Welcome to the July community update – I hope you’re staying cool in the heatwave. This month we’re introducing the Pinecil V2 – an evolution of the original Pinecil – and taking a closer look at the Star64 RISC-V SBC. We also discuss Quartz64’s development progress, the PineBuds and receive an update on PineTime and InfiniTime from JF.
Let’s get into it.
You can watch the synopsis of this month’s community update on YouTube (embedded below) as well as on Odysee and PeerTube. To stay up-to-date with PINE64 news make sure to subscribe to this blog (subscription widget at the bottom of the webpage), follow the PINE64 Telegram News channel, the announcements channel in Discord as well as our Twitter and Mastodon.
I’d like to thank Alex (clover) Ben Brown (Ralim), pillow, JF and PizzaLovingNerd for their contributions to this community update.
N.B. Comments on the blog post need to be in English and follow our Community Rules and Code of Conduct.
July Update video synopsis
TL;DR
Before we get to all the news of this month there are a few housekeeping items. In case you missed it, the most recent episode of the PineTalk was released earlier this month. In this month’s episode, Zed – the show’s editor – replaces Brian as co-host and joins Justin in discussing PINE64 news, gaming on Linux and gaming-related Linux hardware. The latest episode can be found on PineTalk’s page along with all previous episodes. If you haven’t done so already I suggest you subscribe to the podcast’s RSS feedor follow it on the streaming platform of your choice.
It is time for the quarterly Q&A. The Q&A is a community event which provides you with an opportunity to ask questions live in the chat and have us answer you live. In theory, we should hold the Q&A this month, but due to the holiday season – Marek is unavailable at the end of this month and I’m away at the beginning of the next – we’ve decided to schedule the Q&A for August 13th, 8:00 PM CEST (UTC+2). Just as last time, you will be able to pose questions on Discord, IRC, Telegram and Matrix and we will be live-streaming on Youtube. Many people asked us to simultaneously also live-stream the Q&A session to PeerTube – we will try to accommodate this request. A reminder about the Q&A will be released closer to the date via Telegram, Discord News channels and Social Media. Stay tuned!
The last community Q&A in case you missed it
In other news, it is once again possible to pay using alternative methods in the Pine Store via Stripe. For the past few months people only had the choice of PayPal at checkout, but now credit / debit card payments as well as payments via Apple and Google Pay have been reintroduced. We are well aware that people want multiple choices at check-out, and the Pine Store actively explores ways to provide an array of payment options.
Speaking of the Pine Store, in case some of you missed it the Pinebook Pro is once again available for purchase. I know that many of you have been waiting for the Pinebook Pro to make a return, so if you’ve been waiting to pick up a unit then now is your chance. Dsimic is working on a write-up about the PCB of this Pinebook Pro batch, so keep an eye out for his post on this topic in the coming weeks.
The ANSI variant is avaialble now. There is no schedule for the ISO variant.
Lastly, this month PINE64 EU finally launched following a delay due to various regulatory and technical hurdles. The store’s opening has been met with overwhelming positivity and I am humbled by the response. I am also thrilled to see the number of people who decided to check PINE64 EU’s website out and to pick up hardware – I never expected this much traffic. There is more stock on the way and as I stated in the past the selection of available gear will gradually grow. Early August will see the addition of the PinePower, USB-C cables and swag, as well as a restock of the PinePhone and the PineTime. I invite you to follow PINE64 EU’s Twitter, Mastodon and Telegram news channel to be notified when hardware gets restocked.
Two years ago, the Pinecil V1 was announced, and since then has become PINE64’s most popular consumer hardware. Sadly, I’m far worse at Riddles than the great Lukasz, so to cut to the chase, today we are announcing the follow up version two, which is more of an evolution than a revolution. Featuring improved hardware and accessories it retains the same ergonomics and design as the original Pinecil, and it will work with any accessories you already have. It will obviously work with any external accessories, such as cables, but also with existing cases and tips.
