc. and even went so far as to shut off all system sounds. 'Til I noticed in pavucontrol that the volume slider was 150% As I had basically disabled sound everywhere else, I zeroed this too... and battery life went up! Not by much, but enough to notice. Apparently there is a difference between not playing sounds at 150% amplification to not playing sounds at 0% amp.
This made no difference whatsoever on another LM19 Cinnamon laptop though. So may or may not help you.
Also... have you checked Timeshift's configuration? I'm pretty sure that under 18.3 it doesn't default to enable snapshots on boot, but on the one LM19 Lappy I know I haven't altered beyond updating a fresh install it does do a boot snapshot. That's a significant few minutes of HDD access every time it boots, with corresponding power use...
Another possibility is wifi configuration. I'm unsure whether there's a difference here between LM18 & LM19, but if there is, more frequent polling would certainly shorten battery life. (I'd configured my NAS to prefer samba on a Win based LAN, then forgot all about it. As my current set up doesn't use smb, there was a period of confused handshaking before it'd connect. My bad.)
Perhaps monitoring with
top in a console would give you a better idea of whether something is constantly running differently to an LM18 machine?-
Ascaris
Re: Higher power consumption LM19 compared to LM18?
Andy,
Given that on my laptop that has the same issue, Cinnamon is always listed in the top few entries in System Monitor in Mint 19 when sorted by CPU usage (meaning it never drops to 0), while in 18.3 it would sometimes appear for a few seconds and then disappear into the sea of 0s, the same way that Xfce does in Mint 19 Xfce, I would think that would be a good place to start looking for the culprit. Whatever it is, it is apparently happening within the Cinnamon process space, and during my looped displaying video test, Cinnamon was usually the top CPU using process, above that of VLC, which was the program actually displaying the video.
If anyone has Mint 19 installed and does not have Cinnamon constantly using > 0 CPU, I would be very interested in knowing what the difference is. I've noted this behavior in all of the PCs I have running Mint 19 Cinnamon (three of them, total).
I would love to switch back to Cinnamon on my laptop, as I like it better than Xfce, but as long as it persists in draining my battery faster, it's a no go for me. On my older laptop that has such terrible battery life that I use it like a desktop, I still use Cinnamon, and I haven't gotten around to upgrading my actual desktop yet.
I tried rolling back to 18.3 with Cinnamon on the laptop (the same one I referred to in the first paragraph, not the older one), and while its battery life was better than Mint 19 Cinnamon, it was still not as good as Mint 19 Xfce during the video test, and I don't know if that's a function of the DE or the version of Mint. I read that there were some changes to hardware accelerated video on Intel iGPU setups in 18.04, so it could be that the Ubuntu base is responsible for the longer life I am seeing in Mint 19 Xfce vs. Mint 18.3 Cinnamon.
Given that on my laptop that has the same issue, Cinnamon is always listed in the top few entries in System Monitor in Mint 19 when sorted by CPU usage (meaning it never drops to 0), while in 18.3 it would sometimes appear for a few seconds and then disappear into the sea of 0s, the same way that Xfce does in Mint 19 Xfce, I would think that would be a good place to start looking for the culprit. Whatever it is, it is apparently happening within the Cinnamon process space, and during my looped displaying video test, Cinnamon was usually the top CPU using process, above that of VLC, which was the program actually displaying the video.
If anyone has Mint 19 installed and does not have Cinnamon constantly using > 0 CPU, I would be very interested in knowing what the difference is. I've noted this behavior in all of the PCs I have running Mint 19 Cinnamon (three of them, total).
I would love to switch back to Cinnamon on my laptop, as I like it better than Xfce, but as long as it persists in draining my battery faster, it's a no go for me. On my older laptop that has such terrible battery life that I use it like a desktop, I still use Cinnamon, and I haven't gotten around to upgrading my actual desktop yet.
I tried rolling back to 18.3 with Cinnamon on the laptop (the same one I referred to in the first paragraph, not the older one), and while its battery life was better than Mint 19 Cinnamon, it was still not as good as Mint 19 Xfce during the video test, and I don't know if that's a function of the DE or the version of Mint. I read that there were some changes to hardware accelerated video on Intel iGPU setups in 18.04, so it could be that the Ubuntu base is responsible for the longer life I am seeing in Mint 19 Xfce vs. Mint 18.3 Cinnamon.
