They seem to come with Windows ARM pre-installed, but if one wanted to (also?) run Linux on them, which distro(es) would be available?
sangnoir [3 hidden]5 mins ago
An experimental build of Ubuntu 24.10 for Snapdragon X Elite ARM computers is available[1] today from Canonical. It's denoted as "Experimental" because its very bleeding edge, and under active development. It boots and is installable, but is not ready for casual users yet.
It supports standard UEFI boot, Ubuntu has a WIP image you can use that has some extra patches, and with some fiddling you can get other distros, assuming you use a really recent kernel with a supported device tree and maybe you need some firmware updates and a kernel patch or two (search 'x1e' in the Linux tree to see if yours is in there.) Wifi, USB-C etc work last I checked. Not sure about BT or general sound support or anything like that.
Overall though, things "work" but as far as usable goes, it seems to remain in a pretty poor state; especially given the lack of accelerated graphics combined with no NUC/SFF options. I guess WSL2 is probably okay. But it's a shame, because the hardware does seem pretty nice.
When some NUC options come out you'll at least be able to use it in a headless form factor, I guess.
betaby [3 hidden]5 mins ago
None. I was down-voted during the launch for saying that there won't be any meaningful support from Qualcomm and 6 months later ... there is none. Browse reddit thread to see how thing are (non)working on Qualcomm based laptops.
sangnoir [3 hidden]5 mins ago
> None. I was down-voted during the launch for saying that there won't be any meaningful support from Qualcomm and 6 months later
You were - and are - wrong. Qualcomm mainlined enough drivers for in the same window (past 6 months) so one can boot and install the latest Linux kernel on several Snapdragon X laptops from different OEMs. There are people running Fedora and an Ubuntu Experimental[1] builds. Its still rough, depending on ones device tree, the peripherals line microphone, webcam, and/or fan control may not work right now, but it's a platform that's actively being developed for.
With UEFI lots of arm64 distros both Linux and BSD should boot, but quite possibly have very poor periferal support. pcie and graphics seem like likely candidates. If pcie isn’t working you may be able to load a ram disk, but it wouldn’t be usable as a laptop. (Given nvme storage)
6SixTy [3 hidden]5 mins ago
IIRC my X13s has experimental right next to booting Linux, and I doubt any of the newer models are better. Honestly not a great sign even though Windows forced the use of UEFI.
First of all, the post wasn't downvoted, and secondly, someone pointed out that it's being worked on, as another commented pointed out here as well. I hope being corrected twice is enough to prevent repetition.
onli [3 hidden]5 mins ago
It might have been downvoted at first.
wmf [3 hidden]5 mins ago
Weirdly, it had good Linux support before launch and then it somehow went away.
ekianjo [3 hidden]5 mins ago
it's wrong. I have tried a Debian distro running on a Snapdragon based machine just last month. Not everything was perfect. but it worked well enough that you could use Box64 to run some Windows games.
kristopolous [3 hidden]5 mins ago
really? I know there's quite a few arm64 distributions ... that's pretty strange. What about WSL2 and what about NetBSD?
betaby [3 hidden]5 mins ago
"Really" as a surprise or you are confronting the veracity of my statement?
No, NetBSD is not working on Qualcomm based laptops.
I also know quite a few arm64 distributions ... like LineageOS, which works OK-ish on my phone.
This thread however is about laptops in a sense one expects from the laptop, i.e. hardware accelerated video, WiFi 5 and above, bluetooth, woking sleep/resume.
I don't understand what exactly is strange here for you.
kristopolous [3 hidden]5 mins ago
They must not care about Linux as much as I think. I thought it's kinda standard these days to release laptops that plausibly work with at least Ubuntu, in some limited way.
I must live in a bubble.
betaby [3 hidden]5 mins ago
I would say Linux support even for modern x86 laptops is not great. Take for example Zenbook S series, both AMD and Intel. While most of the things are kind of working, they are not that stable. I would say Linux support peaked somewhere on 12th generation Intel laptops.
brucehoult [3 hidden]5 mins ago
I don't have any problems with Ubuntu 24.04 on my i9-13900HX Lenovo Legion Pro 5i.