There are more improvements than just the teal silicon grip 😉
On the subject of tips, as well as the new Pinecil, we are creating a new line of tips for both the Pinecil V1 and V2. These tips are optimised for a higher power level than the standard Pinecil tips. By reducing the tip resistance from 8 to 6.2 ohms, they allow 65W at 20V USB-PD standard input. These tips will ship with the Pinecil V2 by default, but if you already own an original Pinecil then you’ll be able to pick up the 6.2 ohms tips from the Pine Store and use them with your existing device.
All parts are directly interchangeable and replaceable with the V1
The Pinecil V2 is a natural follow on from the first version, still using a RISC-V processor, but adding noticeable upgrades to the hardware. Key changes from version one:
Things staying the same:
A notable improvement is the new BL706 RISC-V processor from Bouffalo. It is similar to the BL602 in the Pinenut from the Nutcracker challenge. The BL706 features Bluetooth Low Energy; so future IronOS releases will work to enable exposing features over Bluetooth Low Energy. One obvious future feature making use of BLE would be OTA firmware updates. That said, while technically possible, implementing OTA updates will be a non-trivial amount of work. It is something I suspect someone will make at some point for the bl70x/bl60x chips however, and If someone makes it, I’d be happy to incorporate it into IronOS.
Left: Pinecil V2 PCB top // Right: PCB bottom
The other potential future feature that BLE offers is remote telemetry and control. I’m currently thinking about what the future use of it would be, and I am always open to community requests and suggestions. For like a year now there has been chatter in the Pinecil chat about hacking a bl60x into Pinecil V1 to add WiFi/BLE; main driver has been having a remote control screen and possibly integrating into other software to, for example, turn on the air vent automatically while soldering. I suppose one’s imagination and the sky’s the limit, but these two implementations would be legitimately useful additions rather than gimmicks. The inclusion of BLE is directly inspired by discussions that have been going on in the community chat for the Pinecil, and we are looking forward to seeing where the community takes the new feature set.
Additionally the hardware is designed for higher voltages to enable USB-PD EPR support (the higher voltage USB-PD standard is starting to become available). Using this new EPR standard, 28V PD chargers can be used with automatic negotiation.
Pinecil V2 comes in a simpler box, more in line with the aesthetics of other PINE64 hardware
[edit from Lukasz] The Pinecil V2 will be available shortly: it may, in fact, be already available in the Pine Store when you’re reading this. The original Pinecil is the most popular PINE64 device – not counting the industry-focused SOPine modules – and it is also the most counterfeited hardware in our lineup. The Pinecil V2 introduces some measures to combat counterfeiting, which I won’t be discussing as I’m sure you’ll understand, but Ben told me that eventually users will be able to verify the authenticity of the Pinecil V2 on their own. Countermeasures aside, you should be aware that there are no more genuine Pinecil V1 units available and there won’t be any moving forward. In other words, all Pinecils available for purchase now on popular large web-stores aren’t legitimate.
Lastly, many thanks to Ben for introducing us to the Pinecil V2 and for the amazing support from IronOS.
Last month I had the pleasure to announce the Star64, PINE64’s first RISC-V single board computer. If you missed it, the announcement took the form of a riddle, and I was thrilled to see so many people from the community participating and offering responses and guesses. I would also mention that I found some of the inaccurate guesses truly inspiring, and may draw on them for inspiration in the future when we’re working on new hardware. Congratulations to Gav who was the first to correctly decipher the riddle – someone will reach out to you for your shipping details closer to Star64’s release date. For those of you who are interested in the riddle, I’ll do a short run-down analysis at the end of this section.
I covered a lot of the core hardware specifications last month, but just as a recap the Star64 is a single board computer comparable to the Quartz64 model-A – the notable difference being the RISC-V SoC at the heart of the SBC. The StarFive JH7110 64bit CPU features quad SiFive FU740 1.5GHz cores and comes paired with BXE-2-32 GPU from Imagination Technologies. I’d like to note that this is a different SoC to the one introduced last year, not in the least because this one includes a GPU. The board will be available in two configurations, with 4 and 8GB of RAM and we aim to match the Quartz64’s price point of the respective hardware versions. Similarly to the Quartz64 model-A, the Star64 will feature an open-ended PCIe port, USB 3.0 and GPIO. One thing I hinted at last month was that the Star64 will have 2 native Gigabit Ethernet ports. A version with only one Ethernet port will also be available at a later date, and we expect it to cost $5 less.
