It's not answering quite the question you asked, but Shannon - who invented much of this stuff - has some really nice practical arguments in the start of his amazing paper that introduced "information theory" [1]. It's a really readable paper, much less intimidating than you might think, and worth a look.
Hartley was (as far as I know) the first person to recognise the usefulness of log probabilities in the context of measuring information [2]. It's a really amazing paper, for several reasons ... but one thing that really strikes me about it is how it's aged: the first half has an essentially timeless presentation of the essence of information, and the second has a now-quite-irrelevant presentation of how to make a better TV. I guess that was the really interesting problem at the time!
Practically speaking, it's a simple measure of how similar two probability distributions are, minimised (with value zero) when they are the same. So it's often used as a loss term in optimisations when you want two distributions to be pushed towards being similar. Sometimes this motivated by clever reasoning about information/probability ... but often it's more just "slap a KL on it", because it tends to work.
I think it's the usual meaning of shape, it's just perhaps that we don't typically think about what that really means. [I was an author of the Imperial study mentioned in the article, and I had some long arguments with my collaborators about whether it was right to describe our measurement as the "shape of the electron", so I feel compelled to defend the point!]
If I pick something up on my desk and feel what shape it is, what I'm really doing is mapping out the interaction between the density of electrons in the object under test and those in my fingers. I might infer that the electrons of the object are distributed in a spherically symmetric way (it's round), or perhaps something more complex (all the other less symmetric shapes).
So shape in the context of the electron shape measurements means how it interacts electromagnetically. If it interacted perfectly spherically symmetrically, I think it would be reasonable to say it was round. If it interacted in a more complex way (as in, you could grab it and rotate it, because it has some non-spherically symmetric interactions) then it's not round. Interestingly, the electromagnetic interactions of an electron are extremely tightly constrained by it having only 1/2 unit of spin. You can expand any field around a point in terms of spherical harmonics, and you can show the with spin 1/2, the electron can only interact in the manner of the first two spherical harmonics - monopole and dipole. So it can be round (monopole), or round + a more negative and less negative end (dipole). Nothing more complicated than that. (Assuming you believe quantum mechanics. The Wigner-Eckart theorem is the thing to look up if you're interested.)
These measurements, then are measuring the dipolar component of the electron's electromagnetic interaction. The only way in which it could be not round.
As to the point charge thing: well, that's like your opinion man :-) Which is to say, the electron is what the electron is, and it cares not how humans decide to describe it! These experiments are fine examples of a long tradition of measuring and observing to make sure our theoretical descriptions are actually faithful to reality.
You might be tempted so say that if one of these measurements discovered that the electron was not round, then maybe that would be evidence for the electron being not a point particle. It's complicated though ... and you'd find many physicists would start arguing with you if you did say that, because the current description of the electron - while being a point particle in a certain sense - is already pretty complicated (basically, because of interactions between all of the different quantum fields) and many people would say it's already not point-like. Many wouldn't though. Maybe the point here it's quite tricky to be precise about these things without just doing it properly with maths!
Thanks for the clarification! The story I was taught (up to and including a graduate course on relativistic quantum chemistry!) is "delocalized point particle", i.e. it's a point but of course for quantum mechanical reasons it's effectively in many places at once, at least until you collapse the wave function.
Are you talking about the quantum delocalization? And if yes, presumably an isolated electron (since in an atom the shape of the quantum distribution will depend on orbitals?
Sorry if I'm completely off base here -- I'm a mathematical who took some chemistry, but without any particle physics background.
In recent years, a number of prominent computer scientists, along with academics in fields such as philosophy and physics, have lent credence to the notion that machines may one day become as large as humans. Many have further argued that machines could even come to exceed human size by a significant margin. However, there are at least seven distinct arguments that preclude this outcome. We show that it is not only implausible that machines will ever exceed human size, but in fact impossible."
I have an XPS 9500 and have found this infuriating.
