Lots of force while applying some downward pressure, in the direction in the first image. If there's any thermal paste used, it gets even harder since it probably has caked up by now and perhaps is the reason why it doesn't want to budge at all...

Tiido wrote on 2026-04-11, 11:23:

If there's any thermal paste used, it gets even harder since it probably has caked up by now

Do you think adding some drops of isopropyl alcohol on top of the exposed part of the ceramic might help to soften caked paste?

It can’t hurt.

IPA does help dissolving thermal paste.

Though if it's attached with a thermal adhesive, that may be a different story.

Unless some fellow retro enthusiast has been under that heat-sink before I'd say there's a pretty low chance of finding any TIM under it.
While IPA might help I would probably try some contact cleaner first, they usually contain some light lubricant that can wick under the HS.
And the step up from that would be a few drops of WD40 or equivalent.
Also don't forget to treat the tabs that slide =)

Should that fail perhaps our old frenemy heat could help.

And when/if you manage to get it off and suddenly feels an urge to use some TIM think again =)
The only material I'd even consider using would be one of those graphite pads, if it could survive the sliding (doubtful).

mkarcher wrote on 2026-04-11, 12:09:

Do you think adding some drops of isopropyl alcohol on top of the exposed part of the ceramic might help to soften caked paste?

It won't hurt to try but it probably won't really do anything since it cannot really get in there. WD40 might help as suggested by bofh.fromhell, it "climbs" every surface and might reduce friction in the retention mechanism too but it'll take a bit of effort to clean it all out later.

Thanks for the tips! Yeah, I may try some IPA just in case as WD40 would be a bit of a mess to clean there if it gets inside the socket.

Other than that, looks like I will have to apply some force!