It's never crashed, I never reboot it, sleep works fine. It spends almost all its time sitting with the lid closed, sshing in from my Mac, using 13W-15W depending on whether it's in eco/smart/performance mode, then boosting to 160W or so if I ask it to do something serious.
When I do take it mobile, it runs for about 5 hours doing light stuff in eco mode, vs 6 hours when it has Windows on it. The difference isn't really a problem as it disappears if you give it some real work to do. That's not Arm Mac battery life, but it's 2 or 3 hours more than I ever actually need. I've got it set up to keep running closed on AC, sleep if the AC disappears. Works great, and it's going instantly from sleep with a keypress if I open it.
Really, no complaints at all.
kristopolous [3 hidden]5 mins ago
For me, ACPI is the one thing that never seems to work quite right. Maybe I'm just doing it wrong
Muromec [3 hidden]5 mins ago
It's not about the arm as an instruction set (I'm writing this from the arm machine running linux). It's the drivers, bootloader, all the fdt glue. You don't just need support for arm, but for the specific board on the specific SOC.
I have the risc thing too, that magically works on mainline kernel somehow, but is slow.
WhyNotHugo [3 hidden]5 mins ago
postmarketOS runs on the ThinkPad X13s. I don't see a wiki entry for this specific model, so I assume it doesn't work out of the box, and "maybe" with an undefined amount tinkering.
ekianjo [3 hidden]5 mins ago
Debian works on it, but graphics support is still very early.
snvzz [3 hidden]5 mins ago
Works is a bit of an understatement, with a package library that's close to surpassing that of ppc64[0], thus becoming the third largest.
I've used mine as a desktop for a while, but now it's been repurposed as home server.
It runs Debian trixie (still not frozen, just sid + small delay), with a generic kernel (not vendor crap) and root on ZFS, a 1TB M.2 SSD, with an external 22TB disk, also ZFS, on UASP USB3 enclosure.
Compared to the rpi4 it replaces, it is somehow actually faster. Sustains 70MB/s aes decrypt (luks), where the rpi4 gave me <50MB/s.
Graphics support is indeed early and tied to an ancient vendor kernel, but Imagination Technologies is funding the mesa3d effort. It will eventually (hopefully soon, there's been hints recently) simply just work with open drivers.
This Reddit thread [1] says that the battery lasts ~22h based on some benchmarks. The vendor claim was at ~29h but I think the point is to be comparable to Apple chips. Indeed is a game changer.
They put the old (55WH vs the newer 61WH) battery in the shell -- Framework you cheapos! I guess they are trying to get rid of the old, smaller batteries.
Also, word of warning just don't apply the 3.08 bios update if you have a 12th gen. I've had nothing but problems with the 3.08 bios update without real remediation from framework. Really, don't perform this upgrade without doing a thorough review of the support forums regarding the issues people are having.
There was a blog post about the BIOS update that i found here on this site, but i'm having trouble finding it for reference now. Someone had the same issue as i have where sometimes my system just won't boot (no video and error sequence) unless i remove or re-add a memory module. Not that fun.
adgjlsfhk1 [3 hidden]5 mins ago
IMO that's a smart choice. This cpu isn't made to be a daily driver. It only is in a laptop chasis to make R&D easier. No one will miss the extra 5WH
_nalply [3 hidden]5 mins ago
Some people will put it as as small server somewhere, then this battery is like a backup power for servers but built-in. Framework even sells a small case for this sort of usage. If I had skin in that game, I would do this. For developping end-user software, just ssh in and program away!
attila-lendvai [3 hidden]5 mins ago
sometimes my Thinkpad Carbon X1 gen6 refuses to turn on until i disassemble it and disconnect the battery... about once a year.
it's rare enough to scare me for a moment each time that it has died on me...
pjmlp [3 hidden]5 mins ago
That is what happened to my Asus 1215B netbook. Taking out the battery used to fix the boot being stuck on UEFI, until not even that last year.
It took about 4 years to die, and served me from 2009 - 2024, so still quite a long life.
grapesodaaaaa [3 hidden]5 mins ago
I have always had issues with this on my Lenovo X1 series laptops. Why is that?
buccal [3 hidden]5 mins ago
I had same issues with some quite new Dell XPS Carbon and Asus ROG gaming laptops. Battery disconection is first thing I do after laptop is not turning on.