I think there's two separate things going on:
As others have noted, S3 sleep isn't supported, only S1 "sleep to idle" sleep. But I don't think this is the direct cause of the overheating. In S1 sleep, the laptop can average something like 900mW of power usage, which is enough to annoyingly knock a few percent off your battery overnight, but not enough to make the laptop warm, even in a bag.
The seems to be a second problem, specific to Windows, that when in S1 sleep sometimes the power consumption is high (of order 10W), and this causes the laptop to get very hot if not well ventilated. I've never been able to figure out if this Windows actually doing something useful in "modern standby" like Windows Update, or whether it's a bug. Edit: And I should add, it's crazy that there's not a way to disable this if it is doing something "useful".
Either way, under Linux the latter doesn't happen, and the laptop sleeps very cool ... just with the annoying "lose 5% of your battery overnight" problem from sleeping in S1.
There is something OS specific: Windows has some setting to wake up when a WLAN connection is available. I had this issue once, that I hibernated it, put it in a bag and went home. On the way it must have picked up some WLAN somewhere and have turned on, while in my bag, while the lid is closed, while hibernated ... And I believe this was the standard setting. I mean, who in their right mind wants their laptop to turn on, when the lid is even closed only, because they are in range of some WLAN? What a silly setting. This has probably fried many machines and also probably their owners still do not know, that Windows was the culprit, not their hardware. Fortunately my way home was not long at that time, so the overheating was avoided in time multiple times, until I figured out what was going on. Well, now I do not use Windows any longer, except rarely, so no such issues.
Windows keeps waking up for random things, the problem is getting worse with every Windows release and Microsoft keeps removing more and more controls.
Just a few days ago I spent hours trying to fix the constant wake-ups. This time it was a waketimer set by the StartMenuExperienceHost.exe process! [1]
Microsoft already removed the Power Management tabs in Device Manager for most devices (mouse, etc.). They also removed the "Allow wake timers" option in Power Options (Surface pro has very limited power options exposed). They also removed the CSEnabled registry key.
> Windows keeps waking up for random things, the problem is getting worse with every Windows release
I resorted to removing the power plug from the PC every evening after it kept turning itself on during the night semi-randomly. It was in "suspend to disk" mode, whatever that is in more technical power-save jargon terms, not just "suspend-to-ram".
I spent a significant amount of time going through Windows event logs to find what caused the wake-ups, fixed some, others were too broad to do anything about it. The settings are useless. So is the Microsoft help forum. I won't even try to ask Dell (Dell 8500 PC), they are only good at sending replacement hardware but unable to answer anything related to software (as long as they do the former I came to accept the latter).
I too had my (also Dell) laptop overheat after I had suspended it, thinking it was turned off. It was very hot when I took it out of the bag later, fortunately it was just in time. It seems this mishap caught a lot of people off guard.
What REALLY pisses me off is that if I put my computer to sleep for the night, and windows then runs some updates, it doesn't put the computer back to sleep! Why not? Why should it run all night instead of going back to sleep, what the hell microsoft
> I resorted to removing the power plug from the PC every evening after it kept turning itself on during the night semi-randomly. It was in "suspend to disk" mode, whatever that is in more technical power-save jargon terms, not just "suspend-to-ram".
In "Suspend to RAM", the RAM is kept powered when the computer is shut down, and thus doesn't have to be touched on wake, but it means the computer has to stay powered the whole time.
In "Suspend to disk", the entire RAM is written to disk (which can take some time especially with an HDD) then the computer shuts down entirely, on wake the OS will restore RAM from disk before resuming. The need to read data back from disk to RAM makes the wake costlier, but because everything's on disk the computer can be completely shutdown.
The two can be combined into mode where data is written to disk (making going to sleep slower) then the computer enters "suspend to RAM" mode. If the computer is resumed normally it is restored from RAM, but if the computers suffers power loss it's restored from disk. Either way the computer doesn't have to boot from scratch and all the working set should be recovered. IIRC it's the default behaviour for macOS laptops, microsoft calls it "hybrid sleep".
I've seen the same behavior on a 10+y.o. system with Win11 on it. I started unplugging it every night and re-plugging it every day when I went to use it.
