Tinker Board, Asus's answer to Raspberry Pi

And I was just thinking about getting a RPi. Might look into it a bit more

>32bit
kek

>>58596827
Were you planning on slapping 8gb of RAM on that bad boy?

>Rockchip
topkek
>Is it DOA because of the higher price?
Yes, there's plenty of better specced competitors in that higher price range, and they have much better support and less ancient chips

>>58596733
without a support community, extensive documentation and software, is little more than smoother failed RPi clone. The Pi is king because if it's community.

>>58596733
That looks beautiful but we will see how it goes.

>>58596930
While you are partially right, if you have any brains you can adapt almost anything from RPi to the Pi clones because of the similar hardware and architecture. It's only really a problem for the dumb shits who just copy paste everything and then don't know what to do when it doesn't work.

>>58596733
>Rockchip.
>32 bit
>Looks like a fire hazard out of the box.

Ill pass.

>>58596966
People that parrot that shit are basically brain dead, they don't realize that all of these SBC run the same fucking armhf binaries.

>>58596857
Name a better competitor pls

File: prodotti-171631-rel391968ffe9214511a1a7b30d02623665.jpg (586KB, 2548x1911px) Image search: [Google]

586KB, 2548x1911px

Why not play around with the real shit instead?

>>58596827
what would you do with an SBC that could possibly benefit from 64-bit registers and addressing?

>>58597116
64bit science?

>>58597116
Maybe some exotic forced induction or hybrid system. The again a SBC wouldn't make much sense over an LS or LT series engine.

>>58596733
Gigabit LAN is nice, increased processing power doesn't catch my eye since I'm not using the RPi 3 for anything difficult.

>>58596851
>64-bit is only to address more than 2^32b memory

>>58597115
Because prices?

File: coolface.png (861KB, 5000x4068px) Image search: [Google]

861KB, 5000x4068px

>>58597115
>24V

>>58597245
24VAC

http://www.automatikprodukter.se/index_proj_guide_gg0_uk.htm

You should pay attention to the "polarity" even if it is AC when connecting shit.

I fried one when working for an electrical firm.

6000 NOK down the drain.

>>58596857
Name one fag otherwise your full of shit

>>58597017
Do they really? I'm so-so at Linux but I'd like to use it for projects

I'd rather get the ODROID-C2
similar specs plus

ADC pins (!)
USB OTG
Infrared receiver
Optional eMMC slot
64-bit ARM

and it's cheaper too. $40

There's still no open source driver for mali gpus

>>58597115
Is this a programmable logic controller?

File: 10471559-Puzzled-man-at-computer-shrugs-shoulders-and-expresses-lack-of-knowledge-Stock-Photo.jpg (80KB, 1300x863px) Image search: [Google]

80KB, 1300x863px

>be Asus
>be depressed about laptop sales
>look at RasPi sales
>copy-paste product, recolor in photoshop
>tfw

I swear if this garbage has the USB ports and ethernet sharing a lane as well, I will rage.

>>58596733
if it isn't open hardware and/or doesn't support free software fuck off

>>58596733
>still ARM
>56 dollary-doos

>>58596857
Not him. Orange Pi and Banana Pi.

>>58597136
>using a shitty rpi clone for "scientific" computing
literally why

>>58598290
Because you can
99% of what people do with SBC's would be much better served with microcontrollers

Buy a pine64 or odroid c2 instead, cheaper and better.

>>58598225
/thread

>>58598309
yeah I completely agree, which is why spec jerking on them is pretty retarded

>>58597466
Processor is much slower.

64-bit is useless for small SBC workload.

>not using the 9$ chip computer
>2017

Fucking retards

>>58596733
Looks nice. It uses the RK3288 SoC which is in a bunch of other SBCs and some Chromebooks. Probably lots of work from those devices could be carried over.

>>58598074
True. It sounds like lots of stuff works fine without it though.
https://libreboot.org/docs/hcl/c201.html#videoblobs

>>58598146
Open hardware, no. Nothing with an ARM chip is completely open hardware since the processor designs are proprietary. I don't know about higher-level stuff like PCB design though. Free software support, maybe. If it's similar to the other RK3288 boards, it'll probably be possible to run totally free software if you don't need video acceleration.