Crappy BIOS/power management controller does not depend on price of the laptop or manufacturer.
IshKebab [3 hidden]5 mins ago
Interesting CPU choice. Seems like it's actually intended for image processing...
Consumer image/video processing is one of the few applications where requirements of low cost, high performance, and tolerance of poor software support and low reliability converge. It makes a lot of sense that StarFive would target that application. Most of the SoC vendors from China are competing in this same space, like Allwinner, Rockchip, Goke, Ingenic, SigmaStar, Ambarella, Fullhan, Xiongmai, etc.
boredatoms [3 hidden]5 mins ago
I hope they can update it with a chip that has the vector extension
plg [3 hidden]5 mins ago
Q: why
geerlingguy [3 hidden]5 mins ago
In this case, I'm guessing targeting a solid portable hardware ecosystem at a development board. The chip on here is slow enough it would be a slog for any kind of day to day use... but if you're testing drivers for something or evaluating RISC-V for some remote use case, a battery powered laptop is a little more convenient than toting around a power supply and dev board + enclosure.
I'm guessing the sales numbers will be extremely small—some people who get excited about RISC-V and would consider buying this don't know how slow it is in comparison to even the cheapest Arm SBCs right now.
snvzz [3 hidden]5 mins ago
>don't know how slow it is in comparison to even the cheapest Arm SBCs right now.
I own a VisionFive 2 with this SoC.
CPU sits between rpi3 and rpi4 performance.
GPU is easily 3x that of rpi4. I/O speeds are similar to rpi5.
This is not anywhere as slow as you're making it to be. And certainly it isn't slower than the cheapest ARM SBCs.
But yes, you can get a much faster rpi5 and it is cheaper.
Current RISC-V SBCs and laptops are for enthusiasts and developers, who know full well what they're getting.
Next iteration will be faster. The next, faster still.
MobiusHorizons [3 hidden]5 mins ago
At this point it feels more like a marketing stunt for framework than a legitimate product. Lots of people expect this to be a proper laptop experience in the ballpark of arm or x86, just because of the form factor. I am imagining a bunch of people buying the upgrade kit with the vague idea that it will be in line with some “open” or “libre” goals, but will be severely disappointed at the experience.
arp242 [3 hidden]5 mins ago
Marketing for who? The few thousand people who care?
And they have a clear disclaimer it's a dev-focused board; if you expect a "proper laptop experience" then that's on you the basic description of the product.
"Targeted towards developers, tinkerers, and early adopters, this platform allows exploration and development within the RISC-V ecosystem. [..] This is very much a developer-focused board to help accelerate maturing the software ecosystem around RISC-V, so we recommend waiting for future RISC-V products if you’re looking for a consumer-ready experience."
MobiusHorizons [3 hidden]5 mins ago
I’m reacting much more to the initial launch ideo from framework [1] and the hype around it than the product page. I’m glad to see they are calling it out so well at this stage.
I do not think we need to resort to calling it a marketing stunt. It is for developers and early adopters. This will make a nice device for that. Being able to reuse the rest of the platform is pretty great.
abdullahkhalids [3 hidden]5 mins ago
They explain very clearly in the previous linked blogpost [1] in OP. Some snippets
> There is excellent philosophical alignment between RISC-V and Framework. Both are built on the idea that an open ecosystem is more powerful than the sum of its parts... x86 ... is used in the processors for each Framework Laptop ... The x86 ISA was invented by Intel, extended on by AMD, and is proprietary, with Intel and AMD being effectively the only two companies able to use and create processors around it. ... What makes RISC-V unique is that it is an entirely open architecture, which means that anyone can extend on it and create their own processors that use it without paying a fee. RISC-V International is the collaborative organization that exists to help develop the standard and define common versions to ensure cross-compatibility of hardware and software. There are hundreds of companies now developing cores and chips around RISC-V, but most of these have been hidden away in embedded applications....
