[flagged]
Oh no, someone’s an optimist… get him!
The Framework Desktop is a beast
281–290 of 464 posts
Re: The Framework Desktop is a beast
#282Earlier quoted context omitted.
Lenovo is a Chinese company.
I would rather say it's a Chinese conglomerate. E.g. Think.* are said to be designed in the US, Taiwan, and Japan. They are often assembled locally (e.g. my ThinkStation was assembled in Hungary). They have offices in many local markets and can be held accountable for warranty, issues, etc. I agree with your general point though, 'Chinese' does not mean bad quality, warranty, etc. It's more a property of a bunch of C…
> As an aside, Lenovo are pretty awesome.
These things would be important if they made products that actually work. My T14s Gen 3 AMD simply doesn’t work. Half the time I go to wake it from sleep, the firmware has crashed and I have to hard-reset it. I spent months trying to get Lenovo to fix this. They did replace the motherboard twice (once on-site, once shipped to them) and eventually replaced the entire laptop with a new one. None of this is useful when they can’t make a laptop that doesn’t crash while it sleeps.
Re: The Framework Desktop is a beast
#283Earlier quoted context omitted.
> the Remote Device Management baked into firmware as Apple does with its hardware? What do you mean? Linux had SSH (and before that rlogin) for a very long time already.
"Device Management" means "hardware-level backdoor."
Only if it's your hardware. If it's corporate hardware, it's their hardware and you're a guest.
Re: The Framework Desktop is a beast
#284Earlier quoted context omitted.
The M3 Ultra also consumes a lot more power because Strix Halo is actually a chip for laptops. The Framework here is a desktop but that doesn't change what AMD prioritized when building this chip.
The M3 Ultra is more efficient both on the CPU and GPU side.
I'd hope so since M3 Ultra is using a newer/more efficient manufacturing process than Strix Halo!
Re: The Framework Desktop is a beast
#285Earlier quoted context omitted.
Apple devices support MDM. When you purchase the device, the device's firmware is configured to check in with an account owner. The firmware has an integrity feature such that this configuration cannot be removed by a user: https://it-training.apple.com/tutorials/deployment/dm005/ If OP just meant remote management through a BMC then that's not common except for server hardware, and it would have features like Redfis…
I don't think it's fair to describe MDM as a firmware-level feature. I think it's entirely implemented and enforced within the environment of a booted macOS; the firmware isn't going to be bringing up a whole network stack to phone home. If you had Linux on a MDM-enrolled Mac there wouldn't be anything MDM-related running during or after the boot process. But presumably any sane MDM config would prevent the end user…
Yeah, your point about implementation is correct -- much of the MDM functionality runs within macOS.
But, eh, I still think it's fair to describe it as a feature of the firmware. The enrollment and prevention of removal have firmware-level components through Apple's Secure Boot and System Integrity Protection. A user can't simply disable MDM because these firmware-level protections prevent tampering with the enrollment.
Case in point, getting Linux installed in the first place would be blocked by firmware-level boot policies, right? I'm not too knowledge about this, and maybe you are more so.
Re: The Framework Desktop is a beast
#286> The Framework Desktop with 64GB RAM + 2TB NVMe is $1,876. To get a Mac Studio with similar specs [...] you'll literally spend nearly twice as much [...] The Framework Desktop is simply a great deal. Wow, someone managed to beat Apple on price?? I don't know that it logically follows that anything is a great deal when it undercuts Apple. Half sounds about right -- I thought Apple was a bit more competitive these day…
He's comparing to a studio when he should compare to the mini for this performance. They are almost the same price at a 64gb RAM + 500gb storage config (CAD).
- Framework, Max+ 64GB: $2,861.16
- Apple Mini M4 Pro, 64GB: $2,899.00
Apple does charge way too much for integrated storage, but Apple is only a 25% premium at 2TB not double (if you compare to the mini instead of the studio). Plug in a NVMe SSD if you're building an Apple desktop.
Re: The Framework Desktop is a beast
#287Earlier quoted context omitted.
Actually the pricing is pretty similar. Framework Desktop price with default selections, 32GB of RAM, 500 GB storage: $1,242.00 USD Mac Mini with 32GB of RAM, 512 GB storage: $1,199.00 Post changed a bit since I started replying, so: > For the purposes of running LLM models, a Mac Mini The M4 Max is the one that actually gives you a shit load of memory bandwidth. If you just get a normal M4 it's not going to be espec…
> Mac Mini with 32GB of RAM, 512 GB storage: $1,199.00 You're looking at the wrong Mac Mini. The model with the M4 Pro is the right comparison, on account of also having a 256-bit memory bus giving substantially higher bandwidth than a typical desktop computer. The M4 Pro model doesn't have a 32GB option. The M4 Max (not available in a Mac Mini) has an even larger memory bus, giving it far more bandwidth than either…
I was just going for a head-to-head comparison, that's the closest you can get in price/performance. The closest M4 Pro Mac Mini is already a lot more expensive than the baseline Framework Desktop.
