bkeys has quit [Remote host closed the connection]
<organizedglobals>
how do rv64 SBCs compare to A311D, RK3588, S905D3, or an Intel Atom x7? im looking for something for my parents to use thats small
<organizedglobals>
theres a lot of hype around running actually competent microkernels on rv-128 since its (rv128) IPC delay is nearly as small as Zen 3 and Itanium.. im hoping that just means i wont get called to come fix it when they've corrupted the sandboxed web browser or when grub decides to die
<organizedglobals>
s/(rv128)//
<davidlt[m]>
I don't think you can compare them :)
<organizedglobals>
well the khadas boards use the A311D or S905D3, 96rock and industrial control panels use the rk3588 and other Rockchip SoCs, and some small devices use Intel Atom, but it was mostly just benchmarked around the same speed as the S905D3 and A311D
<davidlt[m]>
Most current available SBCs don't have a "proper SoC".
<organizedglobals>
those dev boards are way overpriced right
<davidlt[m]>
Allwinner D1 might the the 1st one to look more like a proper mass produced SoC, but it has it's own issues.
<davidlt[m]>
It has some non-compliant bits for RISC-V.
<davidlt[m]>
PLIC is not compatible, RVV 0.7.1 (not compatible with the final spec), usage of reserved bits (non standard)
<davidlt[m]>
It's unknown when/how/if (most likely will) support upstream Linux.
<davidlt[m]>
But vector (RVV) on that thing probably will never be supported by upstream.
bkeys has joined #fedora-riscv
cyberpear has joined #fedora-riscv
King_InuYasha has joined #fedora-riscv
davidlt has quit [Ping timeout: 260 seconds]
<rwmjones>
davidlt[m]: where is the source for the t-head cores?