IRC channel logs
2024-09-17.log
back to list of logs
<daddy>howdy o/ thought i was still in here, i was not. <daddy>oriansj: are there any plans for a new stage0-posix tag soon? i'm trying to update live-bootstrap within stagex and it needs a newer stage0-posix for unxz that has yet to be released <Guest50>Hi, I found the stage0 project, which sounds very interesting <Guest50>The readme says, under "Future dierctions" <Guest50>> Implement the Knight processor in FPGA and then convert into TTL. <Guest50>I guess it's not just the transistor-transistor logic technology, e.g. as opposed to CMOS <Maja>i assume shit like the 74xx logic series <Googulator>It needs to be something doable without modern lithography equipment that itself relies on firmware and modern lithography for its own existence <Googulator>Bootstrapping lithography would be a formidable challenge, mainly because virtually all relevant lithography techniques are proprietary and trade secret <Googulator>Achieving 6502 or maybe 80386-level performance is the "easy" part - going from there to something with Intel Core 2-level performance is the more problematic task <nikolar>i mean there's like 7 orders of magnitute between 80386 and core 2 transistor counts <aggi>Googulator: and it's not desirable to do it for modern monstrosities of what's sold ax x86/64 nowadays <aggi>64bit in general seem a little unrealistic, because an intermediate FPGA deployment is either expensive or not feasible <aggi>a 32bit x86 however, this would be great if such existed, fully opensource with permissive license, without relying on proprietary tools <nikolar>i am pretty sure 32bit x86 patents have expired <nikolar>so sounds like that should be perfectly doable <Googulator>IIRC the original patents on AMD64 also expired in 2021 <nikolar>so you could even create an x64 chip <Googulator>yeah, SSE2 definitely was on the market in 2004 already <nikolar>you can't really have x86_64 without it <nikolar>i am just not sure which version of sse it was