[forevernoob]

> ...the first Framework Laptop 12 motherboard is going to use Intel's 13th-generation Core i3 and i5 processors

I _really_ hope they launch an AMD version (perhaps with an iGPU) soon after that. That and preferably with Libreboot support. This would make it the ideal portable laptop for me and thus I'd be able to (finally!) replace my X220T.

[justahuman74]

Libreboot would be amazing progress for moving open-source downwards in the stack.

It feels odd having a load of closed source EFI stuff and then putting linux on top of it. Sure linux injects a bunch of firmware into hardware later in the boot, but it's still progress

[MisterKent]

I am a fan of open source and being able to tinker etc. But I've never felt the need (advantage?) to do more than just use the bios/efi to boot or configure a few basics.

I've been burned by a small SBC that had poor support, but on laptops/desktops never felt limited.

But people always sound so excited to libreboot their personal computer... Am I missing out or is it just nerd cred?

[panick21_]

There is security.

The other things is bug-fixing. The way it currently works is that bugs often only get fixed for one version and other issues like that.

Getting this stuff open and having upstream heavy fixing of bugs will imporve the ecosystem.

The other things is performance. Faster boot times are quite nice.

The other things is being able to better play with more advanced boot security concepts, booting with Bluetooth/NFC verification and so on.

And once its open, and you can more easily work with it, and its more available, hopefully more people do more interesting work with it.

That said, its not an absolute priority for me, but its just a better long term solution. So companies that push it are a big plus.

[forevernoob]

I'd say that after numerous revelations in regards to UEFI vulnerabilities and such, an open source BIOS / EFI has become a necessity for me rather than something just nice to have.

[p_l]

A lot of those revelations are due to majority of EFI code other than vendor specific drivers being quite open source unlike the bad old times.

Still would like an end to end open source UEFI bootchain, most groups end when they hit the 60% of stuff they need for their limited use cases : - |

[brunoqc]

Why would you prefer AMD? price, heat/fan noise?

[starkparker]

Framework shipped AMD 7040-series and 13th-gen Core i-series alongside each other for the 13.

The 13th-gen Intels had miserable battery life and heat issues under load. If you could manage that, all four USB-C ports were full Thunderbolt ports equally capable of driving displays, PD, and USB 4 throughput.

The AMD line had considerably better performance-per-watt but rougher firmware support (and early on, really broken Linux kernel support that required Fedora or other rolling kernel release distros). It also couldn't deliver the same "every port does everything" promise that the Intel boards did, with some ports not supporting displays or USB 4, which significantly reduced the value of the expansion-card model to kind of a novelty.

On the 12, if it's likely also going to have a smaller batter than the 13, going only with 13th-gen Intels means it likely will be either a further step back in battery life vs. the 13 or throttled to extend the battery.

[phonon]

The two CPU models are i3-1215U and i5-1334U [1] which are 15 W parts with 2 P cores. They should be OK.

[1] https://youtu.be/-lErGZZgUbY?t=729

[starkparker]

The issue isn't TDP so much as performance-per-watt. The equivalent Ryzen 5 PRO 7540U can run at the same base TDP, had 2 full and four 4c cores on a smaller die, and outperformed the 1334U almost across the board.

Both chips were Q1 '23, so the timing's not a great excuse. They were in HP's 2023 EliteBook G10 840 (1334U) and 845 (Ryzen 5 PRO 7540) laptops, and the 845 was better on both single- and especially multi-thread Cinebench R23, _much_ better in GPU loads and gaming, _and also_ lasted longer on battery.

I think Framework mostly just wants to target an education market with a mainboard experience that's lower maintenance than AMD has been for them. Fewer USB-C restrictions, less firmware drama.

Still hard to get excited about it being the _only_ available option, though.

[phonon]

I would suppose they got a good deal on them, as they're a little out of date, but they're good enough, and overall provide a good package. If you're trying to hit a price point, there will always be compromises.

[aljgz]

I don't know about the GP. I won't buy anything from Intel unless things change dramatically. My last Intel laptop had serious thermal throttling problem that could be completely avoided if Intel cared a bit about users. The one before had some other problems. In past 20 years, anytime I bought (or was given by a company) AMD I was happy, and as time goes by I get less and less happy with Intel.

[genewitch]

Thermal issues on a laptop I'd be pulling heatsinks and putting better paste on. Also a laptop cooler is nice unless you get junebugs.

[omnimus]

Framework also mentioned they use PTM7950 in everything from now so repasting wont help if they didnt mess up assembly.

[iamtheworstdev]

i don't know about other people's experience - but my framework with intel cpu is always running that fan relatively maxed out even when it's not doing anything. And it has massive issues staying asleep which is some sort of driver issue with Windows. But I can be an airport and all of a sudden my backpack feels like it's about to combust and i can hear my laptop fan rippin', even though it should be asleep.

[bityard]

That is by design, it's a feature from Intel and Microsoft called "modern standby." It basically means your laptop cannot actually sleep anymore. Instead, it enters a slightly-lower-power mode so that it can download emails in the background, run Windows updates in the middle of the night, and generally pretend to be a phone even though no one wants that and the hardware/os was not really designed for it.

[p_l]

Modern Standby requires that deepest sleep states take less power than S3 standby, the issue is mainly drivers and applications either waking the system or not putting peripherals into deep enough sleep (or shutting them down entirely if possible)

[WD-42]

Some new laptops cannot actually fully suspend to ram. It sounds crazy but I had this issue even after installing Linux on a laptop. It’s a hardware limitation. You can thank Microsoft for trying to make sure they can sneak in OS updates when you think your laptop is asleep.

[p_l]

It's not sneaking in OS updates, it's faster wake-up mainly.

Combined with properly prepared HW taking less power in modern standby than in S3, but it requires hardware/firmware/driver/userland confluence to work well

[pmontra]

By the way, how do people carry those laptops on airplanes if regulations mandate that every radio equipment must be off?

[wkat4242]

They don't. Only cellular connectivity. Most airlines even offer WiFi now

[jazzyjackson]

Mandate is a strong word.

[p_l]

Airplane mode is still a thing.

[amatic]

I'm not sure I had the same issue, it is a different laptop, but it would not go to sleep. The current solution is to make it hybernate instead of sleep.

[forevernoob]

Considering Intel's track record on hardware vulnerabilities, I'd much rather prefer AMD.

[zeroq]

just an anecdata but recently I was building a HTPC/NAS. Initially I wanted N100 for pure NAS, but ended up with an 5500 AMD and I was blown away by the capabilities of IGPU. Turned out to be a quite capable gaming machine.

[moltopoco]

And a 2-in-1 laptop in tent mode would be perfect for (casual) gaming on the train with a gamepad; much more ergonomic than holding one of the many heavy gaming handhelds.