>>58598355
TempleOS requires 64-bit architecture
It can't be ported to ARM otherwise

>>58598355
>64-bit is useless for small SBC workload.
That depends. Some ARM processors are much more efficient running 64bit code than 32bit code. Is that the case on the processors commonly used in SBCs? I don't know.

>>58597017
Yeah okay. You run a program on a pi that needs proper timing and see how that works out.

>>58598804
Modify it then.

They really need to make these things with 4GB of RAM. 2GB struggles with web browsing. Rockchip should make a simple a73 quad core chip as well.

>>58596733
Gigabit LAN is the only thing that I would really like from the rPi

I wish someone would build actual PCs to use instead of these stupid boards with way too much useless shit slapped on for the 1% that will actually use a Raspberry Pi for tinkering

>>58596851
>what is virtual memory?

>>58597115
Kek Get a Siemens Simatic S7 300 at least. Captcha north Eaton

>>58596733
>blobby video drivers for non outdated kernels
There is some hope but it will probably die anyway.

>>58597017
Nobody cares about the binaries. Compiling binaries for a different architecture is so trivial even a baby can do it. The hard part is to get driver support and ARM is particularly crappy in this area. All their mali GPUs need closed source drivers.

>>58596733
>Mali-T760
>"As of November 2016, drivers are still being in development"

Honestly don't care about the price. Just wish it has a 64 bits CPU.

It's DOA, not because of the price, but because of the support behind it.
Unless Asus can compete against the whole open-source community, it's dead.

>>58597245
Standard for industrial applications.
Impractical for hobby use, though.

>>58596733
It's dead on arrival. For $35 you can get an orange pi plus 2e with 16gb emmc and otherwise the same specs.

>>58599662
RPi uses the USB bus for it's NIC.
The USB 2 is limited to 480 Mbits. So even if it gets a gigabit NIC, it still can't use it fully.

Take a look at Orange pi pc 2. Has Gb NIC and uses it's own bus

>>58598318
This famalam, my odroid c2 is glorious, using it as seedbox, mpd, tt-rss, syncthing and soon vpn node as well

File: Orange-Pi-PC.jpg (120KB, 725x482px) Image search: [Google]

120KB, 725x482px

As much as I hate the RPi and the Pi Foundation they have the advantage of getting in early and having a huge amount of support to draw upon. If any other SBC wants to properly succeed, the manufacturers need to focus on that. My Orange Pi PC is much faster than my RPi but it doesn't do anything half the time because it lacks the advantage of community support.

>>58601279
I'd be inclined to agree but I'm not a huge fan of the mess that the BCM283x chips are, there's a lot of problems with the design of the chip that make it arguably much worse than anything Allwinner has (ARM AXI transactions being forced non-secure, incomplete Cortex-A15 integration, lack of any TrustZone peripherals for ARM which ties in with my first point, lackluster clock and power management) as well as the fact that we still don't know how to initialize half of the peripherals without using the blob.

A64 or other Allwinner boards seem to have all that, though I've not personally tested it myself and from what I heard the secure peripherals will not work unless a certain eFuse is burned.

>>58596733
no SATA?

>>58596733
I literally don't need such specs.

The Pi3 is still cheaper.

>>58597088
>>58597410
ODROID, Banana Pi, Orange Pi, Beaglebone Black, so many, many more.

Just look through the single board computer database.

>>58602110
>ODROID
ODROID doesn't let you run code in EL3 which kind of sucks and require a signed SPL that drops down to EL2 or EL1 IIRC.

>Orange Pi
Orange Pi runs a cheaper version of sun50i, should probably go for Pine64 instead. Pine64 is reasonable (it runs on Allwinner A64/sun50i) but see what I said in the above post. Though at least you get to have your code run in EL3 on dev boards.

