<x​fedex:matrix.org> hello! i've got a question for tevador, if you are available. tevador/equix seems very interesting, do you have an idea how performant it could be on GPUs? and how hard to build an ASIC for it?

<x​fedex:matrix.org> (github.com/tevador/equix)

supposedly slower than on CPUs

there's no implementation to test it however