Funny thing has happened in the meantime, since getting a MB Air I haven't plugged it in for like 2 months now...
Mine will wake the screen and make "device connected" and "device disconnected" sounds while the computer isn't sleeping, regularly. Besides being annoying (I eventually disabled those specific sounds), it probably meaningfully hurts the lifespan of my monitors to power cycle them every five freakin' minutes.
I've never been able to figure out what exactly is happening here, trawling through event logs and such (I'm no Windows guru nor do I want to be).
At this point, in my particular case, Linux actually feels more "hardware compatible" than Windows. It can keep the monitors off.
I really wish Windows was good enough to not raise my blood with things like this, because sometimes you just need to use it.
I noticed that it wakes from sleep for windows updates now, which not only is annoying (Bluetooth devices suddenly waking up too), but windows update often requires intense CPU usage as defender does its dirty things, .net re-JITs, etc.
In the US at least, laptops with LI-ion batteries are not supposed to be in the cargo hold.
I don't know whether the checked-luggage scanners catch this or what happens if they find a laptop in luggage.
The passenger isn't present during checked luggage scanning so it would be complicated to try to give the laptop to the passenger. The obvious alternatives look like theft and are extremely inconvenient. (Oh joy, I'm at my destination and my laptop isn't.)
Same in Europe. No idea what they do either. I've only put a laptop in my checked luggage once, in 2000 when it was still permitted. It was one with NiMH cells by the way which don't have a tendency to catch fire but they suck in other ways (energy density, memory effect). That's why nobody uses them anymore.
It arrived with a cracked screen so never again...
My previous PC would wake up every night around 1am and I could never figure out why. I disabled every single wake timer, all kinds of wake permissions on devices, yet it would always wake up in the middle of the night. The power event in pc management only said "woken by: unknown source".
I do some support work for a friend and his kid's school laptops.
All Lenovo/Win 10.
He complained about the same thing, in this case on laptops that were completely shutdown the night before. In the end we tracked it down to the Lenovo Vantage service.
I assume it was powering on the laptop to check for updates but I could find no log or record of it doing so. But, once we removed that software the issue went away completely.
Anecdotal I know and I even told him it could be something else, that removing the software may have changed something related but not from Lenovo, etc. But in the end, a few weeks later that is, he confirmed that since we did that, they did not have the problem again.
For desktops at least, Wake on LAN is still a setting in EFI/BIOS. On my AMD desktop using a high end ASUS motherboard (Dark Hero), I switched that off along with a couple things that looked like a WoL setting in Windows, and that PC has stayed asleep all night ever since.
I haven't seen this behavior from my ThinkPad X1 Nano, but that may be because it shuts itself down entirely after being closed for an hour or two without being connected to a power source.
You can disable network connected standby with group policy. I did try that when running Windows on my machine, and it didn't improve the randomly-getting-hot-when-asleep for me.
Generally speaking - GP GUI is just a front end for making changes to the registry for local group policies. You can find the mappings and make the registry setting changes yourself.
Hardware manufacturers could pressure Microsoft to stop doing this. I hope there will soon be legal action prompting the manufacturers to take action (perhaps a class action against Dell by people who have been denied warranty on this basis? - though in the USA at least, Dell probably has that blocked by arbitration clauses.)
I have an XPS 9500 under Linux (this is a >$4000 machine), no amount of fiddling with BIOS settings (which are few) or systemd/GNOME settings will make it work reliably:
- short suspend time (it can't really stay overnight when suspended with a full battery)
- sometimes it wakes up and overheats in the bag
- very short battery duration when unplugged
Apparently I'm not alone, and even under Windows it's a generalized problem; and when you look for solution on Dell's forum you find this FAQ, which tend to say XPS are more foldable desktops than real laptops.
I have an high-end XPS 13 from 2020 (9310), allegedly with native Linux support, and have the exact same issues.
My previous XPS 13 from 2016 was suspending properly but now I cannot suspend my "laptop" for more than a day (it will die) or store it in a bag (it will become lava). Hibernate does not work either.
