[HN Gopher] Linux Runs on Raspberry Pi RP2350's Hazard3 RISC-V C...
___________________________________________________________________
Linux Runs on Raspberry Pi RP2350's Hazard3 RISC-V Cores (2024)
Author : walterbell
Score : 141 points
Date : 2026-01-03 16:34 UTC (6 days ago)
(HTM) web link (www.hackster.io)
(TXT) w3m dump (www.hackster.io)
| sylware wrote:
| I wish I could buy RP2350 with the ARM cores being hard-fused
| disabled, "cheaper" since no ARM royalties would have to be paid.
|
| That said, I wonder how much they did improve their hazard3
| design, because we all know the future is no PI locked ARM cores.
| I wonder if they are sharing part of the design of other open
| source RISC-V cores.
|
| If those efforts are kept significant, the future is looking good
| and better there. Hopefully, all that will be a success (=latest
| silicon process, ultra-performant RISC-V implementation in
| mobile/embedded/desktop/server).
| danhor wrote:
| The Hazard 3 is _basically_ a hobby project of Luke Wren, a
| Raspberry Pi Employee. He 's contiuing to evolve it further,
| but I don't think it's ready for a full replacement of the
| Cortex-M yet, especially in regards to the Security Features.
|
| The source code is all from Luke Wren and I don't think other
| cores use the source code directly, but improvements to test
| harnesses or general implementation patterns as well as better
| software support help other cores:
| https://github.com/Wren6991/Hazard3
|
| For the SoCs I would expect to see an off-the-shelf Risc-V core
| (certainly no Hazard3 as the main CPU), but we'll see.
| sylware wrote:
| linux started as a hobby project...
|
| I am curious to know which RISC-V design they'll go for in
| this SOC.
|
| M. Wren getting real hard experience on RISC-V is going only
| to help RP to select and audit more seriously any RISC-V
| design which would make its way in their SOCs.
|
| I just don't want to contribute to arm IP racketering (and we
| have mpeg and hdmi to take into account too with avX and
| eDP/DP).
| Findecanor wrote:
| The Hazard3 in the Pico 2 are bigger, more capable cores than
| the Cortex-M0 in the first Pico, and therefore in general
| faster at the same clock for compiled code.
|
| You're supposed to be able to just recompile most Pico
| projects to use them as long as there is no ARM assembly in
| it.
|
| They are only inferior to the ARM Cortex-M33 cores in the
| Pico 2.
| PunchyHamster wrote:
| So they come inferior in the only place it matters, gotcha
| ipdashc wrote:
| I know the "basically" is probably doing a bunch of heavy
| lifting, but dang, that's still awesome to think about. I
| didn't know hardware development was at the point where a
| hobby project CPU, apparently mostly developed by one guy,
| can realistically end up in a mass produced product like
| that.
|
| Quick edit: sounds like "basically" wasn't doing that much
| heavy lifting after all, wow
| https://www.raspberrypi.com/news/risc-v-on-raspberry-pi-
| pico...
| Tharre wrote:
| You can probably already get that if you order a somewhat
| significant amount of chips directly from Raspberry Pi. They
| seem to already have everything required for it, it's literally
| just setting a bit differently during factory programming.
|
| But I'm assuming you're talking about for consumer use, in
| which case my question is why? There is absolutely no way
| you're ever benefiting from them spinning up an extra SKU with
| significantly less volume (most people want the ARM cores).
|
| Even if they decide to eat the costs for the benefit of
| consumers, at most the chip would be what, 15 cents cheaper? I
| really struggle to see how that's a meaningful difference for
| hobbyist use.
| phkahler wrote:
| >> I wish I could buy RP2350 with the ARM cores being hard-
| fused disabled, "cheaper" since no ARM royalties would have to
| be paid.
|
| Better yet, put 4 RISC-V cores on there!
| rmu09 wrote:
| It would probably make more sense to target something like this
| https://www.hackster.io/news/andrew-bunnie-huang-prepares-th...
