TAGG commented on own score:

 – DIE says 950Mhz but there's no such chip in the DB, it also defaults to 900Mhz, so i guess it's working as intended?

TAGG is right , there is no Thunderbird 950 Slot A in the database.

You should apply here (if you are 100% sure it's an original 950)
https://community.hwbot.org/topic/40155-please-add-other-hardware-thread/

edit

https://www.cpu-world.com/CPUs/K7/AMD-Athlon%20950%20-%20AMD-A0950MMR24B%20A.html

5 hours ago, eviledeath said:

Hi Tagg, what was the code printed on the processor cartridge? It was very common to have different dies under the cartridge.

no idea, most of my Slot A cpus are from the scrap and i received them as a naked PCB, but it defaulting to 900 makes me think it's indeed a downbinned DIE.

3 hours ago, TASOS said:

You should apply here (if you are 100% sure it's an original 950)
https://community.hwbot.org/topic/40155-please-add-other-hardware-thread/

nope i'm more like 80% sure it isn't a 950 chip as default multi without GFD connected is x9 (and there's no obviously knocked off resistors in the area to set multi so i think it's stock)

My 900 is a down binned 1K.

4 hours ago, TAGG said:

no idea, most of my Slot A cpus are from the scrap and i received them as a naked PCB, but it defaulting to 900 makes me think it's indeed a downbinned DIE.

nope i'm more like 80% sure it isn't a 950 chip as default multi without GFD connected is x9 (and there's no obviously knocked off resistors in the area to set multi so i think it's stock)

Becides the soldering marks and traces ... where the specific resistors are placed (which must appear clean)

There are 2 more elements that verify the original speed of a SLOT A cpu. (for both pluto,orion and thunderbird).
Close photo's of both sides of pcb are needed in this case.