A look at the final layout of the Star64
Much of the work on Star64 work has now finished and the board in its final layout stage. There is still some testing needed, which will help us characterize the single board computer’s qualities and performance. The initial review has yielded some very positive results, partly because the SoC runs cool without the need for passive or active heat dissipation, even under load. The SoC running cool without a heatsink is great news, as it opens the door for the platform to become a basis for future devices. In any case, the initial impressions are very good and we have high hopes for the Star64 becoming an opener for our RISC-V single board computer range. It will still take some time before Star64 finds its way into the Pine Store, but the engineers are working hard to make the launch happen sooner rather than later. I will keep you posted on Star64’s progress in the coming months.
Finally, for those of you who are interested, here is a short breakdown of the riddle. I titled the riddle Victoria Line Station, because a station on the Victoria line underground in London is called Seven Sisters, which is the common name for Pleiades star cluster. First stanza: people who sing, act and dance and are recognizable are usually referred to as ‘stars’. The final line of the first stanza is a reference to a falling star. Second stanza: aside from the sky, stars can also be found in the sea – sea stars. I also associate stars with magic depictions in kids books. The final line of the second stanza indicates that 64 should be added at the end of the root name. Final stanza: here it would appear that I’ve made a mistake – stars first turn to blue giants, then to red giants, and only then explode. I should have gone back and double-checked my highschool knowledge.
It has been a little while since the last update on the PineSound project and the PineBuds in particular. The PineBuds were first introduced in April and we have since been working hard to accelerate the pre-manufacturing process to bring production units to the market as soon as possible. I now come bearing news that the PineBuds tooling is finished and plastic molding has started. Obviously the plastic chassis of the buds and their case are just one part of the manufacturing jigsaw puzzle, but it is also a very important one. Granted the plastics turn out well, and no major changes to the mold will be necessary, then the next step is to deal with the electronics.
A decision was made to go with no branding on the buds
It may seem a bit counter-intuitive but, in a sense, the electronics are easier to deal with than the plastic molding. The most challenging part in this day and age is perhaps selecting components that can be reliably sourced in the long term. At present, we expect the production of PineBuds’ electronic components to be fairly straightforward, especially given that the prototypes have been thoroughly tested and found to be very reliable over the course of the past 5 months.
First plastic mold of the PineBuds carry case
If everything pans out, we hope to be able to make PineBuds available for purchase in early October. I hope that the Pinebuds will turn out interesting to a wide demographic by offering solid core functionality to those who simply want good in-ear headphones, as well as a deep and easily accessible array of options for those who wish to tinker and hack. In this sense the PineBuds are a very similar product to the PineTime and the Pinecil. Needless to say, we hope that the PineBuds will see the same level of success as the PineTime and the Pinecil have enjoyed.
Before I get into some of the Quartz64 news, I would just like to mention that the first batch of QuartzPro64 boards have gone out to developers. I was glad to learn that the first round of developers have already received their boards and began working on supporting the platform. This was the first dispatch of QuartzPro64 and many more are yet to come. We have received a good response from the development community and the boards sent out earlier this month have found their way into the hands of very skilled developers. We are, however, still actively seeking the best possible candidates for these early development boards, so if you are someone who could contribute to the bring-up process of the RK3588 then make sure to submit your application to purchase a dev unit at half the price.
Thankfully you have quite a few heatsink mounting options Sergio – got a RockPro64 or (very) old GPU heatsink laying around?
Now, let us discuss Quartz64 developments. Earlier this month DietPi added support for both Quartz64 and SOQuartz. The Quartz64 is now officially supported by DietPi, Armbian and Manjaro, but instructions for running Arch Linux and NetBSD have also been added to the Wiki in recent weeks. We also hope to get other OSes on board supporting Quartz64 in the coming weeks. As a side-note, if you are a maintainer of a distro that added support for the Quartz64 then please reach out to one of the mods and notify them so they can add relevant entries on the Wiki. The reason why we’re seeing an increase in the number of OSes supporting the platform is the pace at which the Quartz64 matures. As I showcased last month, the current development pace has helped push the PineNote closer to a more general early-adopter release. Obviously the benefits are mutual, and development on the PineNote has also helped the Quartz64.