| with linux, has MMU and more RAM.
| sho_hn wrote:
| This one runs Linux on the ARM cores rather than the RISC-V
| cores in the same package, so it's apples and oranges, but
| still pretty neat for something that comes essentially with a
| companion MCU and is in the same form factor:
|
| https://milkv.io/duo
|
| In embedded systems it's very common to have a bigger SoC and a
| satellite MCU to handle e.g. network comm and power lifecycle.
| I've still not really tried out my Milk-V Duo, but it's
| interesting to get a combo like this in a hobbyist board form
| factor.
|
| They also claim desktop-class performance for this RISC-V-based
| miniITX board, it's a bit weird though since it doesn't claim
| RVA23 compliance:
|
| https://milkv.io/titan
| arjunbajaj wrote:
| I'm building Fostrom (https://fostrom.io), an IoT Cloud
| Platform. We have Device SDKs to simplify integrating
| devices, powered by a small Device Agent written in Rust.
|
| I wanted to support RISC-V boards too, so I went with the
| Milk-V Duo S as the test device. I have managed to get
| Tailscale working, and our Device SDK works too, with the
| bundled Python.
|
| The experience of using the Milk-V Duo is definitely not as
| straightforward as the Pi Zero, but it does work, and is
| easily available in most places, unlike some of their other
| products. The Linux distro they provide is quite barebones,
| and I wasn't able to get Debian working. The docs for the
| device are pretty decent. I hope we get better support for
| Debian/Alpine/Arch for these kinds of boards soon.
| jacquesm wrote:
| Availability is a major factor in projects like this.
| jokoon wrote:
| I don't even know what would be the point of such device, and if
| there really is a market for it for tech companies. I would say
| yes?
|
| As a consumer, all I want if a very minimal "phone", with wifi,
| touch screen, battery, but no 4g or mobile networking, and linux
| on it. Just the cheapest, smallest, wifi, battery-powered,
| touchscreen LCD device that could exist that can run executables.
|
| People are going to say "but just buy a cheap phone", but I
| cannot really run custom software on those, I can't expect to
| install a custom system image, and generally even cheap phones
| CPU SOC are way way too powerful. Open source phones are
| generally crazy expensive and very powerful, and I am not going
| to buy those.
|
| With the range of hardware that exist out there, I think such
| device could cost about 60 euros, and it would be more
| interesting than a RPI.
|
| The RPI is an amazing product, but it lacks an all integrated
| consumer device with an actual screen and battery. Of course I
| can already build one with a compute module etc, but it's not
| really portable and not designed around a flat battery.
| tecleandor wrote:
| First, note that what's on that article is not an Raspberry Pi,
| it's a Raspberry Pi Pico.
|
| I'm not sure about what you want. You don't want a phone, so
| you want... A tablet, maybe? Or a Raspberry Pi 3/4/5 with a
| touch screen? Do you want a small screen? Big? Keyboard/Keypad?
| Touch only?
| bee_rider wrote:
| They don't make cellphones, they make little computers.
| Hypothetically you could probably use an Arduino or an esp32 to
| make a cellphone, but you wouldn't say those organizations lack
| an all-integrated consumer device, right? They just aren't in
| that business.
| mghackerlady wrote:
| You can buy touchscreen enclosures for the various pi models.
| If you want to put in the effort to make your own PCB you could
| really slim it down by using a pi compute module
| numpad0 wrote:
| > I think such device could cost about 60 euros
|
| Making a thing made, shipped and delivered at 60 euros each on
| credit cards is HARD. 600 is easier. 60m consider it done. I
| wouldn't want to even make and sell, idk, a lens cap for a
| camera, at that price, without first owning couple factories
| running at healthy levels of utilization. Maaaaybe if my
| customers would be happy with a 3D printed artisanal version of
| it. But even then it'll technically be delivered at a slight
| loss.
