Why cant we have Open laptops with maybe an ARM CPU that can last hours and hours on a single charge? There are pretty beefy ARM CPUs (Helios x20, the Exynos) and GPUs (Mali T880) these days, with insanely low power consumption, which can even handle 4K video
There are some GPU brands which offer fully open firmware without any proprietary binary blobs too
>>54268637
I'd love a cheap 11 inch heliox10 laptop with 1080p screen and 50w battery minimum.
Running a Linux distro like Ubuntu
Because we have tablets and freetards are irrelevant
>>54269167
Not just freetards. I'd love a netbook using an 8 core arm chip and a massive battery.
>>54268637
>Why cant we have Open laptops with maybe an ARM CPU that can last hours and hours on a single charge?
We can but people don't like the wooden cases or lugging a car battery around.
>>54269327
like this?
basically a laptop form-factor smartphone
>>54268637
>implying arm is more power efficent than x86
>wanting gimped laptops
>>54269883
Yep, like that, but have the battery extend under the k/m.
I'm talking about 1cm thick thanks to killer battery.
They should do a dual CPU thing kinda how Samsung did it with there processors.
16 core ARM chip would be pretty fucking good for battery life.
I just wish Win10 could work with ARM
>>54270368
Or just use an operation system that does work and also respect your freedom
>>54270377
>Operation system
Nice one, freetard. I like Linux a lot, but still find myself using Windows because I need certain shit.
>>54270687
But for netbook uses Linux should be enough.
I have an asus transformer, I use it for web browsing and typing up notes while listening to music in the background. YouTube and shit too.
I'm sure there are video players for people who want to watch movies and shit too, but that's why existing 8 core arm chips are good enough, hell even something outdated like the SD800 in my Nexus 5 might do the trick.
NO ONE WILL BUY A NETBOOK FOR COMPLEX COMPUTATIONAL TASKS.
The problem is the SoC vendors, they won't provide datasheets or reference designs, they don't respond to requests for info from any company that isn't worth 100 million+ and they refuse to sell their chips to distributors.
I found a company I can get a 8 or 16 core A72 product off right now, it doesn't cost much and I could probably easily crowd fund some nice products based on it. But when I try to contact them my emails go to a third party company they use for sales in my region, they don't want to sell them and try to sell me a FPGA. I even tried to sign up for an NDA and dev tools through their website directly and all they did was forward it to the same company who tried to sell me another companies FPGA products yet again.
I think the only way to get anything out of them is to have contacts that work in the semiconductor industry in CA and have them ask friends who work at the company.
>>54268637
Not a bad idea. There are 8 Core 2.4GHz ARM cpus (64-bit) at 45W TDP in servers. put ubuntu mate on a lapatop like that, and you're good to go.
>>54269167
>posting this on a technology board
>>54269652
Who said it can't be thin and good looking, while being open?
The main idea behind an open laptop would be it's modularity, easy serviceability of parts, none of which would make it thick or unattractive
>>54269932
ARM is more energy efficient than x86, why do you think majority of phones use ARM?
Not to mention x86 costs more, and there are way more proprietary binary blobs, whereas ARM counterparts may have close to none
Because x86 won. x86 is now officially the most energy efficient cpu architecture in terms of performance per watt.
We already have dual-core i5/i7 chips that give laptops/ultra portables crazy amount of battery life.
>"The touch-screen version of the XPS 13 provides a small but insignificant increase in battery life over its immediate predecessor. On the Laptop Mag Battery Test (continuous Web surfing over Wi-Fi at 100 nits), the quad-HD XPS 13 lasted 8 hours and 8 minutes."
http://www.laptopmag.com/reviews/laptops/dell-xps-13
You have to remember that the touchscreen XPS 13 has a 13.3" 3200 x 1800 screen.
Also nothing based on ARM can compete with pic related.
Because x86 is more power efficient and can run x86 OSs and thus x86 software. Fact is ARM just plain blows and is only used in phones because of how cheap the chips are.
>>54268637
There are, look on aliexpress.
>There are pretty beefy ARM CPUs (Helios x20, the Exynos)
They're slower than Intel's Atom gear at equivalent TDPs, only Apple's ARM is faster at the same TDP.
>which can even handle 4K video
Video is universally handled by separate SIP blocks, they aren't software decoded.
Rule of thumb is:
If you ONLY want Linux - Go ARM
If you want Windows ever, go Intel Atom (AMD isn't making <5w x86 SoCs. yet)
>>54268637
protip: cpus actually dont really use all that much power compared to other laptop components.
there are some arm chromebooks though
>>54271340
45watt TDP isn't the same as <5watt, Xeon-D is a shit comparison.
Atom Z8x00 (Cherry Trail) is the comparison you want, and it still rocks.
>>54269883
Sounds heavy and prone to damage
>>54268637
Just get a tablet and put it in a bluetooth keyboard case, goddamn.
>>54271377
>If you want Windows ever, go Intel Atom (AMD isn't making <5w x86 SoCs. yet)
Actually they already have but there is no incentive to use them when intel atom chips already exist.
A10 Micro-6400T
- 4 x86 mullins cores @ 1.2-2.2 GHz
- passmark multi-core: ~1,900
- 4.5W TDP
>>54271340
>>54271399
In the <5W TDP CPU department, ARM still beats x86 in performance per watt, while delivering more performance. In the >5W, atoms beat ARM CPUs.