>Beaglebone Black
Bootrom will exit secure mode, which kind of sucks. Also old pre Cortex-A15 OMAP3, so no LPAE or virtualization and no AArch64 (obviously).

Bottom line is a lot of these boards have glaring problems, but Allwinner based ones are probably (?) the best unless you want to buy Tegra based Jetson TX1 but it's pretty expensive (dev boards do not have a key fused so you can run code in secure/EL3 mode).

>>58596733
>no SATA
ya blew it

>>58602203
So um, you're saying this Rockchip-based Asus Tinker doesn't have any of these problems?

How many things do I need to create a private network?
Can I put both DNS-server, router and NAS on RaspPi at the same time and connect it to my single laptop?
Well, sounds retarded, actually, but still? I just want to toy with network pentesting with RaspPi servring as prey.

>>58602257
>Router
>RPi
Best thing you'll get is an Access Point, you can't really have something with a single Ethernet port work as a router. You could get a USB dongle and try that, but it isn't ideal and you're likely looking at shit speeds all around because of how both ports will be using the same USB bus for Ethernet, and it's on a Raspberry Pi that is running a few other things.

>>58602286
What about wifi connection for pentesting? Like it is lying flat, emitting waves into the air, and I am trying to penetrate from my laptop?

>>58602254
I've never dealt with Rockchip SoCs so I can't talk about them much, from what I've heard they're pretty crappy and you'd be best of settling for something else.

Also, should probably get an ARMv8 based SoC (that SoC is Cortex-A17 so it's ARMv7-A with a bunch of extensions), AArch64 is pretty cool, while not necessarily useful if you're just going to run Linux on it, if you like low level stuff, it's a whole different world.

>>58602320
If you want to penetration test WiFi things, you're going to want a WiFi dongle that has packet injection support and monitor mode under Linux, so look into something like the Alfa AWUS036NHR or TP-Link TL-WN722N. At that point, you'd want to hook that up to your RPi, install reaver, aircrack-ng, and whatever else you want to pentest with, and SSH into the machine and pentest.

>>58597217
Outside of platform specific stuff that was conflated with a 64bit rollout (such as x86_64s extra registers), that is almost exclusively what it is for. You can also process 64 bit numbers in a single cycle, but I doubt you'll notice much improvement in 99.99% of programs running on this.

>>58599900
You can accomplish the same thing in software with deeper page tables. Not sure if this CPU supports it, but it IS possible.

>>58602414
Dear anon,
>RaspPi as prey
not the other way

>>58596733
How difficult would it be to set up one of these as a Wi-Fi router?

>>58602566
Then set it up as an Access Point, and then pentest on your laptop. There are some guides out there for setting an RPi as an Access Point.

>>58596733
>ARM
>no sata

Seriously, why fuck they keep selling that useless toys

>ARM
yeah no

>>58602520
Cortex-A17 supports the LPAE extensions that were introduced with Cortex-A15 (which also ended up being the AArch64 page table format) but that's mostly for the sake of virtualization, you still won't be able to address more than 4GB of memory without a page table switch.

>>58602672
>>58602667
>ARM
What exactly is wrong with ARM?
>no sata
I think you're missing the whole point of "embedded" hardware.

>Still no USB3
>Probably needs proprietary drivers and modified kernels and shit that will never get updated

>a new developer board announced every fucking day
>not a single one of them truly open hardware or even fully functional with free software

And nothing remotely promising in sight that would fit these simple criteria. To this day, not even the RPi community (which is huge) could make their board work without proprietary garbage. As far as I know, anyway. So these dev board things remain pointless.

>>58602714
>What exactly is wrong with ARM?
No standardized boot process, have to deal with shitty undocumented bootloaders
Drivers are almost nonexistent, if you can find them, good luck patching your own kernel with non-forward compatible binary blobs the manufacturer released for Linux 2.6.36
All available interconnects are garbage, but that's not an issue with embedded hardware.
If ARM got these sorted out properly, it could be a damn fine architecture.

