The previous QV transpile benchmarks were all loading the circuit from qasm2.0 files. However, a real quantum volume circuit should be defined with haar random SU(4) matrices and the compiler's job is to synthesize that to a realization that is executable on the hardware. With the exception of staq all the toolkit's we're running this benchmark on support representing a quantum circuit like this, and we already had code to generate a quantum volume circuit for construction benchmarks. Staq is the exception here as it is unable to represent an arbitrary 2q unitary operator as it only supports qasm 2.0 as the input format. This commit makes this test a hard failure as the library lacks support for this where everything else that has a transpiler supports it.
The previous QV transpile benchmarks were all loading the circuit from qasm2.0 files. However, a real quantum volume circuit should be defined with haar random SU(4) matrices and the compiler's job is to synthesize that to a realization that is executable on the hardware. With the exception of
staq
all the toolkit's we're running this benchmark on support representing a quantum circuit like this, and we already had code to generate a quantum volume circuit for construction benchmarks. Staq is the exception here as it is unable to represent an arbitrary 2q unitary operator as it only supports qasm 2.0 as the input format. This commit makes this test a hard failure as the library lacks support for this where everything else that has a transpiler supports it.