so is there actually a good reason to separate, physically or logically, the CPU from the GPU? couldn't you have an ISA that mixes both general-purpose and "graphics" operations? obviously you'd have to have vastly more cores than a standard CPU but you'd think you'd benefit from not having to shuffle data back and forth across a bus, and be able to just work directly on the same memory
i know very little about this topic so maybe this is completely retarded, idk. it just seems weird that even integrated CPU-GPUs logically separate the two components so you have to transmit data back and forth with e.g. glBufferData
i know very little about this topic so maybe this is completely retarded, idk. it just seems weird that even integrated CPU-GPUs logically separate the two components so you have to transmit data back and forth with e.g. glBufferData
bonifartius 𒂼𒄄
@velartrill gpus are heavily parallelized simd architectures. they are more like batch jobs of old times. you load a bunch of data, press start and then retrieve the results.
@bonifartius yea i have some idea of the basic principles, i just don't see why that couldn't be integrated into the CPU itself and share an address space. like, most(?) CPUs already have some level of SIMD capability
Follow
@velartrill i think it just doesn't fit the bill. i guess you could add it as a submodule. it just isn't useful for many things except graphics or number crunching. essentially it's just a bunch of alus.
futhark-lang.org abstracts over cpus and gpus, it looks nice but i don't have any use for it currently.
i'd really like to work on something heavy parallelizeable, it's currently only used for things like KI and cryptomining because nobody (i know of) really bothers to design new algorithms.
@bonifartius huh interesting