>>58602714
>What exactly is wrong with ARM?
it doesent run my shootan games
cant pretend im a real soldier without shootan gaems

>>58602714
the hardway is strong enough for a home media server (or rather, it's way stronger than what's needed for one). SATA would be nice for attaching a disk so you can stream stuff.
unless it has USB 3.0, then it should be good enough, I guess.

>>58602795
>No standardized boot process,
This is changing:
https://wiki.linaro.org/ARM/UEFI

>>58602802
This is logically sound.

>>58602714
>What exactly is wrong with ARM?
isn't not x86

>>58602933
What's so special about x86?

>>58602939
old habits die hard

>>58602986
What old habits do you have that are bound to x86?

>>58602795
>No standardized boot process, have to deal with shitty undocumented bootloaders
Well, we have device trees if you're talking about higher level stuff (like booting Linux). First stage bootloaders generally differ since everyone uses different EMIF controllers for example, and different ways of managing clocks and power. ARM is not to blame, since it can be licensed out, different SoC vendors choose to do stuff differently.

You can also have UEFI on ARM and treat the "proprietary" (which in a lot of cases it isn't) bootloader as the firmware and happily boot like you would on x86(_64).

>Drivers are almost nonexistent, if you can find them, good luck patching your own kernel with non-forward compatible binary blobs the manufacturer released for Linux 2.6.36
What are you even on about? Most peripherals like DWC2 USB have standardised drivers aside from some platform specific stuff usually. That's literally one of the stupidest things I've heard in a while,

>All available interconnects are garbage, but that's not an issue with embedded hardware.
You mean AXI? Again, what's wrong with AXI/AMBA?

Or were you referring to SoC<->external peripheral (CLCD for example) interconnects? Again, I still don't see the issue, ARM is mostly for embedded systems, you don't need exposed general purpose interconnects like PCIe.

>>58602821
Honestly, UEFI on ARM is just bloat, we really don't need boot services on ARM. The main argument that people are bringing up in this thread is a differently worded "ARM isn't x86_64" so they're unfamiliar with it and like talking shit about it without understanding the primary market most ARM SoCs are aimed at.

>The dream of open Linux hardware is being crushed by proprietary blob ghettos of SBCs and Android Phones

RMS WAS RIGHT

>>58603111
ARM desperately needs some sort of unified infrastructure.

https://www.youtube.com/watch?v=fuAebQvFnRI&t=20m57s

>>58603003
Your arguments basically boil down to "ARM doesn't target SBCs and servers". Guess what the entire point of this thread is about?

File: 1470758451505.jpg (52KB, 320x401px) Image search: [Google]

52KB, 320x401px

>>58597466
>>58598318
>>58601231
>>58602110
My niggas.

ODROID C2 is far better and cheaper.

>>58602997
window vidoe juegos

>>58603393
Well ARM can target servers but again, by SoC vendors that customise their SoCs in such a way, that's the beauty of ARM licensing.

>SBCs
Well, I never actually said that, what I implied is that is that your typical ARM SBC user doesn't give two shits about what the boot chain does as long as it eventually boots Linux. Just treat it as you treat firmware on x86_64 machines since that's what it essentially is. I don't see where I implied that ARM doesn't target SBCs.

>>58603501
This is why proprietary software is holding back progress.

>>58603559
fk u

File: 702c34f0301e4c7e1c04d7bc70a258070ebf9e35.jpg (16KB, 326x315px) Image search: [Google]

16KB, 326x315px

>Binary Blobs
>No mainline kernel support
>Undocumented bootloaders

Is there any reason to get these instead of a mini-ITX + Celeron/APU combo?

>still not a single sbc running freetard firmware
fuckkkkkkkkkkkkkkk

>>58597115
What would be the benefit?

>>58603781
>what is texas instruments beagle bone black

>>58601279
What makes you hate the pi?

>>58603781
>still not a single sbc running freetard firmware
Allwinner dev boards (ie. Pine64), Nvidia Jetson TX1/TK1 (Coreboot instead of NVTBoot), OMAP3 dev boards, there's an open source firmware for RPi, etc.