-
lazarus
Re: Higher power consumption LM19 compared to LM18?
On my main LM18.3 PC, Cinnamon never, ever drops below 0.7%CPU or 1.1%RAM usage. The minimums are about the same on my LM19 laptops, although the peaks do seem to reach higher values. They're still similar enough that I can't say this is a LM18 vs LM19 based difference, as the OP is discussing.
But I agree with you that there are definitely major variances in power usage between the MATE, Cinnamon and KDE desktops of any given LM release... with Cinnamon being the hungriest of the lot. (I'm slowly rolling back my laptops to LM18.3 MATE for desktop speed & battery life reasons, although I also would prefer to keep them Cinnamon.)
I've assumed the OP is using the same DTE in both versions; comparing power usage of LM18.3 KDE to LM19 Cinnamon is akin to comparing badgers to beavers.
-
Ascaris
Re: Higher power consumption LM19 compared to LM18?
Andy,
It may be the same desktop environment in both releases (and I would strongly suspect it is) in the OP's PC, but knowing whether it's Mate, Xfce, or Cinnamon in both cases tells us a lot. If the problem is Cinnamon having gotten even more power hungry, we would expect that other DEs would have similar battery life in their 18.3 and 19 versions. If other DEs also showed a drop in life going 18.3 -> 19, it would suggest that there's something other than the DE at work.
I just opened up the System Monitor in Mint 18.3 (Cinnamon) on my desktop and found the same as you, it never drops to 0%. This has no noticeable effects on my desktop, though, as you might expect. On the laptop (the Swift, not the older one), when I looked at CPU usage in Cinnamon (Mint 18.3), it did drop to 0 periodically, as one would expect while it's idle. That in and itself raises a question as to why this difference in behavior is occurring on different PCs running the same version of Mint.
It may be the same desktop environment in both releases (and I would strongly suspect it is) in the OP's PC, but knowing whether it's Mate, Xfce, or Cinnamon in both cases tells us a lot. If the problem is Cinnamon having gotten even more power hungry, we would expect that other DEs would have similar battery life in their 18.3 and 19 versions. If other DEs also showed a drop in life going 18.3 -> 19, it would suggest that there's something other than the DE at work.
I just opened up the System Monitor in Mint 18.3 (Cinnamon) on my desktop and found the same as you, it never drops to 0%. This has no noticeable effects on my desktop, though, as you might expect. On the laptop (the Swift, not the older one), when I looked at CPU usage in Cinnamon (Mint 18.3), it did drop to 0 periodically, as one would expect while it's idle. That in and itself raises a question as to why this difference in behavior is occurring on different PCs running the same version of Mint.
-
LordSaddler
Re: Higher power consumption LM19 compared to LM18?
Same problem for me (as I written in a new thopics), placing mouse on battery icon, I have this result:
- Manjaro say 7 hours;
- Mint 19 Cinnamon from live usb say 7 hours;
- Mint 19 XFCE Installed say 7 hours;
- Mint 19 Cinnamon Installed on my pc say 4 hours (and actually it seems that it is draining too fast);
I am using TLP on installed system.
- Manjaro say 7 hours;
- Mint 19 Cinnamon from live usb say 7 hours;
- Mint 19 XFCE Installed say 7 hours;
- Mint 19 Cinnamon Installed on my pc say 4 hours (and actually it seems that it is draining too fast);
I am using TLP on installed system.
-
MintBean
Re: Higher power consumption LM19 compared to LM18?
I usually keep my laptop plugged in so I wouldn't know, although it wouldn't surprise me. Mint 19 is using significantly higher CPU than 18 at rest.
Re: Higher power consumption LM19 compared to LM18?
Same issue here... On Mint 18.3 my battery was lasting hour or so... maybe even 90 minutes (depending on the load),
Once done clean install of Mint 19 that time was halfed... I can hardly get 30 minutes out of it even on idle.
This also included change of HDD... to SSD... so imagine my surprise when I got less than 30 minutes of battery life with brand new SSD and Mint 19.
Old HDD with 18.3 still on it can last that 60-90 minutes still...