I'm learning with this thread that it also happens on Windows and I'm struggling to understand how Dell could decide to sell "laptops" at this price and not test one of the most basic features of a laptop.
I'm still rocking mine from 2016. It's been a stellar little linux laptop. Hopefully this gets sorted before I have to upgrade but I'm really good at stretching hardware on linux.
I've tried that and while the laptop sleeps, I wasn't able to get it to wake up again. Which of course undermines the utility of it sleeping quite a bit :-) I didn't persist too much with it, so it's quite possible that one can make it work ... but from what I read when I was trying it, it might well be the case that BIOS support for S3 sleep just isn't there.
My wife had a similar problem with her XPS-15. Randomly during the night the laptop would warm up and cause the fans to spin up. Apparently this is caused by windows waking up. Initially I thought it was windows deciding to do an update, but even disabling that (if it is possible at all), the laptop would wakeup. Random trailing through the internet pointed to the useless Killer network drivers, or better the management application, to be the cause and uninstalling them was supposed to fix the issue. Anecdotally it seemed to have worked, but when I last checked the drivers seemed to be back.
I have the same problem on my new Thinkpad. I found a setting in the BIOS for Linux mode and that seems to have fixed the problem for now Windows has S3 available again. If I switch it to Windows mode, then modern sleep prevails and the fans run all the time.
> just with the annoying "lose 5% of your battery overnight" problem from sleeping in S1.
Just 5% would be manageable -- I probably wouldn't even notice it. It's more on the order or 20-30% for me, so if I happen to not use my laptop over the weekend, I'll come back to a completely dead laptop.
A couple times, I've opened the laptop to have it scream at me to plug it in ASAP.
It's frustrating when you put a completely charged laptop to bed and can't get going without having to hunt for a power cord.
Agreed. Every time I tried windows it was the same set of frustrating behaviors. Linux helps but needs some tinkering. My first step on a new laptop would be to wipe clean and install Linux.
I run Ubuntu on a X1 Extreme Gen 2 for a while now, no problems whatsoever. The discrete nVidia GPU is as power hungry, and running as hot during gaming, as it did under Windows (so nether a big surprise nor a real problem, anyone remember how harsh Battletech was on GPUs initially?). Battery live, depending on use (and excluding gaming) is around 4 hours (didn't have less then 3+).
The heat issue is mitigated somewhat by putting the machine on some kind of stand to have air pass underneath it. That being said, even under Windows the machine went to sleep just fine, also woke up again. If anything, that works better now under Linux.
There is a difference between T440 and T440s. T440s is already power optimized and runs integrated GPU vs T440 which is more normal laptop with discrete graphics.
I used mostly powertop but this is less reliable source of information when you are running discrete graphics card.
As an i7, is it quad core? I had a T440p with i7-4700MQ and it had terrible power consumption compared to prior and later Thinkpads I have experienced.
What I saw in powertop was that it never got into the better "package" power states. The cores could all be spending 90%+ of the time in C7 but the package as a whole was in C2 or worse. I never found a way to fix this.
Mine had the NVIDIA GPU as well as Intel iGPU, and I primarily used it with the iGPU. The NVIDIA was of the sort that was not connected to a physical output, so it supposedly could be powered down when not being used. I had no way to truly verify this, of course.
On my T440p, the Intel and NVIDIA GPUs would show up via 'lspci'. I don't recall whether there were any BIOS options other than internal vs external display during POST/boot.
I hate it too. I’m not just saying that, I ran Linux as my only desktop for 3 years and exclusively AOSP phones for over 4.
Ultimately, however, some of us have stuff to do and can’t spend our entire time tinkering to maintain basic functionality. There are only so many hours in a day.
"Modern Standby" is such a crock of shit. As far as anyone knows, it doesn't do anything useful, but it consumes FIFTY PERCENT (50%!) of the battery each night.
Luckily you can turn it off, at least in Linux, and then the machine functions more-or-less normally.
