Microsoft Teases Multi-Day Battery Life For Upcoming ARM-Powered Windows Devices (techspot.com) 72
An anonymous reader quotes a report from TechSpot: Microsoft late last year announced a partnership with Qualcomm to bring the full Windows 10 experience to ARM-powered devices. Terry Myerson, Executive Vice President of Microsoft's Windows and Devices Group, promised at the time that Snapdragon-powered Windows 10 devices would be efficient in the power consumption department. We're still waiting for the partnership to bear fruit but in the interim, new details regarding efficiency (and a few other subjects) have emerged. With regard to battery life, Pete Bernard, Principal Group Program Manager for Connectivity Partners at Microsoft, said that to be frank, battery life at this point is beyond their expectations: ""We set a high bar for [our developers], and we're now beyond that. It's the kind of battery life where I use it on a daily basis. I don't take my charger with me. I may charge it every couple of days or so. It's that kind of battery life."
Device? Since it won't be a phone ... (Score:2, Funny)
...it must be a dildo.
Can't wait for the first complaints: "Oh, no! Not rebooting _now_ for an update!"
Re: (Score:3, Funny)
Ah, but this is Microsoft...the only product they could build that wouldn’t suck is a sexbot.
Re: (Score:2)
Ah, but this is Microsoft...the only product they could build that wouldn’t suck is a sexbot.
didn't they have to turn that bot off?
Charge every few days (Score:5, Funny)
Charge it every few days because it has so few applications that you never use it?
Re: (Score:2)
Charge it every few days because it has so few applications that you never use it?
Exactly!
It’s multi-day battery life as long as it&rs (Score:2)
If my Surface RT is any indication of Windows and ARM, after a year and tons of patches it will last maybe 4 hours of normal usage.
Re:It’s multi-day battery life as long as it (Score:4, Interesting)
Also, who are they going to sell these to? The people who bought Surface RT won't want to be fooled again (presumably) and everyone else already has an iPad or some sort of Android tablet, or maybe a Microsoft Surface device which runs x86 and so has access to whatever Windows software the user might need.
This looks like all the other Microsoft "new market" type efforts. Doomed to failure and then irrelevance, then death.
Re: (Score:2)
It runs win 32 apps, RT didn't.
They will sell them to the 90% of people who do light to moderate tasks and want a 2 day battery.
Re:It’s multi-day battery life as long as it (Score:5, Interesting)
I highly doubt it will get days of battery life in any realistic scenario.
There is no magical way to make common tasks 2-3x more efficient. There are no magic RAM chips that draw 25% normal power, no special GPUs that decode video using half as much energy as every other one. Did someone invent a super efficient LED backlight?
And people expect their Windows devices to stay connected to networks and more active than a phone, so that Skype calls can come in etc. Android and iOS sleep for to to 15 minutes, with no IP connectivity and delayed notifications, to save power.
Re: (Score:2)
You could just put a realistically sized battery in there. Power draw minimization is one thing, power capacity is another.
Re: (Score:1)
I disagree. I've used both Windows Phone and Android, I can say without doubt I got more mileage on the battery of my old Samsung Focus Windows Phone than I did my present Samsung Galaxy A3 Android. On Windows Phone I always had +40% battery at the end of a day on a single charge, whereas my Android phone gets charged twice a day. So if they've made big improvements since v7, on the right device, it could easily be true.
I think the biggest irony for Microsoft is that Windows Phone is both the finest OS
Re: (Score:2)
If they meant Nokia's Windows Phone devices, they have a fair point.
Re: (Score:2)
I think the critical failure for the Windows Phone was the app store. If they'd have baked some way to run Android apps on it that would have dramatically changed the mobile world.
Unfortunately the company culture wasn't right to do that then, and now the opportunity has passed. Today's Mcrosoft would do it in a heartbeat.
Re: (Score:2)
Today's Microsoft killed the Astoria Android compatibility project: https://arstechnica.com/inform... [arstechnica.com]
Re: (Score:2)
I think the biggest irony for Microsoft is that Windows Phone is both the finest OS they've ever produced
Depends on how you define "finest", but for me that would be user experience (aka UX.) I think Windows 7 was therefore the finest OS they produced. Ranked as follows: 7 > 10 > Vista > XP > 8 > Phone > Mobile. (Yes, Vista had some technical issues when it first launched, but I'm only going based on the latest version, which had most of that ironed out. Windows 7, which seems to be the most popular, is basically Vista SP3.)
