People have been trying to launch Linux tablets for years from PenPod 700 to Jolla Tablet, or more recently NTablet. You may not know or remember about those, as Linux tablets that actually shipped never really gained traction.
But in early 2019, Pine64 started to mention development work on PineTab, an Allwinner A64 powered BSD/Linux tablet, and the company/community is really good at developing low-cost hardware and providing decent firmware support, so hopes were high. After COVID-19 related delay, Pine64 has now announced the first PineTab tablets would go for pre-order at the end of the month for $99.
- SoC – Allwinner A64 quad-core Cortex-A53 processor with Arm Mali-400 MP2 GPU
- System Memory – 2GB LPDDR3 RAM
- Storage – 64GB eMMC flash, MicroSD card slot, M.2 slot for SATA SSD
- Display – 10″ MiPi 720p Capacitive LCD
- Video output – Mini HDMI up to 4K @ 30 Hz
- Audio – Speakers and Microphone
- Cameras – 5MP rear camera, 2MP front-facing camera/webcam
- Connectivity – WiFi, optional 4G LTE M.2 card (multiplexed with SSD)
- USB – 1x USB 2.0 A host port, 1x micro USB 2.0 OTG port
- Expansion – M.2 slot for multiple expansion boards for LTE, LoRa and SATA SSD
- Misc – Volume rocker and Home button, optional magnetically attached keyboard
- Battery – 6,000mAh battery
- Charging & Power Supply – 5V/3A power barrel jack
- Dimensions – 260 x 175 x 11 mm
- Weight – 600+ grams
The PineTab itself will go for $99.99, and the optional detachable keyboard for $19.99. Just the PinePhone Braveheart Edition, PineTab Linux tablets offered at the end of May will be for early adopters who may not mind some small defects like a missing pixel on the display.
The tablet will support various Linux distributions, including PostmarketOS, Arch Linux Arm, and UBPorts Ubuntu Touch. A demo of the latter has been uploaded on YouTube to show how it well works so far, and it’s pretty good considering the relatively low-end hardware.
A few more details can be found on the Wiki, and PineTab should be up for pre-order on Pine64 store by the end of the month.
Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.
17 Replies to “PineTab Linux Tablet Coming Soon for $100. Watch an Ubuntu Touch Demo in the Meantime”
OK, this thing has a Cortex-A53 processor and costs $120 USD with the keyboard. But this thing:
…also has a Cortex-A53 processor and starts at $1,000 USD! As of today (18-May-2020) the $1,000 device’s campaign on Crowd Supply has raised $135,398 of the $115,000 goal with 117 backers and 31 days left to go. So it looks like they’re going to make it.
Sheeesh, I must be missing something to justify that $1,000 price tag – either that or all of their 117 backers are getting taken for a ride.
the thing about the 1000$ laptop is that it has no vendor binary blobs, you got access to all hw and source code. this makes the design proccess a nightmare and requires a significant sw effort. the pinetab uses the chepest components and relies on community sw efforts this makes it much cheaper.
> the thing about the 1000$ laptop is that it has no vendor binary blobs
…except the DRAM initialization so in other words: it can not even boot without a BLOB.
While I really appreciate the Reform approach trying to be as open as possible it simply outlines the challenges involved. BTW: with old Allwinner hardware like an A64 you get a full open source boot process since linux-sunxi community even reverse engineered DRAM initialization.
even then you hardly get the HDL files for the chips used, not even the tape out designs from the fab, buhu.. i guess the closest you can get to a fully open design is a fpga based soc using an open source toolchain all the way..
What blobs are required for the A64? I believe that there is a free graphics driver for the Mali GPU?
As far as I can tell none
RAM initialization might be a problem. I read the Allwinners might also carry a secondary processor that might run proprietary surveillance code.
The A64 has FOSS DDR3 initialization code now. There also is a secondary processor but we have fully control over it: https://github.com/crust-firmware/crust
You’re confusing with your beloved broadcom chips Jerry.
@Anton Fosselius said: “the thing about the 1000$ laptop is that it has no vendor binary blobs, you got access to all hw and source code. this makes the design proccess a nightmare and requires a significant sw effort. the pinetab uses the chepest components and relies on community sw efforts this makes it much cheaper.”