| PunchyHamster wrote:
| after https://dmitry.gr/?r=05.Projects&proj=35.%20Linux4004
| nothing's really impressive when it comes to making it run linux,
| doubly so that it runs just normal Debian userspace
| ge96 wrote:
| What about Commodore where it took 7 hrs to boot (video YT)
|
| Edit: oh damn this one is even worse 8.4 days ha
| mosura wrote:
| That is too practical. For ultimate HN bait it needs to include
| Rust deployed (over QUIC) as WASM on RISC-V.
| IgorPartola wrote:
| But the CPU needs to be build in Minecraft.
| apitman wrote:
| Rust and QUIC are bloated, but I'll take WASM and RISC-V
| RobotToaster wrote:
| Has anyone run linux on thermionic valves yet?
| dmitrygr wrote:
| I'm on it
| balou23 wrote:
| You'd better be, we have certain expectations from you now
| /s
|
| But more seriously... I don't think that Linux has ever
| been booted on a non-monolithic CPU (I wanted to say
| 'discrete cpu' first, but there's some PDP-11s with 4 chip
| CPUs)
| dmitrygr wrote:
| one could argue that 4004 is it. It does not really do
| memory ops at all. all decoding for memory ops happens
| inside the memory chips. but yeah - not non-monolithic
| enough
| b00ty4breakfast wrote:
| Is that the same Dmitry that got the linux kernal to run on an
| atmega?
| PunchyHamster wrote:
| also on 3 8 pin chips. I listened to interview with him and
| he decided to wait 10 years to give competition a chance then
| try to do it on raw transistors or vaccuum tubes
| hdndndb wrote:
| Running native vs running under emulation is a complete
| different beast
|
| Any turning machine can boot Linux under emulation
| PunchyHamster wrote:
| turing machines are far superior to 4004 to, normal machines
| don't have infinite memory for example
| nxobject wrote:
| ...Linux nommu, and the system busses of these platforms,
| won't get the complete environment that Dmitry wants,
| unfortunately.
| Perz1val wrote:
| Linux in PDF was impressive in a different way
| mghackerlady wrote:
| I'd really like to see a pi zero 3 that runs on RISC-V, it might
| be the kick in the pants RISC-V needs to go "mainstream" at least
| in the public view. People know raspberry pi, nobody outside the
| RISC-V scene really cares or even knows about MILK-V
| eek2121 wrote:
| RISC-V needs an inexpensive, performant part for it to be used
| in the consumer space.
|
| The fastest available RISC-V consumer chip is orders of
| magnitude slower than a Raspberry Pi 5. Example:
| https://browser.geekbench.com/v6/cpu/compare/15998376?baseli...
| lenerdenator wrote:
| So we've got a truly free OS; how free is this implementation of
| RISC-V?
|
| Honest question, I don't follow much on the hardware front.
| Zhyl wrote:
| It's an open source core design [0], so pretty open.
|
| [0] https://github.com/Wren6991/Hazard3
| walterbell wrote:
| Linux on Pi Zero can be used as a sidecar for iPads.
|
| https://schwarztech.net/articles/my-ipads-raspberry-pi-sidec...
| This method configures the Raspberry Pi [Zero] in "gadget" mode,
| creating a private link that appears as a wired Ethernet
| connection over a USB cable. The Pi's address will never change,
| either as it is also a DHCP server for the iPad. As an added
| bonus, the same cable powers the Pi, creating a setup as simple
| as plugging the Pi into your iPad and waiting for it to appear ..
| you have a full-blown Linux computer working in tandem with your
| iPad and the iPad acting as a directly-connected console that is
| also providing power.
| 0x457 wrote:
| Any non RP2xxx pi can be used like that. Unrelated to this post
| at all? I only fall back to that when there is no interwebs,
| otherwise connecting over ssh to a normal server is far more
| productive.
| Aurornis wrote:
| The Pi Zero is a full SoC which naturally runs Linux. Any Linux
| SoC with USB (so basically all of them) could use used in the
| same way.
