After looking at the CHES Issue 2 submissions, efficient software implementation as a research subfield of cryptography is kinda dead.