So I pay $1,000 for dirt cheap laptop hardware but the $1,000 price is a “reward” for the seller to release “open” source/drivers that anyone can use? I don’t get this business model. If the HW/SW is truly “Open” it will take literally a days or weeks before there will be “clones” on the likes of Aliexpress. Again, if things like proprietary graphics drivers, SDRAM/UEFI lock-outs etc. are truly bypassed and/or work-arounds while the source is “open”, how do they make money to pay for all that “significant sw effort” as you said?
So, basically you need to pay insane amounts of money to get a product the way it has to be done. Good business, they make things wrong, closed and as shitty as possible to keep them low cost? I’m sorry but I’m not buying that. Things should be open by default and by law, if you make hardware, I’m sorry, bear with it. Stopping the progress because you are stupid and have the power to stop it should be punishable. If companies have community support, there is no excuse at all.
In theory that would be ideal. The practice is far away from this, simply because it’s terribly difficult. I am an opensource developer and at my work we make products around the project. You probably have no idea how difficult it is to make a build system that is portable and reusable out of your own environment, to document it with variations that don’t depend on environment, even to enumerate all your components, figure their licenses and the original download URLs for the sources.
For example, a long time ago we used to assemble pre-built packages, and if one package resisted a little bit, someone finally managed to compile it by hand so that we didn’t have to go through that pain anymore. In such a case you only focus on 5% of the whole system’s code and trust the rest that you once validated. But once you start to wonder how to provide all these sources, you discover that some of them are lost, that some projects do not use versioning and that the old version of a lib you depend on is not available anymore for download, or that you’ve lost the trick you used to build one part.
All this is not a big problem when you’re transforming an internal product into something you’re going to sell to customers who insist on getting your solution, and all the time you think “we absolutely need to make all this more industrial” but this is a huge pain that only slows down your process with no visible benefit. Eventually you decide that you absolutely want to be able to rebuild your product out of your environment and you discover that it takes one year full time to a developer to automate each and every package, to document the process etc. This is also why we’re now solely focused on a somewhat distributed build system that strives to work anywhere, in order not to step back by accident into this old painful situation.
As such, it seems very important to me to be tolerant against companies that sell products that are not well documented or which do not appear completely open. It’s important to express to them that this matters a lot to you and that you’re not going to buy their products without that, this will give them some real incentive to invest the required amount of energy on this, but it’s also true that as long as their products sell without making that effort, they would be fool to double or triple their development costs for no expected return on investment.
Look at Android. People say it’s opensource because it’s available as source. But it’s not used as an opensource system when shipped by e-waste vendors. A lot of shitty products are only available with Android and not Linux, while only the kernel depends on the hardware and it’s supposed to be the same, and only the rootfs is supposed to differ. It’s exactly the same packaging problem. They are unable to provide in a short time something documented which can be rebuilt out of their own environment and they can’t even extract the automatically patched kernel from their build process. Sometimes some system images may only be built on a single developer’s PC! They know that if they take 3 extra months before releasing their product, it will arrive too late and will not sell at all. So they jump into the crap ocean with all the other ones.
That’s the reason why I have a lot of respect for companies like NXP, intel or AMD who upstream and document a lot of stuff very early. I don’t want to completely blame the other ones, I just want them to make efforts to improve their process with each new product or version.
Interesting. So this should also run Armbian, thanks to the open source community around it?
I can’t help wonder though…why not just buy a cheap $30-35 Amlogic S905 or AW H6 Android 6.0/7.0 TV box available for the last couple years with same or better internal specs, add your own keyboard, mouse, USB hdd + sata etc, and SD dual boot with Armbian on a PC monitor or TV screen?
Not forgetting Odroid C4.
Also the line of still sold, plastic Netbooks with Wondermedia tablet SoCs and ancient software too. bad !
You are mixing things? Like tablet with TV and bunch of extra external devices.
As others mentioned a tvbox and a tv and a mouse are not a tablet. If your TV box doesn’t have a mainline dtb you are stuck with their official kernels *shrug* (I’m unable to write those for myself)
Also pine should provide spareparts, and is usually quite transparent with regards to change management.
Looking on Aliexpress there are atom based Win device around the price . Then the Allwinner A33 BDF Netbooks Android 6, yuk.