>current-gen ARM is all low-power SoC trash
Will we ever see the return of socketed ARM CPUs for desktops?
>>60421179
Unlikely, unless the patents go public, there's no way ARM manufactures will route funds to r&d and production lines for something else than the mobile market.
The only viable alternatives to x86 on the desktop are Russian VLIWs or whatever the Chinese are working on(i think it was a CISC architecture).
RISC barely existed on consumer desktop computers to begin with.
Shit barely existed to begin with, even later Acorn desktops ran chips designed for embedded devices and PDAs, desktop ARM is a waste of time and effort for zero gain other than jerking off phonetards, bikeshedders and nostalgic hipsters who think it's still 1993 when RISC alternatives in the PC space were actually viable.
People on /g/ really need to stop making threads asking stupid questions that they would already know the answer to if they bothered reading about the subject for 15 minutes.
>>60422718
Oh, they've read about the subject, just nothing actually informative, just literal marketing filler and meme hype.
>>60421179
You can already get Cavium ThunderX or ThunderX2 if you want a full ARM computer although I'm not sure ThunderX line has integrated GPU for video out.
Linaro keeps trying to get a decent desktop ARM chip made. We need to nag ARM on twitter to help them fund it.
See this, I have messaged that account a few times and they reply so I know they read it. They don't really seem to know people would buy a desktop ARM chip to play around with.
https://twitter.com/ARMCommunity/status/863816435092553728
Same issue with the hikey 960, it's pretty powerful but ARM won't provide mali linux drivers for the SoC. No one ever complains so they don't think it matters.
Maybe Nvidia would make one
>>60421179
Maybe, I think your likely to get arrays of socs for server use taking over from VMs in many cases along with a resurgence of computational clustering