(The above requires the package 'sysfsutils' to be installed on Debian / Debian derived distributions).
On my work supplied xps-13, bluetooth does not survive the laptop being unplugged in "deep" sleep (even for a couple seconds). Fixing bluetooth requires suspend to disk aka "hibernate" or regular reboot to restore. No reloading modules etc. helps. Other than that annoying bug, proper sleep works fine on the hardware.
But, I also had to add,
/etc/modprobe.d/i915gpu-fix-xps13-crashes.conf:
To solve the laptop crashing when idle. Originally I disabled c-states on the gpu to fix the crashing, but some other kind soul on the Internet shared the above which solves the crashing, but doesn't kill battery life like my fix.
Looking to update an older XPS. Seriously considering a framework, but might just buy a desktop considering I spend 90% of real dev work in the same spot with a 4K.
Serious question — why do people so frequently sleep rather than shut down? Is it to save time on booting up? Is it because you want other processes running? All of the pain points below are reasons why I just always shut down. I think I developed the habit at a time when sleep just never worked on my linux machine though, but that was years ago now. I’m speechless reading below that Windows wakes just to install updates… and then remains awake!
> Serious question — why do people so frequently sleep rather than shut down?
So we can right back in the context we had? Who wants to spend the first 15 minutes digging out Jira, the issue you worked on, the two related ones, start IntelliJ, digging out the correct screenshots etc etc.
... or even: who wants to stop debugging and turn of the computer just because we are leaving the office to catch the train?
I first learned this on a 486/66 DX2 IBM aptiva desktop in 1995.
I've too had times when it didn't work on Linux but today boil-in-bag seems to be a Windows feature, not supported out of the box on ordinary Linux distros on mainstream hardware ;-)
I prefer to put my computer to sleep rather than shut it down. Closing the lid on my computer, putting it in my backpack, and heading home for the day is very natural. When I get back to work the next day, I open it up connect the cables, scan my fingerprint, and everything's back where I left it. I usually go at least a month between reboots, but that's highly dependant on when updates come out (and get approved by work, of course).
This is on a Mac though, so (for me anyway) this is something that just works without any fuss.
As a slight aside, I prefer to shut down and be sure that things are shut down. It really bothered me when Mac laptops started turning on just because the lid was open.
It makes it impossible to confirm the laptop is truly off, because opening it to check turns it on. This was especially annoying during my "boarding a plane" ritual where I check everything is right before settling in.
And on a Mac it works either way, as you will get almost all of your context back even if you shut down rather than sleep. Windows still seems to start up as a blank slate, unless I’m missing a setting somewhere.
These days, Windows will try to recover your context, but the implementation (like many other modern features like display scaling) is dependent on the app. As far as I've been able to tell, roughly most Microsoft and Electron apps will recover their states on reboot; most other apps won't.
Like a sibling commenter, I've found this pretty hit-and-miss too.
Something like Outlook or OneNote will restore its state reasonably well, as will Safari/Chrome/Edge/Firefox. Others, like Activity Monitor, Enpass, and iTerm2, decide to "helpfully" open a window for me even though the previous state was "running with no windows open" (which is perfectly valid for many applications).
People who use Macs have been just closing and opening their computers for about a decade and not thinking about it.
The wifi is still connected, the ssh session I was in is still connected... The tests I was running when the doorbell rang... they pick up where they left off. It's nice.
It's stuff like this that keeps me begrudgingly coming back to Apple for laptops.
> People who use Macs have been just closing and opening their computers for about a decade and not thinking about it.
Yes and no. There have been weird issues where the machine wakes but the screen stays dark and another where externally connected monitors won’t necessarily work on wake. Possibly some of this is connected fo the hell which is the dongle lifestyle.
Apple seems to have come out the other side of these issues with recent laptops/OS releases (last 18 months).
Yes. I agree. It seems to be getting harder to set up hibernate as an option. Regular (non HN) users might not even figure out that hibernate is a possibility. But it definitely is a better option than sleep.
> But it definitely is a better option than sleep.
