WHY I STARTED MAKING THESE MOUNTS
It started the same way most bad ideas start, gaming on a weekend with mates, talking about FPS like it matters more than rent.
We’d just watched a video of a 5090 running the exact game we were playing, absolutely crushing it. And for about five minutes we all did the same thing “Imagine that FPS…”
Then reality kicked in. None of us are buying a 5090. Not now. Probably not ever.
And that got me thinking… how far can older, cheaper, more realistic hardware actually go?
I already knew cold helps. Everyone knows cold helps. But how cold is “cold”? And what does it really do once you push past “better case airflow” and into “this is probably a fire hazard”?
The problem was always mounting. I’d tried sticking CPU coolers on GPUs before and it was always messy. Crooked pressure. Uneven contact. “Looks fine” until you load a game and watch temps sky rocket.
Then my wife bought me a 3D printer for my birthday, and suddenly it all started making sense.
With a printer and a pair of calipers, I could actually measure things properly… the block, the spacings, the die, the standoffs, and build a mount that gives real contact every time.
(Almost every time. This is a garage, not a lab.)
I used to envy people with a 5090.
Now I’m glad I don’t have one.
If I did, I don’t think I would’ve found the drive to keep pushing. The whole fun is chasing it, every degree lower, every frame higher. That’s the game now.
HOW A MOUNT HAPPENS (THE SIMPLE VERSION)
If you want to do this yourself, the process is basically:
Sit the cooler/waterblock on the bare die
Measure from the top of the cooler/block down to the PCB
Measure mounting holes and clearances
Design a mount in CAD to match those measurements
Print it, test fit it, adjust it
Repeat that whole thing 3–5 times until it’s perfect
And if you don’t get it on the first try, don’t stress. Neither do I. You’ll get it.
Or… if you want to skip the “print 4 versions and lose one screw into the carpet forever” part…
You can just buy mine.