On Mint 19 I can hear CPU fan working all the time... even though temperatures do not require that high speed... and my laptop do not have FAN modules.... tried with lm-sensors... and Psensor... powertop.... what not
Also, tried with XFCE... but same result... (default DE is Cinnamon)
Once done clean install of Mint 19 that time was halfed... I can hardly get 30 minutes out of it even on idle.
This also included change of HDD... to SSD... so imagine my surprise when I got less than 30 minutes of battery life with brand new SSD and Mint 19.
Old HDD with 18.3 still on it can last that 60-90 minutes still...
On Mint 19 I can hear CPU fan working all the time... even though temperatures do not require that high speed... and my laptop do not have FAN modules.... tried with lm-sensors... and Psensor... powertop.... what not
Also, tried with XFCE... but same result... (default DE is Cinnamon)
Code: Select all
dragan@origin:~$ inxi -Fxz
System: Host: origin Kernel: 4.15.0-33-generic x86_64 bits: 64 gcc: 7.3.0
Desktop: Cinnamon 3.8.8 (Gtk 3.22.30) Distro: Linux Mint 19 Tara
Machine: Device: laptop System: CLEVO product: P15xEMx serial: N/A
Mobo: CLEVO model: P15xEMx serial: N/A BIOS: American Megatrends v: 4.6.5 date: 05/07/2012
Battery BAT0: charge: 23.4 Wh 92.7% condition: 25.2/77.0 Wh (33%) model: Clevo X51 status: Charging
CPU: Quad core Intel Core i7-2860QM (-MT-MCP-) arch: Sandy Bridge rev.7 cache: 8192 KB
flags: (lm nx sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx) bmips: 19953
clock speeds: max: 3600 MHz 1: 2893 MHz 2: 2893 MHz 3: 2893 MHz 4: 2893 MHz 5: 2893 MHz 6: 2893 MHz
7: 2893 MHz 8: 2893 MHz
Graphics: Card-1: Intel 2nd Generation Core Processor Family Integrated Graphics Controller bus-ID: 00:02.0
Card-2: NVIDIA GF114M [GeForce GTX 675M] bus-ID: 01:00.0
Display Server: x11 (X.Org 1.19.6 ) drivers: modesetting,nouveau (unloaded: fbdev,vesa)
Resolution: 1920x1080@59.93hz
OpenGL: renderer: llvmpipe (LLVM 6.0, 256 bits) version: 3.3 Mesa 18.0.5 Direct Render: Yes
Audio: Card Intel 7 Series/C216 Family High Definition Audio Controller
driver: snd_hda_intel bus-ID: 00:1b.0
Sound: Advanced Linux Sound Architecture v: k4.15.0-33-generic
Network: Card-1: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller
driver: r8169 v: 2.3LK-NAPI port: d000 bus-ID: 03:00.2
IF: enp3s0f2 state: down mac: <filter>
Card-2: Qualcomm Atheros AR93xx Wireless Network Adapter driver: ath9k bus-ID: 04:00.0
IF: wlp4s0 state: up mac: <filter>
Drives: HDD Total Size: 240.1GB (26.7% used)
ID-1: /dev/sda model: SUNTRSI_SSD_S660 size: 240.1GB
Partition: ID-1: / size: 38G used: 15G (41%) fs: ext4 dev: /dev/sda1
ID-2: /home size: 182G used: 46G (27%) fs: ext4 dev: /dev/sda2
RAID: No RAID devices: /proc/mdstat, md_mod kernel module present
Sensors: System Temperatures: cpu: 94.8C mobo: N/A
Fan Speeds (in rpm): cpu: N/A
Info: Processes: 269 Uptime: 27 min Memory: 2864.2/15934.7MB Init: systemd runlevel: 5 Gcc sys: 7.3.0
Client: Shell (bash 4.4.191) inxi: 2.3.56
Re: Higher power consumption LM19 compared to LM18?
Just to confirm that it is LM 19 issue.
I am reporting now from clean 18.3 installation, with all available updates, including Kernel and Cinnamon... and CPU fan is not "screaming" like it was on LM19.
Battery lifetime is back to 1.1 hours, as it was... and I have managed to install and update LM 18.3 on one battery charge... with 15 minutes to spare.
Have not added Nvidia driver... but I use Intel one anyway....