It's the better option if you don't mind the longer time waking up, since RAM has to be restored from disk.
Otherwise, sleep should be backed by hibernation on all machines (it certainly is on mac laptops, and I think it is on windows as well): in case of power loss during sleep, the machine falls back to waking from hibernation, but if there was no power loss it wakes way faster. This is especially useful for people who move around a lot during the day and will close the laptop, move around, and reopen it. While things have gotten better with modern SSDs, having to restore 16 to 32GB from disk to RAM is far from instant-on.
Macs do this. They hibernate to RAM and after about 1 hour (configurable with pmset I think) they wake up momentarily to hibernate to SSD.
The reason they don't write the hibernate image straight away and just power down after an hour is to eliminate writes to the SSD but I believe you can set that wait to 0.
Sure, but it definitely avoids a bunch of the problems mentioned in this thread. It's annoying that you have to dig so hard to even make Hibernate visible next to Sleep, Shutdown, and Reboot on the various power menus.
I recently switched from an XPS 15 to an M1 MacBook Pro, and it's glorious that I just don't have to think about it anymore. The XPS 15 had all the problems I'm reading above, and then some.
I don’t choose to sleep or shutdown. I want instant context saving and I will close my laptop many times a day. I expect opening it to restore me to where I was within 5 seconds with same application state and windows in the same place.
I have a MacBook which performs this flawlessly and my Linux desktop also pauses and restores flawlessly except immediately after installing new Nvidia GPU drivers.
I haven’t thought about different CPU sleep stages in approximately 9 years when I had to make my XPS M1330 handle them in the end years of its life.
> I think I developed the habit at a time when sleep just never worked on my linux machine though
I did the same in my ~15 years of Windows & Linux desktops and laptops, with the sole exception of an IBM (yep, was still IBM) Thinkpad on which suspend to disk would magically Just Work on Linux if you created a partition at the correct location, with the correct size, and with the correct type ID. The BIOS handled it somehow, I think, which seems crazy but it did work flawlessly. IIRC I didn't even have to tell Linux about it, and I ran Gentoo at the time so I doubt it was doing anything for me automatically.
Switching to Mac a little over a decade ago broke me of the habit, eventually. It was one of many coping behaviors I didn't need anymore and had to un-learn. Unfortunately, now that I'm used to shit actually working semi-correctly a fair amount of the time (to be clear, Mac is far from perfect, everything else is just so much worse that it's like no-one else is even trying) without my having to spend time forcing it to work, it's hard to go back.
Primary reason: I need to type a 16-character bitclocker password and then a 16-character Windows password after shutdown. But with all those sleep and overheating issues that's what I do now.
I've seen issues with Modern Standby on my fleet of Lenovo ThinkPad Yoga L13 and L13 Gen 2. Usually everything is fine, but sometimes they get super hot in standby as well. It's like something is keeping the CPU awake but the rest of the hardware is asleep (including the fans). My older units (ThinkPad Yoga 260, ThinkPad Yoga 370, ThinkPad L380 Yoga, ThinkPad L390 Yoga) never had any problem like this because they don't support Modern Standby.
It's very annoying that MS doesn't allow us any way to disable Modern Standby. OEMs still haven't figured out how to make old school sleep perfectly reliable. Springing a new standby model on them was doomed to be just troublesome.
Modern Standby is an Intel thing, not a Microsoft thing. Intel has just pressured Microsoft and other OEMs into supporting it. IMO, it's a huge steaming pile-o-crap, and one of the biggest reasons I want my next PC to have a non-Intel CPU. Intel has shown over the past decade that they are quite simply incapable of implementing properly functioning power management, and I'm tired of having machines die because of their stupidity.
I've had two Surface Pro 4's (one of the first "Modern Sleep" devices) develop battery bloat because of this Intel's power mgt incompetence. Microsoft replaced both, but what is this crap costing all of us, both in higher hardware prices and environmental waste?