DietPi announcing support for Quartz64 this month
I spoke to Pillow this past week to learn what advances have been made in the past month. I was told that they submitted patches enabling PCIe and analog audio output for the Quartz64 Model-B to mainline Linux. A fix for fast SD cards not working on the Quartz64 has also been submitted upstream and will hopefully find its way into previously mentioned OSes with the next Linux installment. Last but certainly not least Pillow patches enabling the JPEG encoder on the RK3566, which benefit not only the Quartz64 but also the SOQuartz and the PineNote, have now been merged. This is a tonne of fixes in a relatively short period of time, and these are fixes that benefit the entire RK3566-based ecosystem. I am thrilled to see the platform taking a hold in the community, maturing and gradually improving.
We released InfiniTime 1.10 “Yellow Mango” shortly after last month’s community update. This new release is the result of nearly 3 months of work by the community: 54 pull-requests were merged, totalling 139 commits and 242 files changes. Thanks to everyone who participated in this new version of InfiniTime! InfiniTime 1.10 contains a lot of changes under the hood: code cleaning and refinements, improvements to the build system (automatic font generation), update of the toolchain, and more. These changes are not visible to the end users but are very important for the maintenance of the project in the long term.
Of course, we have also added some features that will please our beloved users. A lot of them were asking for this for a long time for a possibility to dismiss notifications. This is now possible in InfiniTime – all you need to do is swipe on the notification from left to right to dismiss it.
Here’s a video I shot when testing this new feature on my devkit using the PinePhonePro:
Testing notification dismiss from a PinePhone Pro with keyboard case
We’ve also tweaked the display settings and gamma to improve the color displayed by the LCD, improved the Timer app interface, and reduced the time needed for the watch to go to sleep. These are just some of the more noteworthy refinements, but this release has many other improvements and bug fixes.
InfiniSim, the InfiniTime simulator, has also received a couple of new features: it now supports InfiniTime animations (vertical and horizontal transitions) and allows recording screen capture in GIF format. This is a very useful feature that allows us to easily do demonstrations and share progress with other people. We’ll probably also use it extensively in our documentation in the near future.
Screen capture from InfiniSim showing animations and dismissing of notifications
Siglo also received a few updates and released version 0.9.9 at the beginning of this month. In contrast, InfiniLink, the iOS companion app for InfiniTime, is still looking for a maintainer. Xan-m, the original author, announced a while ago that they would not be able to maintain the project and that they would transfer the project to anyone interested to take over it. In the meantime, the Github project was transferred to the InfiniTime organization on Github. We also plan on transferring the app to an Apple developer account managed by Pine64 to ensure that the app will be available on the app store when Xan-m’s account expires.
InfiniTime is an open source project. It means that the code is publicly available and that anyone is free to do whatever they want with it as long as they comply with the license of the project (GPLv3). Therefore a lot of people contribute to the project by creating pull-requests and asking the core-developers to review and merge their contributions into the original project.
However, some changes cannot always be integrated in the project, and for various reasons: they do not fit with the project vision or priorities, developers do not have the time to review them or… the changes consume too much memory!
I’ve already mentioned this in the past: the PineTime is based on a tiny hardware that embeds a very limited amount of memory – 512KB of flash and 64KB of RAM. InfiniTime does not fully use the memory yet, but I take great care to avoid hitting the limit as that would make the maintenance of the project very difficult.
In such cases, some developers decide to maintain a fork of InfiniTime with their changes. A fork is basically a copy of the project that they maintain on their own. To this day, InfiniTime counts 584 forks! Among these forks, you’ll find Sec42‘s fork that adds changes to make notifications easier to read by “old” eyes and Dmlls‘s fork which adds the Infineat watchface and is very well maintained. These forks are a good thing as they allow developers to make their contributions available to the public even if they cannot be merged in InfiniTime right now, and they provide more choice to the users. Everyone wins!