The Framework Desktop Max+ 395 with 128 GB of RAM, and a 500 GB SSD costs around $2,147.00 USD before tax. The M4 Pro with the 20-core GPU, 64 GB of RAM, and a 512 GB SSD costs around $2,199.00 USD. That's still short 64 GB of RAM, of course.*
The lowest-end M4 Max Mac Studio that can support 128 GB of RAM seems to cost $3,499.00 with 128 GB of RAM and a 512 GB SSD. For that you get 546GB/s of maximum memory bandwidth according to Apple, which is definitely a step up from the 256GB/s maximum for the Ryzen AI Max+ 395, but obviously also at a price that is quite higher too.
Apparently though, 128 GB of RAM is currently the ceiling for the M4 Max right now. So it seems like if you were going for a maximum performance local AI cluster at any price, the M3 Ultra Mac Studios are definitely in the running, though at that point it probably is starting to get to the price where AMD and NVIDIA's data center GPUs start to enter the picture, and AMD Instinct cards measure memory bandwidth in terabytes per second.
* Regarding GPUs: The Framework Desktop Max+ 395 Radeon 8060S seems to be vastly faster than all of the non-Max M4 SKUs, for anyone that cares a lot about GPU performance. The M4 Max seems to outperform the 8060S by a bit though, and obviously it has some stand-out features like a shit load of video encoding/decoding hardware. This complicates the value comparison a lot. The Radeon core definitely gets a much better value for the performance in any case. I'm really impressed by what they managed to do there.
Re: The Framework Desktop is a beast
#288What are more budget friendly options for similar workloads (running web app test suite in parallel)? My test suite currently runs in ~6 seconds on 9700K. Would be nice to speed it up, but maybe not for $2000 :-) Last I checked 13700K or 13900K looked like the price/performance sweet spot, but perhaps there are better options?
>My test suite currently runs in ~6 seconds on 9700K Absolutely nothing. 6 seconds is about the time it will take you to tab to your terminal, press up arrow, find your test task and run it. There's no amount of money that makes it go from 6 to 3, and no world in which there's any value to it. In addition, upgrading to a 13900K means you're playing the Intel Dance: sockets have (again) changed, in an (again) incompat…
6 seconds is the time it takes for the tests to run, after I've switched to the terminal and ran the command. If I switch from 8 cores to, say, 16 faster cores, IMHO it is not unthinkable the tests could speed up to 3 seconds. How much money to invest for this speedup is a subjective question.
I'm thinking about a new system, not upgrading the existing one.
Re: The Framework Desktop is a beast
#289> The Framework Desktop with 64GB RAM + 2TB NVMe is $1,876. To get a Mac Studio with similar specs [...] you'll literally spend nearly twice as much [...] The Framework Desktop is simply a great deal. Wow, someone managed to beat Apple on price?? I don't know that it logically follows that anything is a great deal when it undercuts Apple. Half sounds about right -- I thought Apple was a bit more competitive these day…
The point is that before the AMD Ryzen Al Max+ 395 chip there was only Apple that offered something comparable for the desktop / laptop that could do these AI relate tasks. Were else could you find a GPU with 64-128 memory?
Re: The Framework Desktop is a beast
#290Earlier quoted context omitted.
> We need a more expensive CPU, more expensive motherboard, more power, more cooling, and a larger system. Congratulations you've reinvented threadripper (4 channel), siena (6 channel), Threadripper pro (8 channel), or epyc (12 channel). This is the real story not the conspiracy-tinged market segmentation one. Which is silly because at levels where high-end consumer/enthusiast Ryzen (say, 9950 X3D) and lowest-end Thr…
Well sort of. Apple makes a competitive mac mini and macbook air with a 128 bit memory interface, decent design, solid build, nice materials, etc starting at $1k. PC laptops can match nearly any aspect, but rarely match the quality of the build, keyboard, trackpad, display, aluminum chassis, etc. However Apple will let you upgrade to the pro (double the bandwidth), max (4x the bandwidth), and ultra (8x the bandwidth)…
> Does mystify me that x86 has a hard time matching even a mac mini pro on bandwidth, let alone the models with 2x or 4x the memory bandwidth.
The market dynamics are pretty clear. Having that much memory bandwidth only makes sense if you're going to provide an integrated GPU that can use that bandwidth; CPU-based laptop/desktop workloads that bandwidth-hungry are too rare. The PC market has long been relying on discrete GPUs for any high-performance GPU configuration, and the GPU market leader is the one that doesn't make x86 CPUs.
Intel's consumer CPU product line is a confusing mess, but at the silicon level it comes down to one or two designs for laptops (a low-power and a mid-power design) that are both adequately served by a 128-bit memory bus, and one or two desktop designs with only a token iGPU. The rest of the complexity comes from binning on clock speeds and core counts, and sometimes putting the desktop CPU in a BGA package for high-power laptops.
For Intel to make a part following the Strix Halo and Apple strategy, Intel would need to add a third major category of consumer CPU silicon, using far more than twice the total die size of any of their existing consumer CPUs, to go after a niche that's pretty small and very hard for Intel to break into given the poor quality of their current GPU IP. Intel doesn't have the cash to burn pursuing something like this.
It's a bit surprising AMD actually went for it, but they were in a better position than Intel to make a part like Strix Halo from both a CPU and GPU IP perspective. But they still ended up not including their latest GPU architecture, and only went for a 256-bit bus rather than 512-bit.