|
| This article is about the RP2350's microcontroller cores. It
| doesn't even have an MMU, so running Linux on it is much more
| interesting. It's not as capable as the normal Linux we run on
| bigger boards, but it's still interesting.
| walterbell wrote:
| Could Linux on RP2350 cores work with iPad via Pi Zero form
| factor, with microSD storage to supplement iPad?
|
| https://www.cnx-software.com/2025/07/08/rp2350-pizero-
| rp2350...
| ge96 wrote:
| Gotta get Pi Zero 2, way stronger hardware
|
| side note: I wish it had more ram so you could mount it in VS
| Code vs. developing through SSH
| panzagl wrote:
| Thanks for linking this, I would not have expected this usage
| (more from the iPad side being locked down) and it has inspired
| some thought.
| walterbell wrote:
| It can be used to mirror content for offline viewing, e.g.
| share sheet sends URL to iOS shortcut, which uses SSH to send
| yt-dlp or wget archive command line to Linux sidecar, which
| has microSD storage (more capacity, lower cost/GB). Linux can
| host WebDAV/CalDAV and SFTP content for iPad apps and media
| players.
| MomsAVoxell wrote:
| I carry around an pwnagotchi[1], it serves not only as a cute
| little paper display, but also a battery powered Linux box.
|
| Can do everything you wanna do, in a cute little box attached
| to a backpack.
|
| My battery lasts about 3 hours, but its easy to juice it up if
| we go hiking and need a common filesystem up some mountain
| somewhere, know what I'm saying ..
|
| (PS - the pwnatotchi part is just an app you can disable if you
| just want your Linux box to be a Linux box, but it sure is a
| fun little toy also ..)
|
| [1] - https://www.thingiverse.com/search?q=pwnagotchi
| mbreese wrote:
| I used to do the same thing with a full RPi4. I'm pretty sure
| that was the first non-pi zero that worked in gadget mode.
|
| I was trying to make an attached Linux machine for my iPad
| instead of using a cloud VM. It was pretty effective, but at
| the time the ARM distributions were difficult to run what I
| wanted. Now, however, it would be significantly better.
| walterbell wrote:
| Going in the other direction: recent Intel laptops with
| Thunderbolt should support USB gadget mode. In theory, this
| would allow a Linux laptop to emulate keyboard and mouse
| input to an iPad Mini portrait sidecar, with seamless
| kb/mouse context switching onto USB, as the cursor moved into
| the configured virtual viewport of the iPad "monitor".
|
| Similar to synergy/barrier/input-leap, but using USB gadget
| mode instead of local network, since iOS wouldn't allow
| installation of non-Apple keyboard-mouse remoting software,
| https://github.com/input-leap/input-leap
| fsiefken wrote:
| Alternatively the RP2350 can also run Fuzix
| wbsun wrote:
| Oh running Linux on a MCU without MMU.
|
| I feel there is a gap between MCU and modern CPU, and also
| between the software running on top of them. The missing piece is
| a mid-size computer with: - A processor, single or multi-core,
| with computing power like 20 years ago, but modern fabrication
| process. Maybe without MMU for simplicity. - RAM between
| 100MB~1GB and DDR2/3 bandwidth. - An OS designed and implemented
| for this type of hardware rather than tailored Linux.
|
| I don't think you can use it for working or your daily
| entertainment, so I guess not a good business to attract
| interests.
| synack wrote:
| That was Intel Quark. It was too expensive for the "big
| microcontroller" use case and too power hungry for the "small
| Linux" use case.
|
| The marketing was confusing, I'm not sure Intel even knew what
| it was for, except to show investors they had an IoT play.
|
| https://en.wikipedia.org/wiki/Intel_Quark
|
| I found some of these boards in a box last year and was unable
| to do anything with them... Intel has thoroughly erased all
| documentation and SDKs from the internet. If anyone has those
| artifacts, please push to archive.org
| itomato wrote:
| Buildroot is awesome and deserves more love
___________________________________________________________________
(page generated 2026-01-09 23:00 UTC)