I am reporting now from clean 18.3 installation, with all available updates, including Kernel and Cinnamon... and CPU fan is not "screaming" like it was on LM19.
Battery lifetime is back to 1.1 hours, as it was... and I have managed to install and update LM 18.3 on one battery charge... with 15 minutes to spare.
Have not added Nvidia driver... but I use Intel one anyway....
Code: Select all
dragan@Origin ~ $ inxi -Fxz
System: Host: Origin Kernel: 4.15.0-33-generic x86_64 (64 bit gcc: 5.4.0)
Desktop: Cinnamon 3.6.7 (Gtk 3.18.9-1ubuntu3.3)
Distro: Linux Mint 18.3 Sylvia
Machine: System: CLEVO (portable) product: P15xEMx
Mobo: CLEVO model: P15xEMx
Bios: American Megatrends v: 4.6.5 date: 05/07/2012
CPU: Quad core Intel Core i7-2860QM (-HT-MCP-) cache: 8192 KB
flags: (lm nx sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx) bmips: 19953
clock speeds: max: 3600 MHz 1: 965 MHz 2: 911 MHz 3: 1283 MHz
4: 909 MHz 5: 875 MHz 6: 1008 MHz 7: 1312 MHz 8: 1481 MHz
Graphics: Card-1: Intel 2nd Generation Core Processor Family Integrated Graphics Controller
bus-ID: 00:02.0
Card-2: NVIDIA GF114M [GeForce GTX 675M] bus-ID: 01:00.0
Display Server: X.Org 1.18.4 drivers: intel (unloaded: fbdev,vesa) FAILED: nouveau
Resolution: 1920x1080@59.93hz
GLX Renderer: Mesa DRI Intel Sandybridge Mobile
GLX Version: 3.0 Mesa 18.0.5 Direct Rendering: Yes
Audio: Card Intel 7 Series/C210 Series Family High Definition Audio Controller
driver: snd_hda_intel bus-ID: 00:1b.0
Sound: Advanced Linux Sound Architecture v: k4.15.0-33-generic
Network: Card-1: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller
driver: r8169 v: 2.3LK-NAPI port: d000 bus-ID: 03:00.2
IF: enp3s0f2 state: down mac: <filter>
Card-2: Qualcomm Atheros AR93xx Wireless Network Adapter
driver: ath9k bus-ID: 04:00.0
IF: wlp4s0 state: up mac: <filter>
Drives: HDD Total Size: 240.1GB (2.9% used)
ID-1: /dev/sda model: SUNTRSI_SSD_S660 size: 240.1GB
Partition: ID-1: / size: 38G used: 6.4G (18%) fs: ext4 dev: /dev/sda1
ID-2: /home size: 182G used: 117M (1%) fs: ext4 dev: /dev/sda2
RAID: No RAID devices: /proc/mdstat, md_mod kernel module present
Sensors: System Temperatures: cpu: 53.8C mobo: N/A gpu: 37.0
Fan Speeds (in rpm): cpu: N/A
Info: Processes: 224 Uptime: 20 min Memory: 796.8/15934.7MB
Init: systemd runlevel: 5 Gcc sys: 5.4.0
Client: Shell (bash 4.3.481) inxi: 2.2.35
-
Ascaris
Re: Higher power consumption LM19 compared to LM18?
DraxNS,
That's interesting. even in XFCE, you get the poor battery life. By what you wrote, it seems that you installed Xfce into an existing Mint Cinnamon setup, right? Whatever bit is slurping up that battery in Cinnamon (Muffin?) is still there on your computer, it seems. Installing a new DE isn't like using a version of Mint that had that DE from the start... lots of stuff doesn't get replaced, in order to keep the old DE working too.
Mint 19 Xfce installed cleanly greatly improved my battery life (better than Cinnamon on Mint 18.3), using TLP on both. I got even better (slightly) results with Kubuntu 18.04, perhaps somewhat shockingly, given KDE's reputation for being a resources hog (it actually has a footprint much the same as Mint 19 with Cinnamon on my laptop).
That's interesting. even in XFCE, you get the poor battery life. By what you wrote, it seems that you installed Xfce into an existing Mint Cinnamon setup, right? Whatever bit is slurping up that battery in Cinnamon (Muffin?) is still there on your computer, it seems. Installing a new DE isn't like using a version of Mint that had that DE from the start... lots of stuff doesn't get replaced, in order to keep the old DE working too.
