Well... By 2003 the Alpha has been more or less abandoned in terms of R&D expenditures. It's impossible to estimate how much of its perceived lack of relative performance was due to progresses in x86 designs and how much was simply due to a lack of research in progressing the architecture.