>>58603898
>>58603920
https://www.fsf.org/resources/hw/single-board-computers

>>58603990
>caring about graphics drivers on an embedded device

>>58602286
>Best thing you'll get is an Access Point, you can't really have something with a single Ethernet port work as a router.
>What is VLAN
Though you will get shitty speeds, but it's not impossible

>>58603990
>https://www.fsf.org/resources/hw/single-board-computers

>The BeagleBoard (various versions) as well as the PandaBoard use the TI OMAP family of SoCs. These come with free startup software as well as free drivers for the peripherals.
So do virtually all boards using SGX chips.

>The AllWinner Axx and R8 platforms come in many boards such as the A13-OLinuXino, Cubieboard, Gooseberry, Hackberry and CHIP.
Same goes for Mali GPUs.

GPUs have been traditionally non-free and required (thankfully for you, on ARM, just userland) blobs to function. There's nothing new here, same applies to desktops pretty much. There are open source drivers like Freedreno that work "well" (just like Noveau works except when it doesn't), I guess that's the best example.

>The Raspberry Pi requires nonfree software to start up. It can't reach the point of executing free software unless this nonfree program is part of the installed system software.
There's an open source VPU firmware that you can use to boot if you'd like. And GPU drivers for RPi are actually open source (when I say GPU I mean the QPU).

>Odroid-X, Odroid-X2, Odroid-U, Odroid-U2 and Arndale boards
Don't buy dev boards that are fused for secure boot and require a signed SPL, that I can agree on.

>>58604053
It's not really worth setting up a VLAN to have some crappy Pi do your routing.

>>58597245
>trollface

I miss this meme

>>58603990
This page gets thrown around a lot here, but it's pretty outdated and not even close to complete.
Feels like /g/ needs a proper SBC guide listing all FOSS options and unavoidable proprietary blobs. Some of you guys seem knowledgeable about these.

>58603747
Why go overkill? Why buy a desktop computer when you can accomplish a task with a little credit card sized one?

>>58596930

They are all the same, you fucking shitface. Almost every code written for Raspberry Pi can be used for other arm-based SBCs.

>>58605469
>what are drivers

>>58596966
>because of the similar hardware and architecture

it seems you have never entered the cruel world of ARM, MIPS or whatever-non-x86-based SoCs. it is painful shitfest ranging from proprietary binary blobs to even start the video output to shitty old customs kernels that are full of spaghetti code or violate the GPL alltogether so you can't even attempt to fix shit. x86 might have flaws, but the standardized approach at thing is a blessing. the raspberry is effectively the only board that has proper documentation and things usually just work.

https://www.board-db.org/

>>58596733
>answer to Raspberry Pi

What's the point? Wasn't this shit just meant to be for schools and education? Why are they trying to compete with that?

>>58603886
this is what pros use, or at least close to.

>>58605810
because the majority didnt buy them for fucking school use

Would it make sense to start an artisan-style small business to make CPU coolers for these nanocomputer boards? I know the Pi3 basically needs cooling. But the thermal outputs are still small enough you can have novelty coolers like sterling engines and pico-water

>>58606220
chinese heatsinks for 30c do their job well enough

>>58606220
doubt it.

a cheap and quickly made 3D printed thing makes more sense cus i doubt people would spend more than a couple bucks on heatsink "art" for a cheap piece of hardware like a raspi

>>58597115
Looks nice... but expensive.

>>58596733
The antenna and the gigabit LAN are about the most exciting parts.

>>58598130
USB and Ethernet are not combined. Some sources also claim that each USB port is separate while others claim it's a hub like on the pi.

File: IMG_0516.jpg (26KB, 480x270px) Image search: [Google]

26KB, 480x270px

All these fags bitching about open hardware, but running Intel I series CPU's.

>>58607244
>complain about the lack of decent open source hardware while running closed source hardware
But that makes perfect sense.

>>58597115
Its shit compared to RPi or Tinker Board
http://products.centraline.com/en/pdf/en0z0970-ge51r0116.pdf

>>58596733
im in a robotics group and we got hahold of an NVIDIA jetson chip so we can do vision stuff on the robot

>>58600771
1 GB shared memory

>>58597115
Theses things are shit and people still integrating them in new projects are fucking retarded.

