I remember seeing what looked like a good faith criticism of RISC-V from a computer architecture engineer where she argued that a few details of the ISA weren’t going to be as good for implementing extremely high performance designs as the way the latest ARM ISA did things. But I don’t know to what degree that will hold RISC-V back or if it will be changed in some of the many permutations of the ISA as it is now being scaled up to high performance and probably soon extremely high performance. But last I heard, Longsoon chips had matched Intel’s current CPUs in some IPC tests, though they don’t clock very high and the tested chip was only 4c/8t so overall performance wasn’t great. That’s still far ahead of where RISC-V is right now in terms of desktop computing.
I remember seeing what looked like a good faith criticism of RISC-V from a computer architecture engineer where she argued that a few details of the ISA weren’t going to be as good for implementing extremely high performance designs as the way the latest ARM ISA did things. But I don’t know to what degree that will hold RISC-V back or if it will be changed in some of the many permutations of the ISA as it is now being scaled up to high performance and probably soon extremely high performance. But last I heard, Longsoon chips had matched Intel’s current CPUs in some IPC tests, though they don’t clock very high and the tested chip was only 4c/8t so overall performance wasn’t great. That’s still far ahead of where RISC-V is right now in terms of desktop computing.