@p Here is your new Chinese laptop (they are from Hong Kong in fact): https://canonical.com/blog/worlds-first-risc-v-laptop-gets-a-massive-upgrade-and-equips-with-ubuntu
https://deepcomputing.io/product/dc-roma-risc-v-laptop-ii/
(The second link is behind CAPTCHA for some reason)
Now with AI!
(yes, it has an NPU 😩)
The press:
https://www.theregister.com/2024/06/13/riscv_laptop_ubuntu/
https://www.phoronix.com/news/DC-ROMA-RISC-V-Laptop-II
@m0xee@social.librem.one @p@fsebugoutzone.org wasn't there already another chinese risc-v laptop a year or two ago?
Oh yeah they have.
...I'd rather keep my thinkpad (or buy the /mnt laptop from mr colgate)
@iska
Yeah, it's the same company, but this one's more decently specced and it supposed to be cheaper despite that 🤭
I couldn't find any hands-on reviews of the old model, so maybe it didn't make it into mass production and they decided to just "upgrade" it this way. Going into mass production would also explain the price drop.
And yes, sure, Reform is cool, but I'd prefer RISC-V over ARM if they are otherwise more or less the same.
@p
@m0xee@social.librem.one @p@fsebugoutzone.org I thought reform had a risc-v module ngl
I was about to mention loongson for being mips and decently fast, but dug a little deeper :darkness:
RISC isn't real anyway, that laptop's cpu has 500 instructions
@iska
I don't know, maybe it's possible to have a RISC-V CPU module for reform as the design is originally modular, although it might require extra work, but all currently available offerings, both for original Reform, and for newer Pocket Reform, are ARM-based:
https://shop.mntre.com/products/mnt-reform
https://www.crowdsupply.com/mnt/pocket-reform
Which isn't bad, like I said, RISC-V is more a personal preference in this case.
@p
@iska
And sure, ARM, which is originally de-abbreviated as Advanced RISC Machine doesn't use limited number of instructions either, I suppose, that would make instruction set extensions impossible.
Modern RISC approach is about having to use a particular set of registers for operations, not having different flavours of the same one for different operands like Intel does, but not about having fewer instructions in principle 🤷
@p