Mint 19 Xfce installed cleanly greatly improved my battery life (better than Cinnamon on Mint 18.3), using TLP on both. I got even better (slightly) results with Kubuntu 18.04, perhaps somewhat shockingly, given KDE's reputation for being a resources hog (it actually has a footprint much the same as Mint 19 with Cinnamon on my laptop).
Re: Higher power consumption LM19 compared to LM18?
Yes.. I have installed Xfce alongside Cinnamon... just as a test
Anyway... I had a breakthrough while testing 18.3 and loading drivers....
I have installed Nvidia driver, and intel-microcode and after reboot same thing happened as in 19.0... cooler started screaming again...
So with 50:50 chance, I have choosen to remove intel-microcode and reboot again.... and BINGO!
In 18.3 (honestly do not remember if I have done that also on 19.0) IF you install intel-microcode.... you can expect such battery draining, cpu fan over spinning behaviour....
So, DO NOT install it, or remove it just to see if it will help.
Done wonders for me
Anyway... I had a breakthrough while testing 18.3 and loading drivers....
I have installed Nvidia driver, and intel-microcode and after reboot same thing happened as in 19.0... cooler started screaming again...
So with 50:50 chance, I have choosen to remove intel-microcode and reboot again.... and BINGO!
In 18.3 (honestly do not remember if I have done that also on 19.0) IF you install intel-microcode.... you can expect such battery draining, cpu fan over spinning behaviour....
So, DO NOT install it, or remove it just to see if it will help.
Done wonders for me
-
Ascaris
Re: Higher power consumption LM19 compared to LM18?
The first microcode release for my laptop (Acer Swift 1) was released by Ubuntu on the 24th of August, 2018, which was after I'd already switched to Xfce because of the battery life issue. I first noticed the issue and posted in the thread August 9.
This laptop never got a Spectre microcode update, FWIW, so until the 24th, I was on the same one that was in the initial release of the system firmware, as far as I know. I'm now in Kubuntu 18.04 on this laptop, and the battery life is much better than it was on Mint with Cinnamon 19, sadly. Xfce is usable but a bit too minimalist for my liking, and KDE has sanded off the rough edges that kept me away before... I wish Mint still had a KDE edition, but I understand why they dropped it. Anyway...
The microcode is different for each Intel CPU, so something that affects one won't necessarily affect other Intel CPUs that got an update from the same package. So far, so good with the August 24 microcode on my Sandy desktop and Apollo Lake laptop.
This laptop never got a Spectre microcode update, FWIW, so until the 24th, I was on the same one that was in the initial release of the system firmware, as far as I know. I'm now in Kubuntu 18.04 on this laptop, and the battery life is much better than it was on Mint with Cinnamon 19, sadly. Xfce is usable but a bit too minimalist for my liking, and KDE has sanded off the rough edges that kept me away before... I wish Mint still had a KDE edition, but I understand why they dropped it. Anyway...
The microcode is different for each Intel CPU, so something that affects one won't necessarily affect other Intel CPUs that got an update from the same package. So far, so good with the August 24 microcode on my Sandy desktop and Apollo Lake laptop.
Re: Higher power consumption LM19 compared to LM18?
caolislay and DraxNS,
From your posted info, both of you have Optimus laptops. There is a bug in the operation of nvidia-prime in Ubuntu 18.04/Mint 19, where the Nvidia card is not powered off in Intel Mode, as it was in Mint 18.3, resulting in a greater battery drain.
https://bugs.launchpad.net/ubuntu/+sour ... ug/1765363
For a workaround, see my discussion on Power Management in my fifth post in this topic:
viewtopic.php?f=59&t=275967
From your posted info, both of you have Optimus laptops. There is a bug in the operation of nvidia-prime in Ubuntu 18.04/Mint 19, where the Nvidia card is not powered off in Intel Mode, as it was in Mint 18.3, resulting in a greater battery drain.
https://bugs.launchpad.net/ubuntu/+sour ... ug/1765363
For a workaround, see my discussion on Power Management in my fifth post in this topic:
viewtopic.php?f=59&t=275967
Re: Higher power consumption LM19 compared to LM18?
roblm
with current setup, on 18.3 with recommend nvidia driver installed, and intel-microcode uninstalled I have expected battery life.