> All of this is what makes RISC-V unique from an ecosystem enablement perspective. The actual technology is equally interesting. The base instruction set of RISC-V is simple and streamlined, while there are a number of extensions enabling high performance and specialized compute. This means that RISC-V cores can be developed for anything from tiny control CPUs embedded inside a sensor (the Fingerprint Reader we’ve used in Framework Laptops since 2021 actually has a RISC-V core!) to monstrous multi-hundred-core server processors. The DeepComputing RISC-V Mainboard uses a JH7110 processor from StarFive which has four U74 RISC-V cores from SiFive. SiFive is the company that developed CPU cores using the RISC-V ISA, StarFive is the processor designer that integrated those CPU cores with other peripherals, DeepComputing created a Mainboard leveraging that processor, and Framework makes laptops that can use the Mainboard. The power of an open ecosystem!
We went to the moon because we were losing an arms race and the moon got the population onboard with spending 1% of the gdp on it.
brendoelfrendo [3 hidden]5 mins ago
So the takeaway is that spending funds on an arms race is pretty unpopular, but it turns out that people think cool shit is cool and worth it?
Dalewyn [3 hidden]5 mins ago
The takeaway is that the Space Race was just another front of the Cold War and the Military Industrial Complex got a blank check so long as they could win it. The fact it also captivated peoples' imaginations was a nice side effect.
braiamp [3 hidden]5 mins ago
And they blew up a rocket with people in it [1] in the process, which made everyone there rethink the whole thing from a project planning and design perspective, that it became a blueprint[2] of how to do a successful project
It seems to have no practical purpose. I assume someone funded it to keep RISC-V in the spotlight.
MobiusHorizons [3 hidden]5 mins ago
I believe it was actually designed by the chip manufacturer, not originally by framework. The website says “In partnership with DeepComputing” but I remember an interview where they described it coming mostly from the other company.
adgjlsfhk1 [3 hidden]5 mins ago
the "partnership" from what I understood is that DeepComputing asked a bunch of questions that framework answered (and provided advice) since although the specs for how to make a compatible board were published, this was the first time someone was sinking the time in to do it.
LeFantome [3 hidden]5 mins ago
How does it have no practical purpose? It is a RISC-V developer platform. It allows you to develop and test on RISC-V.
phendrenad2 [3 hidden]5 mins ago
We already have plenty of those, including some that use the exact same SoC.
snvzz [3 hidden]5 mins ago
But not in Framework laptop form factor. Until now, that is.
ThinkBeat [3 hidden]5 mins ago
Cool.
Now make a mainboard with a POWER 10 (IBM)
and
something like the Altra Max M128
Would probably need to create a "hood", extender between the case
and they keyboard to fit. and increase the fans a it </s>
Sadly POWER10 is not entirely open and the reasonning behind having a power laptop/computer was to have an entirely open computer
inkyoto [3 hidden]5 mins ago
Whilst it is true indeed that POWER10 requires an opaque blob of firmware goo, Raptor was supposed to bring a follow-up to POWER9 that uses the Solid Silicon S1[0] instead of POWER10:
Raptor yesterday officially announced that we're not getting Power10 systems. The idea is we're going to be getting something better: the Solid Silicon S1. It's Power ISA 3.1 and fully compatible, but it's also a fully blob-free OpenPOWER successor to the POWER9, avoiding Power10's notorious binary firmware requirement for OMI RAM and I/O.
[…] the S1's specs, and he said it's a PCIe 5.0 DDR5 part running from the high 3GHz to low 4GHz clock range, with the exact frequency range to be determined. (OMI-based RAM not required!) The S1 is bi-endian, SMT-4 and will support at least two sockets with an 18-core option confirmed for certain and others to be evaluated. This compares very well with the Power10, which is also PCIe 5.0, also available as SMT-4 (though it has an SMT-8 option), and also clocks somewhere between 3.5GHz and 4GHz.
I have not heard of Solid Silicon before, and their website[1] does not reveal much other than indirectly ascertaining the non-laptop factor of the S1 SoC(?).
I would legit love to have a POWER10 laptop. No reason it couldn't be scaled down to laptop power profiles, though I have no idea who would provide a GPU for it.
