Live data from Hacker News

Framework's first desktop is a strange–but unique–mini ITX gaming PC

arstechnica.com

411–420 of 554 posts

Re: Framework's first desktop is a strange–but unique–mini ITX gaming PC

#411

Earlier quoted context omitted.

Nvidia project Digits seems to fall in a similar category, no?

It's a bit different. Digits is based on the Tegra CPU, which is an ARM chip with integrated nvidia GPU. It's nearly COTS (commercial off the shelf), but not quite. Tegra CPU support isn't in mainline linux, so you have to run their fork of Ubuntu or build your own kernel. The integrated GPU is a special class in nvidia drivers, and some things just don't work on it (they only work on a discrete GPU) for seemingly no…

Keep in mind the digits is part of their server line and server OS, not the random embedded dev kits often used for developers targeting car entertainment systems and often with abandoned kernel + driver.

The nvidia digits uses DGOS, same as their grace+hopper and similar enterprise/cloud products.

Re: Framework's first desktop is a strange–but unique–mini ITX gaming PC

#413

The Ryzen APU could be interesting for running local AI models and with 128 GB of RAM you can fit quite a large model. Plus it should be relatively energy efficient compared to a full size desktop build with separate GPU. Lack of PCIe 5.0 is a bit of a bummer as you could otherwise plug in some new Samsung 9100 Pro NVMe drives. Would love to see how it performs. It supposedly has a memory bandwidth of 256 GB/sec whic…

They should work just fine, sure peak bandwidth will be limited, but most workloads are not bottlenecked by I/O bandwidth on PCIe 4, but are limited by IOPs which should be decent on the 9100.

Re: Framework's first desktop is a strange–but unique–mini ITX gaming PC

#414

Earlier quoted context omitted.

Laptop chips often only have x8.

Looks like the Ryzen AI Max chips do have x16: https://www.amd.com/en/products/processors/laptop/ryzen/ai-3... But you probably want M.2 slots over a single x16 slot.

They could have added a 16 lane upstream, 32 lane downstream PLX switch to packet switch between those devices.

Re: Framework's first desktop is a strange–but unique–mini ITX gaming PC

#415
Bit of an odd one really, they've introduced modularity and customisation to the laptop market where there wasn't any.

But they have now introduced reduced modularity and customisation in comparison with what's available in the desktop market. CPU is soldered, RAM Is soldered and the GPU is soldered.

Re: Framework's first desktop is a strange–but unique–mini ITX gaming PC

#416

Earlier quoted context omitted.

That sounds misleading. The parent comment says all deliveries in Asia are limited to Taiwan. As far as I'm aware, export restrictions are only placed on certain chips going to China. How do export restrictions prevent Framework from shipping to i.e Japan?

In the case of Japan, it's likely an issue with Japanese regulations. We have a very strict radio law that applies to anything that can produce radio or em waves That includes motherboards, since they can technically emit on those frequencies. From what I have seen, the framework laptop motherboards appear to abide by that law. However, I assume it's just expensive to figure out in the first place.

LTT video specifically states that they were also able to get the Desktop model to be compliant with the (US, I suppose) EMI laws. He also shows the shielded side of the case (only one side is shielded and that was supposedly enough to pass).

Re: Framework's first desktop is a strange–but unique–mini ITX gaming PC

#417

Earlier quoted context omitted.

That is not a framework choice, that is an AMD architecture that doesn't use regular RAM modules as it requires wider data bus

The statement above is that one would expect Framework to have chosen a platform which does not require soldered RAM, not that Framework kept such an option by choosing this AMD part. All that aside, I absolutely can't wait for desktops to decide to go the same route of having 4 memory channels instead of 2. Right now the only way to have >2 channels is to buy workstation/server class stuff or an APU.

From their LTT video, it looked like they chose to do Desktop because of the AMD platform (and not the either way around, where they planned a desktop product and then chose AMD strix halo). Apparently setting up the manufacturing pipeline for laptops built on strix halo is expensive as of now and there are only two laptops in the whole market using it. So Frame.work choose to go the desktop route to save on cost while still making the platform available for everyone.

Re: Framework's first desktop is a strange–but unique–mini ITX gaming PC

#419
post #156

Earlier quoted context omitted.

Still no haptic trackpad!

Why would you want haptic trackpads? Having used modern Macbook trackpads they feel like a massive downgrade compared to either of my Frameworks. The vibration-based simulation of haptics feels uncanny and unsatisfying compared to the real deal.

Strongly disagree. MacBook haptic trackpads feel plenty natural to me and don’t suffer the weird inconsistency issues that plague traditional trackpads. I say this even as someone who uses a machine with a traditional trackpad everyday and has a new FW13 reserved. If Framework ever offered a haptic trackpad upgrade I’d buy it.

Re: Framework's first desktop is a strange–but unique–mini ITX gaming PC

#420
Cool but this shows off the absolute pointlessness of the Framework concept - when you have RAM, CPU and GPU integrated, you essentially have something as modular as a modern Mac, which goes against any idea of modularity.

This has been the future for more than a decade, evidenced by stuff like smartphones and games consoles, and as far as I know, Nvidia is also working on their SoC, and so does Intel (which I think was the whole point of their gaming GPU push).

Post reply on HN