This was not the first submission that I doubt but I want to make it the last.
http://forum.hwbot.org/showthread.php?t=7436 - here's the start of the story.
In short, at first REIBEROCROSS shows a Sempron validation that is not a Sempron at all. He tries to explain it with bridge mods. The only problem is that Sempron CPUs are known to have locked multiplier and can't change their multiplier via bridges.
Then he pretends to be deaf talking about Duron Applebred unlocking (most of which are unlocked and don't require any modifications).
Yesterday, he brings this thread up after 4 years! And shows what he wants us to think is a unlocked via modification Sempron CPU (for the younger, there indeed were none Semprons with default x15 multiplier) - http://valid.canardpc.com/0ccfmn
I want to ask moderators to question him the nature of the mod and verify it by repeating it step-by-step. He doesn't need to make it public, I just want to be sue it's not cheating.
Pretty much the same verification you've done with bob's tweak and others.
If you need any help understanding Socket A modifications or other in-depth things, I'll provide you all the needed info.