looks like the implemented CPU (ULi M6117C) supports a coprocessor interface, so it is entirely possible that a 387sx equivalent could give it floating point capabilities. it’s probably not electrically implemented on this specific device to expose the interface though. otherwise yeah no FP sucks
While an FPU (Floating Point Unit aka math coprocessor) would be nice, the 386SX is still a 16-bit CPU. The 386DX is a 32-bit CPU (and would still need a 387 for a dedicated FPU). In the 486 line SX meant no FPU, while DX meant FPU inside.
ahhh fuck I forgot about that little stupid memory bus cheat that Intel pulled. they’ve been good at obfuscating their screwing of consumers thru weird naming schemes ever since they started lol
It runs Doom, but unplayable speed. They should have run it as 386DX. Had one from AMD, it was running quite well.
looks like the implemented CPU (ULi M6117C) supports a coprocessor interface, so it is entirely possible that a 387sx equivalent could give it floating point capabilities. it’s probably not electrically implemented on this specific device to expose the interface though. otherwise yeah no FP sucks
While an FPU (Floating Point Unit aka math coprocessor) would be nice, the 386SX is still a 16-bit CPU. The 386DX is a 32-bit CPU (and would still need a 387 for a dedicated FPU). In the 486 line SX meant no FPU, while DX meant FPU inside.
I read this in R2D2’s voice!
ahhh fuck I forgot about that little stupid memory bus cheat that Intel pulled. they’ve been good at obfuscating their screwing of consumers thru weird naming schemes ever since they started lol