Framework Desktop
the RAM isn’t replaceable
Absolute deal breaker.
target use case is a server for LLMs where memory bandwidth is very important iirc, not really a desktop for the consumer. this is probably a one off product that was pitched to them by amd
Framework desktop? Why?
Looking at the form factor, it seems to mostly address typical desktop use without having a full blown workstation with a big tower, which one usually builds and upgrades by oneself.
I’m also not sure, if this niche is the right choice for Framework though…
Edit: the more read about it, the less I see the appeal as well…
Maybe we just aren’t the target audience, but I can’t really figure out, who they want to targetYep. Just read the thing. The RAM is soldered. THE RAM IS SOLDERED (the CPU seems to be as well but I’m not sure).
I don’t think there is a target audience. Seems like someone in the boardroom took a big hit of a giant blunt and went “what if took, like, the laptop design philosophy and, like, apply it to a desktop, but like, modular duuuuuude?”
Someone 3d print a case that allows a bigger cooler so we can have a comp that looks like this: https://thumbs.dreamstime.com/z/exterior-view-drag-racing-car-super-charger-air-intake-classic-car-hood-super-charger-159302695.jpg
I support using an Augustiner Maß for size comparsion.
Really tempting but you can get such good computers second hand these days. I got a Ryzen 9 3950X (a few years old but 16 core and still awesome), with 128 GB of RAM and a 1TB SSD for £325. No way I’m paying 6 times that for a new machine that’s 50% faster at best.
For work, we got some of the HP AI Ryzen Max 390 laptops with 64 gigs of soldered on 8000 MHz DDR5 memory, and holy shit, I have never felt such a snappy, responsive computer.
Like the comparison between my laptop with the 13950X and 64 gigs of memory at 4800 megahertz and this Ryzen ai 390 feels very much like when we first started getting SSDs and making the transition from spinning rust to SSD.
And I know that a huge portion of that is due to the fact that the ram is twice as fast. But still, it is simply snappy. It’s nice, and it makes me jealous that these are not computers for me, but for someone else.
You won’t get the AI stuff through. I’m working on a machine like you described and drilled a hole in it to provide power to a 3070 as the power supply wasn’t beefy enough (and server power supplies are too expensive compared to a hole).
This is the first generation where I can play around with local LLMs from what I can tell - even used the hardware for that is way more expensive :(
What’s that panel two under Tux, is that so you don’t have to go all the way outside to touch grass?
I’m just glad to see FW increasing the visibility of SFF form factors
I also really like Small Form Factor form factors
(sorry couldn’t resist)
Is the kind of hardware for AI stabilizing by now? or is it still highly experimental?
Data center? Nvidia, custom TPUs, or GTFO.
Home or small business? All over the map, with different hardware excelling at different use cases.
I have the Asus Z13 Flow 128gb variant on order and “shipped”, to be delivered in a couple weeks. I was on a wait list for almost 4 months to get this, and now with new tariff news, I’m very glad I could get one when I could.
One question I have for this article, which I doubt I’ll see, is “In a desktop form factor, why cant I use DDR5 in addition to the soldered ram?” . They offered only a few words to discussion around the board, but it is really bugging me. Why couldn’t’ we have the best of both worlds with this?
The review is very performance focused, which is… fine I suppose. But it doesn’t cover the most relevant comparison, which would be the Asus versus framework comparison.
Also, its good to see these reviews including most important benchmarks: Training and Inference. But those benchmarks need to be scaled across model sizes, and then also need to come with some kind of accuracy and precision metrics. They’re using things like tokens per second as the metric, or training times, but realistically, consumers have to make a choice about the size/ price of a GPU and what models it will be able to hold. Larger models, likely will perform better, but they come with significant cost especially dependent upon how much can sit in vram versus system ram.
Its a good review but I think we need to see more focus on more relevant benchmarks for users. Tokens/ second is important, but like, train ImageNet and give me model statistics and performance. Like, do more please. Load 5 models (a 4gb, a 8gb, a 16gb a 32gb a 64gb and a 96gb+ model), give them all the same prompts, and display the results.
If I understand right they actually tried to work with AMD to implement an experimental new socket standard or something to get the low latency they needed, but weren’t able to get the results they were after
This device isn’t really for me, but the amount of RAM it comes with should last a very long time, and the fact that it’s soldered unfortunately is important to its unique performance
If I were to hazard a guess, they’re reusing a mobile board design for this somewhat and at least in mobile applications socketed dimms draw 30-50% more juice than soldered. It could also be the npu or gpu requiring the 5-10% extra memory bandwidth they get from being soldered. I do agree that I don’t think it was worth the trade offs from a consumer perspective, but framework seems to generally make good choices so I’m thinking there must’ve been some outside pressure at play affecting the financials of the project or something.
When I’ve seen this come up before, I think the AI Max needs soldered RAM due to latency that socketed RAM introduces. There were some older post where an AMD engineer even hopped in saying they tested but it had a big impact on performance.
First time in hearing this, so thank you!
Soldered RAM is really faster then socketed RAM?
I can’t really see why, when only looking at the connection level.But soldered RAM also means, they only need to support exactly this RAM.
So they can leave out some stuff handing different speeds and can optimise for exactly this RAM.Is that the reason?
I’ve actually never thought about that until now…