I'm just waiting for an affordable AMD Opteron A1100 with any kind of graphics support.
It could easily replace my desktop.

>>58597115
Kek

Why not get a Omron NX7 and some EtherCat couplers and IO. Or better yet, Omron NY IPC's where you can have both Windows/Linux in the same environment as your PLC program.

But if you want something slow as fuck go Siemens.

>>58597283
Hvilket firma jobber du for?

>>58602343
I've got an 8Gb librebooted c201 that has been spun up for over a year now and it has been rock solid for me. I use it with XFCE. 4.8 mainline has rk3288. Debian has top tier armhf support.

If they get libreboot on this thing, /g/anons/ should be all over it. Cuz muh freedums.

>yet another chinkshit board with a junk GPU that needs binary blob drivers

>>58610089
/thread

Chance of Windows running on these devices?

http://www.theverge.com/2016/12/7/13866936/microsoft-windows-10-arm-desktop-apps-support-qualcomm

File: Uten navn.jpg (61KB, 485x516px) Image search: [Google]

61KB, 485x516px

>>58609979
Jobbet der i 10 mnd med NEET penger.

ARM SBCs were a mistake. There's no contributions to the open source community, everyone just takes one of the handful of proprietary chipsets and shits out a crap imitation of a RasPi.

>>58605400
the machine that anon mentioned would be several times faster though

like, I can use a raspi as a desktop machine
but it's not a snappy, enjoyable experience at all unless I use ridiculously minimal software (read: Window Maker as my WM, links2 as a browser, various terminal utilities)

>>58610175
unlikely, but possible

man, why was MS so bad with WinRT
why even have a desktop mode if applications can't use it

shit, the fact that .net practically exists for running on not-x86 and they didn't let you run WPF or WinForms stuff is retarded

>>58610285
Right, RaspPi only works because of the community support that comes from having millions of them in circulation and a lot of developers eyes being on them.

Honestly the company that needs to build something like the Pi is AMD. If they were to get a sub $50 board out with x86 cores and GCN graphics it would immediately be in good shape thanks to the large library of x86 Linux software and the excellent Mesa RadeonSI driver.

>>58610346
It would be, but if the additional speed isn't necessary then it's more size/weight/price for no benefit at all.

>>58610357
And even the RasPi has a problem with proprietary hardware, they just happened to have a community that is willing to overlook(/work around) the proprietary parts.

AMD did have a development board planned based on their "Seattle" SoC but I have no idea what the price was or even whether they released them.

>>58610432
Yeah, I remember them talking about having one but I don't remember ever hearing about them being released. I wonder if the Xbone and PS4 contracts killed them since they could in some ways be seen as console competitors.

If they could even get one out for $199, the same price as Nvidia's Jetson dev boards even that would be a step in the right direction.

>>58596930
>The Pi is king because if it's community.

Yep. I don't get why they aren't making their systems RPi compatible. For example, having an incompatible GPIO pin layout is just stupid -- it shuts them out of the whole existing RPi infrastructure.

It might be a bit much to ask to be able to install a RPi distro on it, but couldn't they at least hire a few engineers to spin a distro that's as compatible with Raspbian/RPi as possible? At least compatible enough so that they can serve as a drop-in replacement for Raspbian/RPi 99% of the time.

It's fine to have extra stuff like SATA or eMMC flash. But if Raspbian/RPi has something, then they need to design theirs as identically as possible. Like the format of the /boot/config.txt file -- that kind of thing.

>>58610432
https://www.gizmosphere.org/products/gizmo-2/

I did find this. It seems like a rather poor entry into the market though. It would be more interesting if it had either higher power hardware and more RAM, or was lower wattage and cheaper.

>>58610803
That's a different one, that one's x86.

https://community.amd.com/community/amd-business/blog/2015/06/23/extending-arm-s-ecosystem-for-server-developers

http://www.phoronix.com/scan.php?page=news_item&px=ARM-Opteron-AMD-Snaps

File: 1478713444371.png (372KB, 513x834px) Image search: [Google]

372KB, 513x834px

>>58606178
You don't know what the fuck you're talking about.