Since it is daily work laptop, I do not have time to install 19.0 again and try same setup ... to find out exactly which part is causing this issue.
Again, in my personal experience, this is how I got it running.
Uninstalling nvidia driver is something I can also try while still on 18.3 to see if there will be even better battery authonomy. And while at it, I can also try no nvidia with intel-microcode installed. But last time it resulted with non functioning speakers... I guess that removing intel-microcode broke something related to sound...
Alsa mixer and all other mixers were showing all things OK, but I could not have sound on speakers. At the same time, headphones jack was working fine...
To try to fix that, I have opened laptop to see if something was actually broken, removed battery, returned old HDD, booted from it and got sound. After returning new SSD with current 18.3 install... Sound was working again.
Before doing this, I have tried fiddling with alsa mixer, rebooting laptop... Even started live session... All without result... Which I found odd..
But anyway, now it is all working as before... So ...if it is not broken... I will not attempt to fix it
with current setup, on 18.3 with recommend nvidia driver installed, and intel-microcode uninstalled I have expected battery life.
Since it is daily work laptop, I do not have time to install 19.0 again and try same setup ... to find out exactly which part is causing this issue.
Again, in my personal experience, this is how I got it running.
Uninstalling nvidia driver is something I can also try while still on 18.3 to see if there will be even better battery authonomy. And while at it, I can also try no nvidia with intel-microcode installed. But last time it resulted with non functioning speakers... I guess that removing intel-microcode broke something related to sound...
Alsa mixer and all other mixers were showing all things OK, but I could not have sound on speakers. At the same time, headphones jack was working fine...
To try to fix that, I have opened laptop to see if something was actually broken, removed battery, returned old HDD, booted from it and got sound. After returning new SSD with current 18.3 install... Sound was working again.
Before doing this, I have tried fiddling with alsa mixer, rebooting laptop... Even started live session... All without result... Which I found odd..
But anyway, now it is all working as before... So ...if it is not broken... I will not attempt to fix it
-
Ascaris
Re: Higher power consumption LM19 compared to LM18?
I don't have an Optimus laptop. The OP has a discrete Nvidia, which nearly always means Optimus, but my laptop (the Swift) is a Apollo Lake SoC model with Intel integrated graphics only. That's the one I have been testing for this thread.
The other laptop in my sig is straight Nvidia, no Intel integrated, and its battery life is so short that I haven't tested it with Cinnamon vs. anything else. Between the 35w TDP CPU and the 35W TDP GPU, it uses up a lot of power even at idle (~25 watts). It is still useful (and slightly faster than my Swift in most things, even though it's 10 years older), but miserly with power it is not.
The other laptop in my sig is straight Nvidia, no Intel integrated, and its battery life is so short that I haven't tested it with Cinnamon vs. anything else. Between the 35w TDP CPU and the 35W TDP GPU, it uses up a lot of power even at idle (~25 watts). It is still useful (and slightly faster than my Swift in most things, even though it's 10 years older), but miserly with power it is not.
Re: Higher power consumption LM19 compared to LM18?
...microcode has absolutely nothing to do with either graphics or even more with...sound loss.
Furthermore, before released to the public, it's been previously tested by:
(partially by) Debian, then (partially by) Canonical, previously also (partially by) the vendors,
and of course, very extensively by Intel itself.
The resulting fluctuation in temp between different versions should be around 1-2°C in normal circumstances -
i'd actually be surprised if it is 3-4°C under semi-exotic hardware combinations.
Someone truly has to hit the jackpot for a different result than that (report back to your vendor if such is the case).
Yet, there is a...41°C degrees difference in the inxi reports above...surely i find it rather impossible we've hit the jackpot here?
So...before unfairly demonizing microcode (for yet another time in those forums)...inxi compare above:
On a side note, using Cinnamon 3d accel over llvmpipe...well, what would you expect?
Especially if it already lasts only 60-90 mins maximum as you say...
No objection at all in that Bionic / Mint 19 underlying base requires more memory than Xenial / Mint 18, and also,
quite likely, that it also utilizes the processor more heavily.
