7 ms·
> Laptops don’t change > New laptops may be more energy-efficient per computational power, but these gains are offset by more computational power Guess this w
by eBombzor 3y ago
> Laptops don’t change
> New laptops may be more energy-efficient per computational power, but these gains are offset by more computational power
Guess this was written before the M1 :)
I also felt the Zen 2 Laptop chips marked a huge step forward on the Win/Linux side at that time: generationally more power efficient than the best laptop chips from Intel, same core/thread count as the desktop equivalents, and cheap enough to be in a lot of budget laptops.
- zakki 3y agoOP claims still holds I guess. There are secondhand M1 in the market.
- Dalewyn 3y agoI'll also point out that quoted argument doesn't even make any sense. The vast majority of people are not going to run their computers at full tilt all day every day all year long. Most of the time is spent in idle cycles or very low usage, so efficiency gains matter a lot. If you are the few people who run their computers full tilt, it still doesn't make sense because you will finish your task sooner meaning you will spend less energy and time overall completing the same task as before.
- turdprincess 3y agoAnecdotal, but my M1 MacBook Pro doesn’t have amazing battery life. The other day I went to a coffee shop with 60% charge and only made it about 2.5 hours doing iOS development.
- sb8244 3y agoSame here. My MacBook air easily could go 6+ hours and it feels like the pro goes 2-3. Same workload.
- Reason077 3y agoI have an M1 MacBook Air and use an app called CoconutBattery to monitor power usage. What I’ve noticed is that while the laptop is usually super efficient (say, 2.5-4W while web browsing giving 8+ hours battery life), occasionally the power draw climbs up to 6, 7, 8W or more with no obvious explanation, draining the battery in 3 or 4 hours. Activity Monitor will show no apps using unusual energy yet the power is going somewhere? I’m suspicious that there’s something with maybe the WiFi hardware or the SSD which sometimes causes elevated power use, unseen by Activity Monitor? Even a reboot won’t solve the phantom power drain while it’s happening, but it does seem to go away on its own eventually…
- lathiat 3y agoWatch your backlight level, which auto adjusts to the room light level. "When comparing battery life, remember that screen brightness makes a huge difference! On an M1 MacBook Air sitting idle on the KDE desktop, you will get around 7 hours of runtime at 100% brightness, and 23 hours of runtime at the minimum (not zero) brightness. Even keyboard brightness matters! Those numbers are with the keyboard backlight turned off. If you turn it on at max, even with the screen left at minimum, your runtime goes from 23 hours down to just 14 hours. So please don't ask us what battery lifetime to expect while doing a certain task and expect a single answer!" https://social.treehouse.systems/@AsahiLinux/110548137464042052 https://social.treehouse.systems/@AsahiLinux/110548137464042...
- beebeepka 3y agoYou are correct. Even plugging in my gaming mechanical keyboard makes a noticeable difference.
- Reason077 3y agoYes, I should have mentioned I always have keyboard backlight turned off [1] I’ve suspected something to do with screen brightness in the past also. System log was often full of messages from the brightness adjustment/monitoring daemon, which raised my suspicions, but I think it might just have been normal behaviour at the time. However the mystery power draw seems to happen independent of screen backlight setting. ie: when the problem is happening, turning backlight right down will reduce power draw a bit, but it’ll still be high. On other occasions, even with the screen brightness at 100% I’ll still get “normal” power draw. I’ve experimented with turning off brightness auto-adjustment too, with no effect. [1] Keyboard backlight is very annoying at night when trying to watch a movie or something, and Apple made it difficult to toggle on and off on recent Macs by removing the hot key for it. So I’ve learned to just live without it.
- JCharante 3y agoI’ve gone 4 work days without charging my M1 MBP
- willcipriano 3y agoI've never gotten a full day
- Gigachad 3y agoI've got the M1 2020 and find it almost lasts 1 day usage.
- beebeepka 3y agoIf you live on Jupiter, with its 10-hour days, then maybe.
- beebeepka 3y agoOh, this was meant for the 4-day guy. Not sure how I responded to GigaChad
- kylec 3y agoDo you have a 13" MacBook Pro with M1 or a 14+" MacBook Pro with M1 Pro/Max? If the latter, it has decent battery life, but it's nowhere near as efficient as the regular M1 MacBook Pro. If it's the former, I guess I'm surprised, I own one (I'm writing this comment on one) and I've loved how the battery seems to just last forever on mine.
- RulerOf 3y agoI have an M1 Max and while I admit that I find myself sometimes disappointed by the battery life, it’s still often getting more than three hours while I expend precisely zero effort to ensure I’m getting the most out of my limited battery life. It’s a far cry from the 20+ hour runtimes shown during reviews of the new machines, but still good enough to cover about as long as I’d want to sit in one spot.
- antigirl 3y agoMy M1 max [14inch] was suffering with really bad battery life. Turned out it was sync plugin for VSCode.
- OrderlyTiamat 3y agoCould you expand on that? You discovered vs code takes a lot of power when settings sync is turned on or something?
- antigirl 3y agoit was a third party extension: https://twitter.com/jordanpittman/status/1526247984349978629?lang=en https://twitter.com/jordanpittman/status/1526247984349978629... https://github.com/shanalikhan/code-settings-sync/issues/1361 https://github.com/shanalikhan/code-settings-sync/issues/136...
- thefz 3y ago> Guess this was written before the M1 :) Depends on what you do, my girlfriend uses her M1 heavily for video calls and the battery lasts the same as any other average laptop.