Cool.
That means a non-bloatware OS should be able to run for a couple weeks. Count me in!
Microsoft says testing of Qualcomm-powered laptops running Windows 10 is well advanced and suggest it will be possible to use the devices for multiple days without charging their batteries. Qualcomm revealed its tie-up with Microsoft in December 2016, when it promised Windows-10-ready 64-bit ARMv8-compatible Snapdragon system- …
Well no, the story was IE outperformed Firefox in terms of battery life. This because M$ continued to run core elements of IE in the background sucking up battery life, whilst Firefox was running. How about boot ups and Windows being quicker, nope, another lie, delayed start of many elements meant whilst you could look at your pretty desktop you could not actually do anything until the delayed start program had finished delaying your ability to actually run programs.
It is near impossible to compare OSs without knowing what programs were still running or not, when you want to fudge numbers. Think disabling all those M$ spyware programs, I'll bet they were not running during battery tests, 'er', corporate version, not the mug punter anal probe version. M$ lies and it lies a lot.
I may be interested, though I just got a 8+h battery for my old AMD E series typewriter. That used to serve me well during travel, but recent changes in Firefox are killing it. The split between web content and the main browser as well as the additional multi-threading make anything that does not have at least 4 cores (regardless of how slow are they individually) a complete and utter dog's breakfast.
So, yeah, 8h+ of battery life on arm is interesting. Provided I can install a different OS and/or dual boot - I actually need to do real work on it sometimes:) Install means install - without 2 days of rebooting and repenting like Linux on Samsung Chromebooks. I do not even try to fight with my Arm Chromebook any more. Last time I had to spend a whole afternoon to upgrade to Jessie on it (it is installed on the internal flash, something which Debian docs claim as not feasible).
You can install any extension not marked as compatible with threading and Firefox will disable threading.
Or you can run Firefox ESR for the duration and wait for the tools to catch up.
Pages similar to mine, but not mine. :)
Chances of this having an unlocked bootloader... pretty close to zero.
I'd say that most consumers want a machine that just works out of the box and gives them access to their familiar applications from a certain well known vendor, and don't really care that much what is making them work. Because of this, it really wouldn't surprise me if devices like this became mainstream so long as MS can avoid their usual standard of advertising campaign.
I'd be queuing up to buy one if I could install linux on it easily, but I'm beginning to suspect that ARM based laptops are about to become mainstream and we're heading to a future where linux can only be installed on a few niche variants of laptop. (e.g. hacked chromebooks.)
I leave home. Sometimes to work on computers for money, even. And yet, I honestly haven't noticed a Redmond or Cupertino based laptop in over six months. Might have something to do with not accepting contracts for Redmond and Cupertino systems anymore ... but even then, their products were fairly common up until the first part of this year. Now they are conspicuous by their absence.
Note that I am a sample of one. Your mileage may vary, etc.
If it runs OS X I'm interested!
I am pretty sure Apple has a similar baby in the works.
Apple were rumoured to have ARM based Macbook Air's several years ago. I suspect it's also not a coincidence that app store submissions are now in LLVM bytecode format; which, once a suitably large body of apps are in this format, should allow Apple to seamlessly ship ARM versions.
I would love a laptop/desktop based on an Apple ARM core. That thing would fly....
Is this going to be another Windows for ARM - which can only run 1 percent of available software, is not-quite-really a full Windows, but a restricted and crippled faint resemblance of the real thing, and will be quietly dropped after 2 years of pumping marketing money into it and people wasting their cash on buying devices? As much as I'd like a multi-day battery laptop, for good or for worse, I have a house full of x86/x64 boxes going back 20 years which are still compatible with most current software and can be put to good use one way or another.
Apparently it'll run x86 applications through an emulator
Not unless Intel blesses[1] the project it won't. They've already made it plain that anyone who trespasses against their emulation patents is gonna get dumped on from on high..
[1] And/or gets paid vast wedges of cash. Which means that the end-user hardware will cost a fortune becuase Intel will price their compliance at a level where it'll be cheaper to buy an Intel processor instead and forget all that silly ARM nonsense. And very expensive end-user hardware (outside of one or two brands) isn't going to sell. Remember Transmeta? I could see Apple pulling it off, but not Microsoft.
