In case you’re wondering how much faster a Framework Laptop 12 is than your older generation XPS or ThinkPad, Phoronix has the benchmarks!

If you want to scale up your ability to run giant models locally, the team at @AIatAMD has published a step by step guide to clustering Framework Desktops:

@openclaw It wouldn't be any dumber than what we're currently tweeting.

Should we give @openclaw running locally on a Framework Desktop control of our Framework twitter account?

If you missed our last YouTube livestream Q&A, no worries — join us for the next one! This Friday, February 27 at 2PM PT, our Director of Software, Kieran, will sit down with Founder & CEO Nirav to talk all things firmware and software at Framework. Have a question you’d like

Interested in making Windows even less secure? You can run @openclaw locally on Framework Desktop using @lmstudio and @AIatAMD and give it full control over your life!

This may look like a 2006 MacBook Pro, but it's actually secretly a Framework Laptop inside! Incredible project by edoogg: https://t.co/khTMIuTH5m

In this video I introduce the new Strix Halo toolbox for Image and Video generation with ComfyUI, supporting SOTA open models like Qwen Image WAN 2.2 and the new LTX2: this is finally stable and performance has been improving. @FrameworkPuter @AIatAMD https://t.co/Ig1izizM18

Prototyping of the Framework Laptop 16 SDR is in full swing 🙌

We’re going live now on the Memory Q&A:

We're also hosting a livestream Q&A on memory and overall silicon pricing on the Framework YouTube channel tomorrow (Feb 12th) at 10am Pacific:

We have more detail in our blog post here:

We've been able to hold our Storage and pre-built system pricing for this month, but expect that we'll need to increase these in the future to handle cost increases. Currently, some of our SSD prices like 8TB are substantially below market pricing.

Both DDR5 and LPDDR5x costs continue to increase, and we've had to update pricing again today. As before, we are setting DDR5 prices as close as possible to our purchase price, and absorbing some of the increase on LPDDR5x.

No purchase required. You can enter the giveaway here:

We're breaking into the rainy day fund for you! Because RAM prices are crazy right now, we're doing a giveaway of 10x DDR4 and 10x DDR5 SO-DIMMs (literally from the bin of random modules in our office.)

https://t.co/NbUIjE7Mn6 I built a 2-node Strix Halo/@FrameworkPuter Desktop setup using RDMA (RoCE/Intel E810) and vLLM tensor parallelism. This took a fair amount of digging into ROCm and RCCL to get low-latency multi-node inference working on gfx1151.

We're building a USB 3.0 dual expansion card into DockFrame (modular usb-c hub compatible with @FrameworkPuter expansion cards), 5 Gbps per port on a compact 4-layer PCB. Read more: https://t.co/3wohzlnUEU #framework #opensource #openhardware #dockframe

Thanks to JLCPCB, sourcing a copper core PCB is actually cheaper and faster in low quantity than a plain copper sheet metal with a solderable surface finish😄

I tried to make Jeff Geerling's @FrameworkPuter mini cluster faster. (new video)