I’ve pre-ordered a Framework Laptop 13 Pro (64 GB RAM, 4 TB SSD), but I’m still going back and forth in the configurator on Ultra 5 325 vs Ultra X7 358H - just want to lock in the right one. Can’t find real numbers online yet since units are barely shipping, so hoping early owners (or someone from Framework) can share rough figures.
What I do with it, roughly in order:
Everyday stuff: web browsing, 4K video
Web/app development (light, bursty work)
RAW photo editing: small sessions of ~24–80 shots (30–35 MB files)
No gaming - I have a desktop for that. But you never know…
The only reason I’d ever want the fater CPU and GPU is possibly running a local LLM (maybe)
What matters most to me: price isn’t really the deciding factor - quiet operation during normal work and low heat on the lap are, with battery a nice-to-have on top.
Questions for anyone with a unit:
Light work (browsing, coding, video) - does the fan stay silent on both? Any real difference between the 5 and the X7?
Rough package temps / fan noise under a sustained load, and does the X7 throttle much on the single fan?
Real battery life on a typical day (please note OS, brightness, and which chip)?
Linux is my target OS, so Linux numbers are ideal, but Windows helps too.
For real performance numbers on the X7 I’d keep an eye on Zack1, since he is doing a lot of work documenting that stuff. You can see his replies here: Profile - Zack1 - Framework Community
The heatsink & fan of the FW13 are pretty small compared to other laptops. From my experience on the X7 is it feels pretty toasty if you are pushing it full-tilt. I don’t particularly care because I have it hooked up to a dock with external keyboard, mouse, and 2 displays anyway. It so far hasn’t shut-down on me for getting too hot, unlike my ASUS Vivobook Pro which loved to do that. The fan noise is only noticeable when you are pushing it hard. Even then, it isn’t nearly as bad as the laptops I had before. Can’t tell you anything about the battery, because something about my board is borked and it doesn’t recognize the battery. So far just seems to be an outlier with my particular unit.
I don’t play too many games now-adays. Mainly just Age of Mythology Retold, Minecraft, and occasionally some Halo. They all run great on the X7 358H in Fedora 44 (KDE plasma desktop).
In Minecraft, I get about 80-100 FPS on maxed out settings at 1080p with the new Vulkan renderer, and about 300-400 FPS on my usual settings (Fast Preset @ 16 chunk distance). I lock the framerate to my external monitor’s 100 HZ refresh rate, and the machine is quiet and cool.
I have the X7 board; it has very conservative fan curves. For light work fan won’t even spin & even when CPU is stressed, fan is very quiet.
Temps can be high, under MT load temps reach 90C+ and X7 throttles from 35W to 32W
Throttling on GPU is less severe. Gaming at ~35W will stabilize the package at around 87C
Running ubuntu 24.04 with 7.1.1 kernel, intel graphics driver preview and lastest mesa, it’s like 8-11 hours of light browsing (50% and 100% brightness respectively) and 6-7 hours of video. Of course one would not be on light load always so I’d say it’s around 6 hours reliably.
I’m on 55wh battery so the 13 Pro would be like 9 hours reliably.
my battery is 55wh at 89% health; I think my panel brightness is also much higher than when they tested it. But that 20h video is likely “theorhetical maximum”
power saver does increase the battery life somewhat, in my observation around 15%
I think with “video playback” you will be looking at the power draw of iGPU which I believe is unaffected by power profile. I would guess the lower tier processor (ultra 5 325) would do better in that regard.
I’m saying “9 hours reliably” which means I don’t care about background tabs having dynamic elements eating battery life, I don’t care about high display brightness and I occasionally do other things when streaming a video.
If one REALLY want to optimize for putting a number on an ad, then there are so many thing they can do.
For example they would turn off bluetooth, remove all non-usb-c expansion cards, close all background tabs & processes and keep display brightness fixed that their benchmark target, and place a WIFI router directly next to the laptop with the latest standard & all the power saving features enabled. Not all videos use the same codecs so they can even pick a video with the most-power-efficient codec to decode.
I’m sure the closer you get towards 20 hours, the more noticeable any of these extreme measures will be
In reality, in normal workloads the differences wouldn’t be noticeable. Maybe a 2-5 Watt different in average power usage, but not really noticeable in heat, fan noise and performance.
However, in high workload at closer to 100% usage there is probably a 15 Watt difference in average power usage, and you will get about 1,5 better CPU performance and close to 3 times better GPU performance but at the cost of noise and heat.
With Total System power with normal screen brightness and usage I would guess Battery life:
Based on this:
Ultra 5
75 watt battery, / 7–11 average total system power = ~6.8 to 10.7 hours
X7
75 watt battery, / 9–16 average total system power = ~4.7 to 8.3 hours
For a more realistic Workday excepted battery life.
Had an opportunity to record the fan noise while downloading Red Dead Redemption 2 from steam. This is as loud as the fan in the X7 will get. Most of the time you will not hear the fan.
Most of the cores were being used during the install, but the 4 performance cores were near 100% usage the entire time. The computer is being run in performance mode on AC.
I don’t have an FL13P, but I tested on multiple laptops and desktops with Intel and AMD CPU, on Windows and Linux. I found that Steam is incredibly inefficient, you can use energy saving mode to cut consumption by more than a half with zero impact on download/installation speed.
Steam game installation is not good for evaluating a machine’s performance because the energy consumption(not steam performance) scales with the hardware, a beefy computer, higher wattage and more heat, same speed.
I bought the 2nd gen usb-a and intel be211 wifi card (not from FW). First, BE211 works fine. But then, so far I’m not seeing obvious gains in battery life. Standby drain is also similar but again I wasn’t testing rigorously.
I’ll believe they do provide some benefit in power savings and leave the testing to upcoming reviews
Also, be211 seems to have much more stable performance with bluetooth headphones than rz616. This is under same BT standard and codec (airpods, so be211 is not using newer BT standards & codecs)
Full support for BE211 is coming in mainline 7.2 I believe, so August 16th. It’s probably not worth using it before then. After, though, it should be really good.