Also ARM chips are dirt cheap for the performance they provide, you can get a high end ARM chip for <30$
>>54271399
Cherry trail is 4w max TDP isn't it?
>>54268637
See >>54271369
Also
>ARM
>open
pick only one
>>54271399
>45watt TDP isn't the same as <5watt, Xeon-D is a shit comparison.
My point is ARM is fucking garbage in performance-per-watt which is why nobody wants to fucking use them in servers/supercomputers.
>Atom Z8x00 (Cherry Trail) is the comparison you want, and it still rocks.
I guess though they're not as good as core-M chips. Too bad each core-M is like $400 a piece else we would see them x86 tablets more
>>54271475
>In the <5W TDP CPU department, ARM still beats x86 in performance per watt, while delivering more performance.
no it doesn't. Explain why the zenfone 2 gets better battery life than the iPhone 6 plus, Note 3, and Honor 5X despite the chinkfone 2 having a cheap IPS display.
http://www.phonearena.com/phones/benchmarks
>Also ARM chips are dirt cheap for the performance they provide, you can get a high end ARM chip for <30$
You also get less energy efficient chips which is a really bad thing long-term. Investing in a xeon-D motherboard/chip combo will always be better than anything ARM has to offer. Yeah they're dirt cheap but by the time you get an array of ARM chips with the performance of a 45W max TDP xeon-d 1540 you will probably be using like 400 watts of energy to power the turds.
>>54268637
My laptop is open whenever I use it you piece of shit
I've been thinking of buying a chromebook and running linux on it. Has anyone here tried this?
>>54271811
chromebooks run linux by default...
>>54271925
Yeah but you get what i mean
>>54271564
There could be a variety of reasons, maybe their browser is better optimized, maybe they made some changes to the kernel to make it better.
It's a web based test, so even if I take to take it seriously, I cannot.
ARM is RISC, so in theory, it always will have an edge over x86 CISC in terms of power efficiency. It's just that a majority of brands do not care about making fine optimizations to the performance of base Android, just because they can get away with it
>>54271936
Then it depends on the chromebook.
if x86, just install whatever distro.
if ARM, research the model further and find whatever difficulties might be involved, some are just as x86, some are near impossible.
>>54271975
https://archlinuxarm.org/platforms/armv7/samsung/samsung-chromebook-2
It turns out it works just fine with the model I have in mind.
>>54271945
>ARM is RISC
>x86 CISC
Neither are really either.
Modern x86 and modern ARM CPUs don't actually have hardware that processes the instructions of either ISA, then instead have proprietary CPU cores with proprietary ISAs with a frontend that includes a decoder.
CISC vs. RISC died with the first Pentium.
>>54272016
this. x86 is still better because of power efficiency and x86 software support though.
>>54271242
>ARM is more energy efficient than x86
Not at performance/watt.
ARM is better at idle power consumption, but if you wanna say encode some video in the least amount of watts x86 will destroy anything.
>>54270761
Yea I agree with that. When I get a laptop ima install Linux.
Like I said I like Linux, but Win is gonna be my main OS.
>>54269883
The batterie should be wider since the touchpad should be the cellphone itselve and be over the keyboard (why the fuck do you want a keyboard on the lower side ?).
btw MIPS is the most relevent architecture OP is looking for
HP SlateBook, nigga
This is the most interesting ARM platform I have seen for this sort of thing, Marvell mochi. They basically split the ARM soc into separate dies that join together via a high speed interconnect they developed. You can pick how many A72 or A53 cores you want, chose a memory controller and chose which other modules you would like to have such as PCIe lanes or GPUs. Then everything is assembled on a silicon interposer. So in theory you can order a 16 core A72 chip with quad channel DDR4 and PCIE lanes without having to fabricate any new chips. In some pictures I saw it looked like you could just connect them with a normal PCB if space wasn't a concern.
https://www.youtube.com/watch?v=3gmiOLJVPEs
I have not found a way to get in contact with marvell to request more details, they're impossible to deal with as they outsource sales to other companies.
>>54268637
>cpu
But screen uses the most battery, why do you care about the cpu?
I just want a Yeeloong
>>54275177
If the CPU consumes less battery the screen can consume more.
>>54275284
No, that's not the point.
Make non shit screens first as screen is the biggest offender, then work on secondary drains.
>>54275332
Regardless the CPU consumes power, one that draws less will increase battery life.
>>54275400
so instead of plugging the hole in your boat that leaks 80% of the water, you will plug the one that leaks 20% first and call it a day?
>>54275460
it's called job security
>>54275460
I can plug the hole that leaks 20% by choosing a current CPU. To plug the 80% hole I have to wait for manufacturers to make something that doesn't yet exist. Obviously the 20% hole is the one you should start with.
>>54275503
aren't we talking about hypothetical hardware here anyway? its not like its gonna happen anyhow.
>>54272171
performance/watt is meaningless on a phone when in order to get that efficiency you're pumping in 40-80 watts
>>54275549
>aren't we talking about hypothetical hardware here anyway?
No, existing hardware in a hypothetical device.
>>54275642
well then take all of the current best energy saving hardware and put it in a 3d printed case.
