News Motherboards and systems with China's Loongson CPUs now shipping to US customers — options start from $373 for a DTX board with processor and cooler

[Admin]

DIY PC builders have another option from AliExpress, with two new motherboards supporting the Loongson 3A6000 System-on-a-Chip from China.

Motherboards and systems with China's Loongson CPUs now shipping to US customers — options start from $373 for a DTX board with processor and cooler : Read more

 

[Amdlova]

i3 12100 (100-130us) cheap h610 board (60-100 us) (64gb of ram 100us) chinese sh.t way more expensive
for that kit max 80us!

 

[gg83]

i3 12100 (100-130us) cheap h610 board (60-100 us) (64gb of ram 100us) chinese sh.t way more expensive
for that kit max 80us!

The title of the article doesn't make sense either. How could these make a good workstation? Maybe if you are building apps for the ChinaOS?

 

[bit_user]

The real question is whether you can build your own Linux kernel for it, or will it only boot a signed OS image?

The only way I'd even consider running something like this is if I could build the OS, with a mainline kernel, myself. I'm done running prepackaged software from China, which is why I won't buy anything higher-level than a KVM switch or unmanaged Ethernet switch from there.

 

[Blastomonas]

Not sure why anyone would want this.. It's not even that cheap considering it's likely performce level.

 

[ezst036]

It's a sad situation that we are getting ATX boards first for Loongson, but still none for either ARM or RISC-V. Only dev/beta boards.

It's like we're being punked and then getting laughed at.

I did not see in the article any mention of socket type so I assume these CPUs are soldered on.

 

[Thunder64]

DoA.

 

[bit_user]

It's a sad situation that we are getting ATX boards first for Loongson, but still none for either ARM or RISC-V. Only dev/beta boards.

Don't say there weren't ARM motherboards first:

ASRock Rack ALTRAD8UD-1L2T Review This is the Ampere Arm Motherboard You Want

In our ASRock Rack ALTRAD8UD-1L2T review, we see why this is the Ampere Arm motherboard that so many have been waiting for

www.servethehome.com

https://www.gigabyte.com/Enterprise/Server-Motherboard/MP72-HB0-rev-30
​

The first is micro-ATX, the second is EATX.

 

[Kondamin]

Not sure why anyone would want this.. It's not even that cheap considering it's likely performce level.

You might be working for a Chinese firm that has to use them or write software for the platform.

I wouldn’t get one for personal use either as it makes zero sense since you can get something much more powerful that will run a lot more software with far fewer issues.

But who knows some other gorilla might get killed because a kid fell in its enclosure and loongsoon might be the dominant arch in 5 years

 

[Pierce2623]

For $400 you can easily get a 7745HX or 12900HX on a MoDT motherboard.

 

[Pierce2623]

Don't say there weren't ARM motherboards first:

​

ASRock Rack ALTRAD8UD-1L2T Review This is the Ampere Arm Motherboard You Want

In our ASRock Rack ALTRAD8UD-1L2T review, we see why this is the Ampere Arm motherboard that so many have been waiting for

www.servethehome.com

​

https://www.gigabyte.com/Enterprise/Server-Motherboard/MP72-HB0-rev-30​

​

The first is micro-ATX, the second is EATX.

Yeah boards for Ampere Altra have been on Newegg over a year. You can buy a combo with the 64 core altra and board for like $1100 last I checked.

 

[thullet]

The real question is whether you can build your own Linux kernel for it, or will it only boot a signed OS image?

The only way I'd even consider running something like this is if I could build the OS, with a mainline kernel, myself. I'm done running prepackaged software from China, which is why I won't buy anything higher-level than a KVM switch or unmanaged Ethernet switch from there.

With mainline kernel support it should work booting your own kernel, haven't heard anything about it requiring images signed elsewhere.
But, it seems like you might need patches for all PCIe-drivers, to workaround their broken PCIe. As reported by phoronix:
https://www.phoronix.com/news/AMDGPU-Patches-Loongson-Bug

They wanted some fix to the AMD GPU driver merged into mainline, and:
Well that sounds like the usual re-ordering problems we have seen

patches for on Loongson multiple times now.

And I can only repeat what I've wrote before: We don't accept
workarounds in drivers for problems cause by severely platform issues.

Especially when that is clearly against any PCIe specification.

Re: [PATCH 1/2] drm/amdgpu: make duplicated EOP packet for GFX7/8 have real content - Christian König