If you define it as features, then it would go 10 > 8 > 7 >
Re: (Score:2)
My current phone gets 2+ days of charge, while checking email in the background, playing music during the day, and being used for regular messaging and web browsing. The magic lies in it not being optimized for thin! over having decent battery life.
Ironically, it's also a Windows Phone, which was a surprising decent OS, despite having poor app availability and being abandoned by Microsoft. I've never been a fan of MS stuff, but wanted to try it out after having iPhones and Android phones. I'm definitely g
Re: (Score:2)
Can you tell us what modem, or what size the battery is? On a Pixel XL with 3450mAh battery and Android I get two solid days with similar use to you, maybe a bit more browsing.
The secret to this is Doze, where the phone can sleep for up to 15 minutes at a time. The only problem is that any messages which arrive during that time are not received, things like wifi and cellular data connections are powered off. Calls still get through. It's okay for messaging apps and email, but for example if someone calls yo
Re: (Score:2)
It's a Lumia 640 with a 2500 mAh battery. I don't use Skype, but calls and SMS/MMS come through immediately. I have sound notifications turned off for everything but calls and texts, so I'm sure the email is just being polled at intervals. I'm not a fan of enormous phones, either, so I'm sure the measly 5" screen cuts the power use down.
I only even considered this phone because my wife had a Lumia 635 and would regularly get a week off of a charge. She would lose her charger fairly often because she so rare
Re: (Score:2)
Yeah, calls and SMS/MMS will come through immediately. It's the data connection that really hogs the battery, so anything that needs data to notify like Skype or WhatsApp etc. will be delayed due to sleep.
Re: (Score:2)
The screen seems to hog the battery more than just the data connection. Unless my phone is particularly good at tearing down the connection and rebuilding it.
(Anecdote follows: For example, the last time I charged my phone was the night before last and I listened to Pandora all day at work yesterday, but never did much texting/email/screen-on activity. My phone sat in my pants pocket last night and when I checked it this morning, it was at 70%. I don't use my work's wifi for my personal phone, so all of the
Re: (Score:1)
https://developer.microsoft.com/en-us/windows/iot
The Win32 API seems to be the thing that's been missing. I am wondering if the "desktop" will need ARM binaries or if there will be some kind of translation from x86.
Re: It’s multi-day battery life as long as i (Score:1)
Re: (Score:2)
I've never seen an ARM set up to run big-endian, even though it is capable. I far prefer big-endian, but apparently the little-endian heresy is now the dominant religion.
Re: (Score:2)
I used to manage a Big-Endian Gentoo distribution for the Linksys NSLU.
I built binary packages for it on bigger hardware since 32MB wasn't enough. Though the 128MB FatSlug did decently. As decently as a 166mhz ARM 5 could.
-J
Re: (Score:2)
I think a great many products just stick with the default unless someone has a preference. The default suppplied with an eval board, the default that a chip has if you haven't touched the boot time pullups, the default that your board bring up contractors decided to use, etc. And that default is based upon people being used to Intel.
I also think that the Intel and Windows dominance has trained a generation of programmers that portability isn't important, and thus there's rarely any thought paid to byte ord
Re: (Score:2)
well if they drop the store only part and let people compile for ARM and x86
Re: (Score:2)
They will sell them to the 90% of people who do light to moderate tasks and want a 2 day battery.
No they won't, those people already own an iPad or an Android device, or they remember the last time Microsoft tried this and then abandoned their devices.
Re: (Score:2)
Leaving aside the battery issue, Win32 is an API that was designed for x86. It's used by programs that run on x86. It's run on computers that use...ARM?
The main advantage of running Windows is that you can run the same crap as your friends. Lots of this Windows crap is native code in x86, and an ARM is not going to do an energy-efficient job of emulating one. This is probably the biggest reason why RT crashed and burned so hard: it was advertised as Windows, but it didn't run Windows software.
And
Re: (Score:2)
"ARM is not going to do an energy-efficient job of emulating one."
How do you know?
Re: (Score:2)
I think even video encoding would require less processing power than Java or VMS. It's normal that you don't get the 10 hours that Apple gets in normal usage benchmarks.
Re:It’s multi-day battery life as long as it (Score:2, Insightful)
I find it bizarre and a bit infuriating when a charge that lasts multiple days is considered a significant advance. We used to have phones that would go a week without a charge, longer if you didn't make a lot of calls, but these days you're expected to charge the device constantly even if you don't use it. A multi day charge should be considered the bare minimum of acceptability.
