Google to Fix Chrome Bug Killing Windows Laptop Battery

Status
Not open for further replies.

Petrovic

Level 64
Thread author
Verified
Honorary Member
Top Poster
Well-known
Apr 25, 2013
5,357
20,490
6,278
Somewhere In nowhere
Users running Google Chrome on a Windows-powered laptop most likely know that in some cases, the browser simply kills the battery, demanding a lot more power than any other browser out there.

Well, it appears that Google has finally decided to look into reports of such a bug and the company has even assigned a developer specifically for the task of diagnosing the issue and delivering a fix.

As Engadget notes in an article today, the bug was first reported in 2010 and was caused by Chrome not allowing the laptop to go in idle mode when not being used, thus keeping the processor working at higher speeds and requiring more battery power.

Forbes launched its very own investigation into the bug and has determined that Google Chrome actually modifies system clock tick rates in Windows, thus making the processor wake up more often to check for events that might require additional power. Hence, increased battery consumption.

At this point, there’s no release date for a new version comprising this fix, but expect Google to introduce it into a future update of its development branches before eventually making it to the stable versions.

Of course, Windows users clearly need this patch as soon as possible, especially since it’s four years old and working on a laptop or tablet on battery power is critical for many these days.
 
  • Like
Reactions: Ink
"Google Chrome actually modifies system clock tick rates in Windows" - How does it do that?
 
I have not tried this on Windows 8 but one reader reports that the performance slowdown is gone on Windows 8. This article in ArsTechnica discusses the move to a tick-less kernel in Windows 8. It seems that some of the cost of just having the timer enabled has gone away with the move to tick-less. Now the cost should be proportional to how frequently applications ask to be woken, which is much saner. I have not verified these changes myself but it sounds encouraging. I have not seen many technical details about the tickless Windows kernel, but a recent article about the tickless Linux kernel explains some of the issues and challenges. In particular it is quite likely that Windows 8 still runs the interrupt on one processor, so that timeGetTime will have its increased precision. The article Timer-Resolution.docx discusses timer coalescing, and disabling of timer interrupts on processors that don’t need them, which is presumably part of what was changed in Windows 8.

it is why Linux Kernel own Windows one, and Win8 own Win7.

people stick with old versions for design/usablity aspects discarding what the real deal is : kernel improvements.

i prefer an ugly Intel core I5 laptop than a shiny one with pentium 3...
 
Status
Not open for further replies.