But that again should give a 5-10% at worst case scenario fluctuation - not draining the battery in...half the time it required before.
If indeed such is the case 'out-of-the-box':
1) A kernel version / revision got severely broken somewhere & is not playing nicely - try changing it.
2) Way more probably (as per above - check roblm's post), something needs to be tweaked / fixed in your graphics' setup.
Furthermore, before released to the public, it's been previously tested by:
(partially by) Debian, then (partially by) Canonical, previously also (partially by) the vendors,
and of course, very extensively by Intel itself.
The resulting fluctuation in temp between different versions should be around 1-2°C in normal circumstances -
i'd actually be surprised if it is 3-4°C under semi-exotic hardware combinations.
Someone truly has to hit the jackpot for a different result than that (report back to your vendor if such is the case).
Yet, there is a...41°C degrees difference in the inxi reports above...surely i find it rather impossible we've hit the jackpot here?
So...before unfairly demonizing microcode (for yet another time in those forums)...inxi compare above:
Graphics: Card-1: Intel 2nd Generation Core Processor Family Integrated Graphics Controller bus-ID: 00:02.0
Card-2: NVIDIA GF114M [GeForce GTX 675M] bus-ID: 01:00.0
Display Server: x11 (X.Org 1.19.6 ) drivers: modesetting,nouveau (unloaded: fbdev,vesa)
OpenGL: renderer: llvmpipe (LLVM 6.0, 256 bits) version: 3.3 Mesa 18.0.5 Direct Render: Yes
Sensors: System Temperatures: cpu: 94.8C mobo: N/A
...Fairly obvious - all in all, it's absolutely totally different drivers loaded.Graphics: Card-1: Intel 2nd Generation Core Processor Family Integrated Graphics Controller
Card-2: NVIDIA GF114M [GeForce GTX 675M] bus-ID: 01:00.0
Display Server: X.Org 1.18.4 drivers: intel (unloaded: fbdev,vesa) FAILED: nouveau
GLX Renderer: Mesa DRI Intel Sandybridge Mobile
GLX Version: 3.0 Mesa 18.0.5 Direct Rendering: Yes
Sensors: System Temperatures: cpu: 53.8C mobo: N/A gpu: 37.0
On a side note, using Cinnamon 3d accel over llvmpipe...well, what would you expect?
Especially if it already lasts only 60-90 mins maximum as you say...
No objection at all in that Bionic / Mint 19 underlying base requires more memory than Xenial / Mint 18, and also,
quite likely, that it also utilizes the processor more heavily.
But that again should give a 5-10% at worst case scenario fluctuation - not draining the battery in...half the time it required before.
If indeed such is the case 'out-of-the-box':
1) A kernel version / revision got severely broken somewhere & is not playing nicely - try changing it.
2) Way more probably (as per above - check roblm's post), something needs to be tweaked / fixed in your graphics' setup.
-
carum carvi
Re: Higher power consumption LM19 compared to LM18?
I am rejuvenating this post, because I am interested in the above findings. Recently I bought a secondhand Thinkpad laptop. I have installed LM18.3 and LM19.2 alongside in a dual boot configuration.
Simply browsing text based websites (with images and videos disabled) lasted for 2 hours and 30 minutes in LM 19.2. I will report back tomorrow with the battery test of LM 18.3. Similar experiment. Only browsing text based websites (with video and images disabled by Ublock Origin).
Simply browsing text based websites (with images and videos disabled) lasted for 2 hours and 30 minutes in LM 19.2. I will report back tomorrow with the battery test of LM 18.3. Similar experiment. Only browsing text based websites (with video and images disabled by Ublock Origin).
-
mint_kiwi
Re: Higher power consumption LM19 compared to LM18?
Another info point: I have 2013 MacBook Air (6,1), previously with Mint 17. I installed Mint 19.1 to look for gotchas when I buy replacement hardware. I never needed TLP on Mint 17, but 19.1 runs much hotter, especially when the SSD got up to 70C doing a Handrake DVD rip and I had to stop it. Normally SSD runs around 60~65C on Mint 17 while doing a DVD rip. With TLP installed fans kick in sooner and harder, but temperatures are still abt 5C higher than Mint 17. Just sayin'.