FWIW, I haven't needed a faster CPU in years - I need more RAM, long, long battery life, and sleep/wake that always works, instantly. If the iPad were capable of being a real computer, it might get me back into the Apple fold, if iOS had a usable UI...
FWIW, it is not an Intel thing. I got an Asus G14 2021 which has nothing Intel (Amd CPU and nvidia GPU) and this does not support S3, only the modern "connected standby" crap.
Talk about ruining a perfectly working solution for almost no gain.
Not sure if this is related, but I have a 2021 Legion 5 Pro, and even fully powered down it seems to lose a lot of battery power overnight. (Maybe 5%-10% charge, IIRC?)
And this is even after changing the BIOS setting so that it's always-on USB port isn't always on.
I'm really curious where the power is going. Or if the supplied battery has internal leakage issues.
It's something about modern standby. I have a laptop with a 10th Gen i7. I put it in my bag once and when I pulled it out after my trip, the fans were running at full blast and the laptop was extremely hot. I'm also fairly sure it damaged the fan because it's never been able to run at higher speeds ever since.
I'm also not sure if it's a bug or something else. I do feel like there's something problematic about modern standby. I didn't have this issue under Linux, which actually does standby properly.
You can force modern standby to disconnect from networks (connected standby vs. disconnected standby). This helps with stopping the system from waking up randomly.
Something seems a bit off with your numbers. If you're losing 5% of your battery "overnight" (at least 8 hours I assume), and you have the larger 86Wh battery, that implies a discharge rate of only 537mW.
900mw - nearly a whole watt - seems very high! Easily enough to cause noticeable warming - a running modern laptop at idle only uses 3-6 watts.
>I've never been able to figure out if this Windows actually doing something useful in "modern standby" like Windows Update, or whether it's a bug.
I asked a very similar question a few months ago. The sleep/standby modes were behaving as other people have reported, however, the fact that the battery was draining rapidly in shutdown mode, on a new Dell with a 3 month-old battery with very few cycles, was a cause for concern. Nevertheless, I tried using the CsEnabled trick, which didn't work. Eventually, after troubleshooting BIOS features, applying the latest updates and using powercfg options -- batteryreport, sleepstudy etc.
I found a solution, which is a compromise at best. The battery drain was 20% within a few hours. But now with S3 sleep state -- it drains around 8-10% in full shutdown mode over a period of 8-10 hours. A similar drain in Standby is over a period of 15-18 hours.
I was going to post the exact same thing. I have the 9500; this definitely infuriated me and is mostly solved by installing Linux.
There are still some things that can annoyingly wake it (usually a bluetooth device trying to pair to it); but it's an odd exception versus the 50% chance every time I went to transport my laptop with Windows.
I've just got the previous version and the coil whine is really annoying. I'm debating whether to keep it. I actually wouldn't mind as much if it's constant, but hearing a "brbrbrbrbrbr" whenever I scroll is very irritating.
There's a lot of static on the headphone port too, which is disappointing.
I had no idea this was actually a thing before this morning :(
It seems to only happen when there is graphics stress -- using "semaphores=1 i915_enable_rc6=7 i915_enable_fbc=1" seems to help, and once I started playing videos with --hwdec it was quieter also..
As it only happens during video, and I usually have headphones or the speakers going at those times I'm not too bugged by it yet..
Doing a kernel compile or something noisy in a term doesn't seem to trigger it..
This also highly influenced my decision to buy a Lenovo T460 which is silent (even no noise under load) and long lasting too and Linux runs well on it too and is a lot cheaper for the same hardware, but does not look that pretty ;)
Hartley was (as far as I know) the first person to recognise the usefulness of log probabilities in the context of measuring information [2]. It's a really amazing paper, for several reasons ... but one thing that really strikes me about it is how it's aged: the first half has an essentially timeless presentation of the essence of information, and the second has a now-quite-irrelevant presentation of how to make a better TV. I guess that was the really interesting problem at the time!
[1] https://people.math.harvard.edu/~ctm/home/text/others/shanno...
[2] http://dotrose.com/etext/90_Miscellaneous/transmission_of_in...