>>58601279
fat british RPI fanboys need to fucking die already.
they're holding the world back

>>58610432
You can buy an AMD ARM system today, but it's either headless with no option for graphics, or very expensive
https://shop.softiron.co.uk/

So stupid that the $599 one doesn't have even 2D graphics. It would be perfectly viable as a development machine.

Linus Torvalds recently said, the big break that ARM needs is a desktop development machine. This doesn't exist yet.

>>58597115
I use my pi as a webserver and tor relay, and that thing is way too underpowered for that.

>>58602520
>99.9% of software do not use double precision floats

File: 5_165123_1.jpg (46KB, 640x480px) Image search: [Google]

46KB, 640x480px

>>58611250
>Linus Torvalds recently said, the big break that ARM needs is a desktop development machine. This doesn't exist yet.
Does anyone know more about this? Rockchip announced support for RK3399 multichip boards. In theory you can have 4 of them giving you 8x A72 core 32x a53 core 4x GPUs and 8 channel lpddr4.

>>58612988
16 a53 actually, I was thinking of that mediatek chip with 8 a53s.

File: 1477307549496.jpg (67KB, 640x853px) Image search: [Google]

67KB, 640x853px

>>58612988
Is it actually 4-way SMP?

Rockchip's press release talks about arbitrary x N performance with as many processors as the vendor would want on a single board, which sounds like some form of parallelism lesser than regular SMP we're used to.

>>58596733
>Another rpi knockoff without displayport
Into the trash it goes.

>>58599999

>>58613083
Maybe not, looking closer ethernet and other things are doubled up.

The marvell armada 8040 might make a good SBC if combined with an AMD gpu. I can't get in contact with them though, I have spent hours sending emails and trying social media, never get anything even when I have my own company and company email.

>>58597115
Is that a PLC? Because it totally looks like a PLC.

>>58599954
why bother when S7-1500 is around?

>>58608459
Fuck you guys
t. poor team

Why can't there be a single board which is:
>unlocked bootloader
>video acceleration without blobs
> <$40
etc

I can imagine the only way around this possibly is a minimalist gpu with the majority of GPU hardware being computing shaders - yeah you'd get reduced performance per dollar or watt, but it sounds like literally the only way to get around literally 100% of GPU acceleration being locked down Jewery. Maybe somebody will make a Pi clone with a GPGPU module which we could use to get around this, but until we do, I literally can't take any competitors seriously because although they might offer much better features (SATA, gigabit, cpu/gpu) there's literally no developer advantage.

Similarly, is there a SINGLE FUCKING BOARD with analog in comparable to audio-in? Or fuck, even any WITH stereo audio in. Normal audio ins are AC-only, but at least it'd be a step-up. Same with the analog out, "analog out" sounds great, until you realize 99% of the time it's like 1khz 10-bit or some shit.

Do any boards you use the transistor/mosfet "open connection" for digital stuff? All the ones I've seen are directly driven and then it's like "but don't go over 100mA you'll fucking blow up the whole goddamn board". A single mosfet logic-level transistor can pass like 8A 50V at 3.3V at the gate, I don't understand why nobody is using them.

>>58613611
P.s. are there any TV chips with an unlocked bootloader? I haven't checked lately but only the original MK802 shit had an unlocked bootloader, while every fucking rockchip after had it locked. It's completely retarded I can use a new kernel on a single-core 512-1GB tv stick but can't on a quad-core 2+GB.

>>58596827
>>58597136
>>58597217
Please don't tell me you actually think 64 bit CPUs are faster at 64 bit arithmetic.

>>58613678
Of course not.

File: 1485082866633.jpg (1MB, 3022x4800px) Image search: [Google]

1MB, 3022x4800px

>USB 2.0

>>58596733
>Quad Cortex-A17 1.8 GHz
>about 2x faster than RPi 3
citation needed