Speaking of memory usage, I’m currently working on improving the situation by leveraging the additional 4MB of external flash memory available on the PineTime hardware. My work is still in the “proof of concept” state, but shows very good results. The goal here is to store data that needs a lot of space (mainly pictures and fonts) in this external memory to free some space in the internal memory. I’ve recently demonstrated a build with the G7710 and the Infineat watchfaces where pictures and fonts are stored in the external memory. There is a slight impact on the performances and display speed, but it opens the road to many new apps and watchfaces in the future!
Testing background image and font loading from external flash memory
As you can see, the community is still at work on anything related to PineTime! However, don’t worry if you notice a small drop in activity – we are just enjoying the summer and taking well-deserved holidays!
Finally, let me remind you that you can purchase the PineTime from the new PINE64 EU store that opened at the beginning of July!
24 responses to “”
I am happy about this update. We are getting a Risc-V board, have a release date on the earbuds, and steady progress is being made on the Pinenote. I can’t wait to hear how things have progressed next month.
Nice, looking forward to build a RISC-V router with the 1st production unit Star64 🙂
Thanks for the riddle, and thanks in advance for the free unit!
I’ll probably be the zillionth user to ask the same question, but:
do you plan to sell circuit boards of Pinecil V2?
Some of us have extra cases, extra tips, extra screws, extra cables, etc.
We’re basically one simple circuit board swap away from making a 2nd Pinecil (V1) out of all the order-parts.
Also, not mentioned here but I presume V2 of the board still supports adding a hall effect sensor?
I love the idea of a Risc-V SBC!
Risc-V has 1 flaw.
It is extremely difficult to find an online schematic circuit, NOT a block diagram, of the Risc-V.
The only one I have been able to find is from a project called ‘Gigatron’ 4 years ago.
Can you provide a circuit schematic of the Risc-V online ?
RISC-V is an architecture, not a board. Your statement is akin to saying you can’t find a schematic for a Pentium or a service manual for a “V-8”. That’s kind of nonsensical.
There are many dozens of boards now with RISC-V processors. In Pine64 family, there are the Pinecone and PineNut with BL602 and Pinecil with GD32V – and soon, Star64. Schematics for those boards have always been available. I have dozens of RISC-V development boards and schematics for most of them are available from their respective vendors.
LOL
If I want a “V-8” service manual.
I can buy one; and I have.
They contain detailed info and pictures about the engine and it’s working components.
If I want a Pentium schematic.
That’s proprietary information.
It contain millions of transistors configured into a micro processor.
Risc-V is a micro controller that
can substitute for a processor at
a slower speed.
Risc-V has major Billion dollar
Computer players taking interest.
Risc-V was developed at the University of California, Berkeley since 1981 and is Royalty & Patent Free Hardware. It has logic circuits
which are composed of transistors.
Architecture is ‘the complex or carefully designed structure of something.’ I.E. a schematic, and
NOT a block diagram. I hope you
know the difference. I don’t want
a board. I’m not interested in vendors. If you would care to share
with those of us who are interested
that would be nice, but I doubt you
have schematics, Not boards.
Thank You.
Being insulting is not assisting your request for help.
You can get a service manual for a specific kind of V-8 engine, just like you can get schematics for a specific kind of Pentium circuit board. RISC-V is a type or class of processor, just like a V-8 is a type or class of engine. There are easily hundreds of different RISC-V processors now, just like there are surely hundreds of V-8s.
One specific chip (that I mentioned) is on one specific board (that I mentioned) and schematics are found trivially at https://wiki.pine64.org/wiki/PineCone. Another specific RISC-V chip is on a completely different board and has schematics also trivially found at https://github.com/starfive-tech/VisionFive. Here’s another with schematics and even KiCad sources: https://www.cnx-software.com/2021/10/30/open-source-hardware-allwinner-d1s-risc-v-linux-sbc/
There are RISC-V microcontrollers and there are RISC-V processors. The $2 PineNut looks very different than the 1,000 core (!) Esperanto designs, like https://www.techradar.com/news/this-1000-core-risc-v-processor-is-generating-buzz-in-the-ai-space.