"Not unless Intel blesses[1] the project it won't. They've already made it plain that anyone who trespasses against their emulation patents is gonna get dumped on from on high.."
If that's so, why haven't they attacked the DOSBOX project or MAME or other programs that emulate x86 CPUs?
If Intel really has pertinent patents, perhaps they'd be bold enough to cite them...
Ah yes, idle.
This Slackware 14.2 machine (assuming I haven't told it to recompile the kernel or similar in the wee hours) will report a load average of "0.00 0.00 0.00" when I wake up in the morning. What would a Windows 10 machine report? How about Debian Jesse?
This machine will STILL report all zeros every morning if I leave it running for the next six months, using it normally day-to-day. Again, what would Windows and Debian report?
IOW, define "idle".
I'd just like a laptop with a super screen like this, but running a lower powered CPU. Most of the time, I just need to connect to a server. Laptop work is limited to browsing and email.
Chrome laptop would be fine if it could just run Windows app (that's not going to happen!)
This post has been deleted by its author
The AMD Exx chips were not that bad. I have two laptops with those - one with the 1.3GHz and one with the 1GHz version. They were tolerable for their ship date. With Linux of course.
They are not really tolerable any more because Firefox separation of web content into a separate sandbox pretty much renders anything sub-1.7GHz and/or sub-4 core unusable.
"They were tolerable for their ship date. With Linux of course."
That's my point. I saw these E1 chips and especially the 1GHz version shipped with Windows 8/8.1 on them and they crawled.
They ran at 100% CPU utilization 100% of the time. I hadn't seen such woeful performance since running XP on a single core with 256MB of ram. This was even with totally clean rebuilds.
I had to install SSD drives in the machines in order to get any semblance of performance and then it was painful,
I remember the customer dumped the machines about a year after purchase. They kept the SSDs though.
Worst mainstream CPUs shipped in the past 10 years IMO. AMD should have been ashamed.
Microsoft is giving up on smartphones, but loves the idea of ARM-powered laptops running Windows 10. Microsoft and Intel had a major falling out over PC processor prices, so Redmond turned to a CPU supplier dying to break out of mobile: Qualcomm.
So forget Microsoft in your pocket. Microsoft wants to be on your lap or desktop, with Windows 10, on Qualcomm ARM, or Intel or AMD x86.
C.
So this version of Windows 10 is a development of the one in my pocket, but with an x86 emulator, so that I can run my Adobe applications, rather than having to wait for Adobe to create a UWP version (which they're obviously not going to do)?
Or is it yet another version of Windows 10? Started from scratch? Yet again?
No,
The systems they are discussing is a port of W10 to run on ARM - not Windows Mobile, which is a cut-down version with replacement parts for mobile, albeit built from the same source and able to execute well-written UWP apps via a re-compile only.
The source code for this system is the same as fill W10 wherever this is feasible (I presume).
Vast amounts of an OS are written in (say) C, or C++ etc. - all of that will compile to ARM and run as intended (when they correct for bad programming). For instance, they may specify that the ARM must run in little-endian mode (I would).
So, the main work being done would be something like:
1. Re-write as much Intel assembler based kernel code in portable C.
2. Where 1 is not possible, reduce the amount to a minimum and write ARM-specific versions of the sections in question, say memory block copies. These would then be as efficient and fast as the originals.
3. In order to actually execute Intel .exes without a re-compile, use said emulation of instructions. I believe Qualcomm are making this easier by baking lots of this into hardware. However, this would only be needed for legacy apps where the money has been spent and the vendor is unwilling to (or unable) to replace the original with a (free or cheap) ARM build for W10 specifically.
In addition, the Cshell project is running to use the same 'Start' system across all W10 platforms (XBox, Hololens, Mobile, Tablet and PC) to avoid having to work on them all separately.
I am ancient but use very few old programs, I am willing to forego Putty if required to gain the advantages of huge battery life and/or size/weight trade offs.
But what I really, really want, is a piece of hardware running on a very small tablet-like device with a touch screen about 15cm across with voice call capability and the very cool ability to run anything a PC will run, especially if a BT mouse/keyboard were connected.