6SixTy [3 hidden]5 mins ago
There's Verisilicon, Imagination, Broadcom, and probably even ARM licensing out GPU IP blocks to anyone with capital.
Getting AMD on board could be a little trickier but could be feasible with a compute SOC deal balancing the deal out (this is a pipe dream).
sylware [3 hidden]5 mins ago
I don't think I can buy this in my country at a reasonable price.
It is very hard to pay online for me as amazon broke interop with noscript/basic (x)html browsers a few years ago (and you can pay using local wallet codes with using your credit card info online).
mort96 [3 hidden]5 mins ago
It's not difficult to use Firefox. You just don't want to.
HN.zip
Times are changing
They seem to come with Windows ARM pre-installed, but if one wanted to (also?) run Linux on them, which distro(es) would be available?
1. https://discourse.ubuntu.com/t/ubuntu-24-10-concept-snapdrag...
Overall though, things "work" but as far as usable goes, it seems to remain in a pretty poor state; especially given the lack of accelerated graphics combined with no NUC/SFF options. I guess WSL2 is probably okay. But it's a shame, because the hardware does seem pretty nice.
When some NUC options come out you'll at least be able to use it in a headless form factor, I guess.
You were - and are - wrong. Qualcomm mainlined enough drivers for in the same window (past 6 months) so one can boot and install the latest Linux kernel on several Snapdragon X laptops from different OEMs. There are people running Fedora and an Ubuntu Experimental[1] builds. Its still rough, depending on ones device tree, the peripherals line microphone, webcam, and/or fan control may not work right now, but it's a platform that's actively being developed for.
1. https://discourse.ubuntu.com/t/ubuntu-24-10-concept-snapdrag...
First of all, the post wasn't downvoted, and secondly, someone pointed out that it's being worked on, as another commented pointed out here as well. I hope being corrected twice is enough to prevent repetition.
No, NetBSD is not working on Qualcomm based laptops. I also know quite a few arm64 distributions ... like LineageOS, which works OK-ish on my phone.
This thread however is about laptops in a sense one expects from the laptop, i.e. hardware accelerated video, WiFi 5 and above, bluetooth, woking sleep/resume.
I don't understand what exactly is strange here for you.
I must live in a bubble.
It's never crashed, I never reboot it, sleep works fine. It spends almost all its time sitting with the lid closed, sshing in from my Mac, using 13W-15W depending on whether it's in eco/smart/performance mode, then boosting to 160W or so if I ask it to do something serious.
When I do take it mobile, it runs for about 5 hours doing light stuff in eco mode, vs 6 hours when it has Windows on it. The difference isn't really a problem as it disappears if you give it some real work to do. That's not Arm Mac battery life, but it's 2 or 3 hours more than I ever actually need. I've got it set up to keep running closed on AC, sleep if the AC disappears. Works great, and it's going instantly from sleep with a keypress if I open it.
Really, no complaints at all.
Want a different board on a soc that already works? Welcome to the world of writing this kind of things: https://github.com/torvalds/linux/blob/master/arch/arm64/boo... and compiling the kernel from some dude's branch instead of going with mainline.
I have the risc thing too, that magically works on mainline kernel somehow, but is slow.
I've used mine as a desktop for a while, but now it's been repurposed as home server.
It runs Debian trixie (still not frozen, just sid + small delay), with a generic kernel (not vendor crap) and root on ZFS, a 1TB M.2 SSD, with an external 22TB disk, also ZFS, on UASP USB3 enclosure.
Compared to the rpi4 it replaces, it is somehow actually faster. Sustains 70MB/s aes decrypt (luks), where the rpi4 gave me <50MB/s.
Graphics support is indeed early and tied to an ancient vendor kernel, but Imagination Technologies is funding the mesa3d effort. It will eventually (hopefully soon, there's been hints recently) simply just work with open drivers.
0. https://buildd.debian.org/stats/graph-week-big.png
[1] https://www.reddit.com/r/hardware/comments/1efsvcj/lenovo_cl...
Also, word of warning just don't apply the 3.08 bios update if you have a 12th gen. I've had nothing but problems with the 3.08 bios update without real remediation from framework. Really, don't perform this upgrade without doing a thorough review of the support forums regarding the issues people are having.