If you want the designs of the cores themselves, many are proprietary, but many are open or licensable. https://riscv.org/exchange/?_sft_exchange_category=core If you want the source to the cores like C906 or C910, companies like Alibaba publish those under open licenses, like https://abopen.com/news/alibabas-t-head-releases-c906-c910-risc-v-cores-under-a-permissive-licence/ and you can find synthesizable RTL at https://github.com/T-head-Semi/openc910 as one example.
So if you want source code for processors, above are examples of how to find them. If you want schematics for boards, above are examples of how to find them. But there are hundreds of wildly different RISC-V processors (and even more boards containing them), just like there are hundreds of very different V-8’s. If you’re looking for something tangible, you need to be way more specific in your help request as looking for schematics for “a RISC-V” is about as vague as asking for schematics for “an amplifier”.
Risc-v is an instruction set, not a CPU, not a microcontroller, not a board. Risc-v is just an open source definition of what instructions the processor understands and what they do. The underlying circuit for CPUs is typically generated through code using one of the HDL languages. This is then accompanied by a block diagram, so, while a full transistor layout of the processor could be generated from that, the layout work is not done there. Unless you are designing your own CPUs or are a security researcher investigating CPUs for vulnerabilities it is highly unlikely that you would have any use for this, and if you were one of those people you would already know this information. What is actually desired and what is needed for working with the processor is the instruction set, not the often proprietary internal design.
Every month’s update says that it’s possible to watch the video summary on the Peertube channel, but no new videos have been posted there since April. :`(
Can we get the video update reposted to our favorite freedom-loving network?
I would like a Pine router… Or a SBC kit appropiate for that…
Can’t you use a SBC with PCIe with a multiport network card ?
Yes, that is what I was thinking with the SBC kit idea.. If the Pine64 won’t release a router, at least, maybe, they could release a kit, containing a proper SBC for it, with probably a recommended PCIe network card, Wifi antenna(s), a plastic case(?), etc.. Something in this direction, you know, “use this kit, and you’ll be fine”..
can i get on the list for Pinecil V2 availability. John (eTech, Ottawa, Canada)
“Seven Sisters, which is the common name for Pleiades star cluster”
But one must not conflate this meaning with the name of the Victoria Line station which was named after a group of seven elm trees planted in a circle around a walnut tree on a an area of common land known as Page Green. Thus the seven sisters were seven elm trees.
Perfectly aware of this 🙂
Looks,like I need to save money because I’m in need of a watch, in-ear TWS headphones and I really like the idea of the RISC-V SBC 🙂
It would be really great if the Star64 had ECC RAM, either standard or as an option.
This would make it perfect for NAS systems running ZFS, and for applications where errors are not acceptable, and would make this board a real killer.
Great news on the Star64! I do hope it has SPI flash too boot from into an OpenSBI-compliant environment allowing distros to have a single riscv64 installer instead of the current ARM mess where each SBC needs its own custom image.
That is pretty cool. I had no idea work was being done with risc-v to solve that.
How repairable will the PineBuds be? Since the PineTime is not very repairable (soldered battery is especially a PITA, not to mention the build design meant to prevent water from leaking in.
I think that the battery should be in it’s own compartment (i.e. “pocket”) that can be easily removed, since that it the part most likely to fail (after all, TWS earbuds are a lot harder to build in a repairable manner than a smartwatch, and the same should be done with the charging case (though that will be easier to build in a repairable manner, so the mainboard should also be easily replaceable).
Obviously, the batteries should be of types that can be easily obtained all around the world (for the earbuds, that will be harder, however if it is efficient enough, a button cell battery similar to the Varta cells that Samsung puts in some of their earbuds would do the trick). Perhaps 2x AAA or 1x AA? Panasonic’s Eneloops would be a great choice in that case, and as Ikea’s Laddas have the same exact cells, those would be even cheaper.
Yes! Please make the batteries easily replaceable.
+1 for repairability!
It’s on my wishlist, too!
As far as a possible RISC-V based Pinebook (PVinebook?) is concerned – what sort of timeframe are we looking at? 1 year, 2 years?