Just gotta eke out this phone for another year.
Intel .exes without a re-compile, use said emulation of instructions. I believe Qualcomm are making this easier by baking lots of this into hardware
At which point, Intel sues them into a very, very large hole in the ground. And even if Qualcomm eventually wins (unlikely - Intel has the lawyers to keep Qually in court until the heat-death of the universe) the process will have taken a sufficiently long time that the processors will be effectively obsolete.
This post has been deleted by its author
"Microsoft is giving up on smartphones, but loves the idea of ARM-powered laptops running Windows 10."
It also loved the idea of smartphones and of ARM-powered Surfaces.
The problem with BigCos (Microsoft is far from alone in this) is that if something doesn't take off quickly enough (and "enough" seems to shrink these days) or its champion leaves the business then the product gets dumped along with all the customers and secondary vendors that had bought into it. Ironically this should make potential customers more wary so that it becomes harder to get something to take off. In reality Barnum applies.
(or at least since I first acquired a laptop, in about 1993 or '94)
I've been waiting for this (kind of thing). Something I could stuff in the backpack for a weekend - or (much better) a week - away, up in the mountains.
I wonder if we can get a spec something like this wishlist? We have the processor and (I imagine) the solid-state storage, so an e-ink screen would complete the trio of hardware fundamentals.
It's coming. Just as I'm getting too old to take advantage, and struggle ever more to lug camping gear and a week's worth of food&stuff over the mountains and still enjoy it.
+1 for e-ink. Suitable in all outdoor use scenarios.
And forget about colour. For most real use cases I struggle to see why colour is essential. Nice yes, essential no. (And perhaps it would make some strategy boutique *onks consider what their "work" looks like printed in black and white on the office laser.
Ok, 1f you're going off into the wide blue yonder and putting some maps on the thing, I agree colour would be more than nice.
I'm sure it will do so absolutely fine. Sounds like the top spec one will have more grunt than my standard i5/16GB work unit. which is currently running all three of those things, along with a dozen other things (inc a VM and chrome+35 tabs), and barely taking up half the available resource.
We all love a good MS bashing, there are enough things to choose from to remain factual and "grown up sounding" in our criticism though.
Multi day life but they turn the brightness of the screen right down and turn every device off every 10 seconds to save power. Every time you want to do something you wait 5 seconds for cpu and wifi adapter to be woken up, in a still throttled state to conserve energy. Registry tweaks needed for wifi adapter to use more than 35% power to get a range of more than 6 metres /sarcasm
Multicore ARM chips are good at what they do and how different cores are used at different volatages as needed but they are hardly powerhouses. They would have to slimline all software down to make it not top heavy, which wouldn't be a bad change of direction for windows software.
No, because the BIG reason WinRT failed was that the programs they wanted to run (x86-based) didn't. Qualcomm's betting they can beat that with an x86 emulator. Now, Intel's made patent threats, but we don't know how far this goes since we haven't seen actual patent numbers yet, plus there's existing prior art in emulating x86 instructions AND there's also AMD to consider.
AMD didn't copy or emulate anything, they have Intel's permission to use any x86 patents.
In order to secure the original IBM PC contract, AMD were handed a perpetual license to use x86 by Intel at the insistence of IBM. Part of this deal also required Intel to hand over all their new CPU designs so that AMD could manufacture them too. Having a second source of parts from another supplier is standard practise in the electronics industry, basically insurance against a single supplier going bust or being unable to supply the parts needed for the product.
Intel played nice until AMD's versions proved not only faster but also cheaper than the Intel equivalents, at which point Intel started delaying handing over their new designs for as long as they could. AMD sued Intel over this - a case AMD won - but ultimately it proved quicker for them to design their own processors so they have ever since.
AMD then developed the 64-bit x86-64/x64 version we all use today, which Intel had to license from AMD.
All right, then why haven't the developers of x86 software emulators (like in DOSBOX and MAME) been sued and ordered to stop development yet? That's clear-cut prior art at this point AND a potential instance of indifference and lack of due diligence, a potential avenue to challenge Intel's patents regarding emulation.