Re: (Score:1)
The Fine Print... (Score:2)
As long as it's turned off, a battery with a full charge will last for at least a couple of days.
Wow! (Score:1)
Battery life and daily usage .. (Score:1)
Meaningless marketing twaddle (Score:2)
They didn't specify the specs of the device that runs multiple days on a charge, so the statement is utterly meaningless. I have numerous computing devices that can go days -- even weeks between charges. But none of them are phones, tablets, or laptops.
Rivals the power of a 486! (Score:3, Funny)
Run Windows 95 for days on end! 49.7 days to be exact!
Good God (Score:3)
Give it up already. It aint gonna happen. Windows NT has been ported to many platforms and they already tried in modern computing history with WindowsPhone and WindowsRT.
Unless maybe they plan some sort of weird hybrid device where the OS runs on the ARM but an ATOM (discontinued) or some x86 takes over to run classic apps I see more money lost.
Even Google played with x86 hardware with Android for x86 with the Asus Zenfone. It failed. Applications/Programs define what hardware/software to run. ARM is stuck on mobile, x86 for content creators and IBM stuff for mainframes.That is just the way it is.
Re: (Score:2)
1) ARM is not and never was big-endian. It was historically little-endian, but is now switchable (though it still defaults to little).
2) This is not an insurmountable issue. Rosetta did dynamic translation of binaries built for a big-endian architecture (PowerPC) to little-endian (x86).
3) Because this is such a common operation, it is highly optimized in hardware. ARM in particular not only has an operation (rev) to do it, you can specify data access operations in either byte order.
4) Network traffic is
Re: (Score:3)
It's the way it is because through the late 90's and early part of the 21st century Intel was able to to deploy a fab advantage over the competition, which was further compounded when AMD came up with a 64bit version of the x86 instruction set that allowed x86 to finish eating the workstation and server market.
The thing is that x86 is a dog of an architecture and it was only able to win through the deployment of capital to create a fab advantage. That fab advantage has in the last couple of years evaporated
Re: (Score:2)
Alpha refutes this. I am a slashdot old timer and remember the days of Alpha being the uber hip cool thing to run Windows and Linux on. Slashdot ran on an Alpha running Debian in a dorm room if I recall.
They were a little more than a PC but a beautiful workstation that supported x86 emulation for Windows NT/2000 RC 4.
PowerPC kicked the pentium ass back in 1994! PowerPC was a serious risk for Intel too. RISC could run twice as fast for the same price or run about the same for 1/2 the price due to less progra
Re: (Score:2)
Considering you can throw a 2GHz 8 core ARM into a mid range tablet it looks like we have finally reached the point where x86 emulation is viable. It doesn't have to match native performance, apps where it matters will be compiled for ARM anyway, it just have to offer compatibility with all those less popular but essential to the user apps that haven't been updated since 2003 and ran well on a 700MHz Pentium 3.
Re: (Score:2)
So could the older Alpha chips running Windows NT/2000 RC 3. But they ran so much slower.
Running emulators will suck battery life right out as the instructions can't be run in a way to conserve power usage. These devices are only good to run the internet on Edge and that is about it perhaps running Netflix and Hulu. The appstore is still limited on Windows
Runs win 32. (Score:2)
For many of you that have posted, WoA will run win 32 apps, this isn't RT.
So many "nerds" here ignorant of that fact.
Re: (Score:1)
Neat.
Intel actually did the same thing in reverse to allow Arm only NDK binaries to run pretty well on x86 Android devices. A lot of games were C/OpenGL applications ported from iOS that ended up being compiled to ARM binaries, not Java and libhoundini got pretty decent performance translating them to x86.
http://android-x86.sceners.org... [sceners.org]
In fact Dec had a translation layer which translated x86 NT binaries to Alpha native binaries. That actually run the code in emulation until it had profiled it and found al
I already get 2 days of battery life (Score:2)
Re: (Score:1)
Apple have an advantage here because they make both the software and the hardware and can tune both to improve performance.
E.g. in one of Louis Rossman's videos he found out that the trackpad in a Macbook supports both SPI and USB. It runs in USB mode when you run Windows via Bootcamp or are in the EFI shell but SPI when you run macOS.
https://www.youtube.com/watch?... [youtube.com]
SPI is obviously a lower power bus, but of course it's easier to add support for HID over SPI to an OS you control than it to one you don't.
So
Didn't they used to do this? (Score:2)