There was a blog post about the BIOS update that i found here on this site, but i'm having trouble finding it for reference now. Someone had the same issue as i have where sometimes my system just won't boot (no video and error sequence) unless i remove or re-add a memory module. Not that fun.
it's rare enough to scare me for a moment each time that it has died on me...
It took about 4 years to die, and served me from 2009 - 2024, so still quite a long life.
Crappy BIOS/power management controller does not depend on price of the laptop or manufacturer.
https://www.starfivetech.com/en/site/soc
I'm guessing the sales numbers will be extremely small—some people who get excited about RISC-V and would consider buying this don't know how slow it is in comparison to even the cheapest Arm SBCs right now.
I own a VisionFive 2 with this SoC.
CPU sits between rpi3 and rpi4 performance.
GPU is easily 3x that of rpi4. I/O speeds are similar to rpi5.
This is not anywhere as slow as you're making it to be. And certainly it isn't slower than the cheapest ARM SBCs.
But yes, you can get a much faster rpi5 and it is cheaper.
Current RISC-V SBCs and laptops are for enthusiasts and developers, who know full well what they're getting.
Next iteration will be faster. The next, faster still.
And they have a clear disclaimer it's a dev-focused board; if you expect a "proper laptop experience" then that's on you the basic description of the product.
"Targeted towards developers, tinkerers, and early adopters, this platform allows exploration and development within the RISC-V ecosystem. [..] This is very much a developer-focused board to help accelerate maturing the software ecosystem around RISC-V, so we recommend waiting for future RISC-V products if you’re looking for a consumer-ready experience."
[1]: https://youtu.be/iMwepyyaj8I?si=AFGEK6He-VWlKQQ9
> There is excellent philosophical alignment between RISC-V and Framework. Both are built on the idea that an open ecosystem is more powerful than the sum of its parts... x86 ... is used in the processors for each Framework Laptop ... The x86 ISA was invented by Intel, extended on by AMD, and is proprietary, with Intel and AMD being effectively the only two companies able to use and create processors around it. ... What makes RISC-V unique is that it is an entirely open architecture, which means that anyone can extend on it and create their own processors that use it without paying a fee. RISC-V International is the collaborative organization that exists to help develop the standard and define common versions to ensure cross-compatibility of hardware and software. There are hundreds of companies now developing cores and chips around RISC-V, but most of these have been hidden away in embedded applications....
> All of this is what makes RISC-V unique from an ecosystem enablement perspective. The actual technology is equally interesting. The base instruction set of RISC-V is simple and streamlined, while there are a number of extensions enabling high performance and specialized compute. This means that RISC-V cores can be developed for anything from tiny control CPUs embedded inside a sensor (the Fingerprint Reader we’ve used in Framework Laptops since 2021 actually has a RISC-V core!) to monstrous multi-hundred-core server processors. The DeepComputing RISC-V Mainboard uses a JH7110 processor from StarFive which has four U74 RISC-V cores from SiFive. SiFive is the company that developed CPU cores using the RISC-V ISA, StarFive is the processor designer that integrated those CPU cores with other peripherals, DeepComputing created a Mainboard leveraging that processor, and Framework makes laptops that can use the Mainboard. The power of an open ecosystem!
[1] https://frame.work/ca/en/blog/introducing-a-new-risc-v-mainb...
[1]: https://en.wikipedia.org/wiki/Apollo_1 [2]: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/197200...
and
something like the Altra Max M128
Would probably need to create a "hood", extender between the case and they keyboard to fit. and increase the fans a it </s>
Sadly POWER10 is not entirely open and the reasonning behind having a power laptop/computer was to have an entirely open computer
[0] https://www.talospace.com/2023/10/the-next-raptor-openpower-...
[1] https://www.solidsilicon.com/desktop/index.html
Getting AMD on board could be a little trickier but could be feasible with a compute SOC deal balancing the deal out (this is a pipe dream).
It is very hard to pay online for me as amazon broke interop with noscript/basic (x)html browsers a few years ago (and you can pay using local wallet codes with using your credit card info online).