Cuttoon wrote on 2022-05-01, 17:52:

Mate, what I wrote about at "General thought on GPU cooling" and on that Sunon cooler, the main point was: It's not an exact sci […]
Show full quote

AppleSauce wrote on 2022-05-01, 16:36:

One other thing , I was reading about fans and there was something about cfm and airflow […]
Show full quote

Cuttoon wrote on 2022-05-01, 15:58:

Well, to be fair, that system is centered around having the value of an average motorcycle in a single retro system. ;) The DX7 […]
Show full quote

Well, to be fair, that system is centered around having the value of an average motorcycle in a single retro system. 😉
The DX7 functionality of that rare GF256 DDR could have been achieved with pretty much any random five bucks Geforce MX card with passive cooling and no apparent difference in use.

That Sunon Cooler: Most fans I encounter are 12 V, but that clearly says 5 V.
So, make sure to check what the card will actually provide and go direct to molex if the card's socket is wrong.
But, apart from that trivial problem, uhm, cooling isn't measured in Ampere.
Not even motor performance is measured in Ampere. But Watts - which only indicate electric power use, not actual aerodynamic effect. Huge differences in efficiency out there.
And anyway, there's only one way to find out - install it and see if it's stable. I'd bet a crate of beer it is or at least it's not about the fan then.

Still, the fancy copper heatsink I've shown there gets sold here for 28 € which if of course outrageous. But if you could find something equivalent - pretty sure it'll fit the slot and suffice, without fan. Copper was almost unknown back then...

One other thing , I was reading about fans and there was something about cfm and airflow

I found both the datasheets for the original and potential replacement but I cant quite understand the diagrams.

UDQFSEH01 (Original)
https://datasheetz.com/data/Fans/DC/P11090-datasheetz.html

GB0545AFV1 (Replacement)
https://media.digikey.com/pdf/Data%20Sh ... 1-8(1).PDF

If you do happen to understand them since you seem pretty knowledgeable , does the replacement have a similar airflow to the original even though it draws less power?

Mate, what I wrote about at "General thought on GPU cooling" and on that Sunon cooler, the main point was: It's not an exact science.

Even if you're an R&D engineer with a PhD in physics, you'll still need years of experience to even assess the cooling solution for one specific chip in one specific case, in advance.
Since there's no way of knowing the latter, what the user will have, those things get over dimensioned routinely, by quite some margin of error.
Your case is easily 10 years more modern and well-ventilated than the average in 2000. Therefore, give it a try - if it runs stable, it's sufficient. That's how you find out. Again, I'm rather sure it is.

That you've found the actual data sheets for those coolers - all hail to you.
But in PC tinkering, no one ever gives much thought to those, if even aware of their existence.
The fancy graphs show you the airflow between "idle" state without any resistance and the resistance or counter-pressure where the air stops moving.
It's called this and any engineering part has at least one:
https://en.wikipedia.org/wiki/Characteristic_curve

That these data sheets exists shows yo that Sunon is an actual, serious manufacturer of those things.
That they use silly, nonscientific units like "inch" and "feet" - not so much.

CFM means "cubic feet per minute" IIRC, which should be forgotten right away as it is a remnant from a primitive, pre modern society and alien to modern technology.
They use "m³ per minute" alternatively, at least.

- The one PDF indicates the fan has 5500 RPM and the other, the large table on page three, shows those range between 4500 and 13000 RPM.
The average Socket 7 fan was around 4 to 5000 I'd say.
The Sunon you have has 5500, that's already quite a lot - and 500 more than the old one.
So, never mind the electric data, but at a certain size, a given RPM will mostly do the same, apart from minor tweaks.
It's not so much a question of choice, the more the merrier. It's just too loud pretty soon.

This is how a Gf256 DDR with passive stock cooling looked like, only as the "quadro" CAD edition:
http://www.vgamuseum.info/images/vlask/nvidia/quadroddru.jpg

Yeah sorry for being a bit ocd , I just wanted to find a good replacement that fit my criteria.

I went ahead and ordered the sunon fan so I guess I'll have to see if it fits the bill when it gets here.

AppleSauce wrote on 2022-05-01, 08:11:

Well it turns out the ebay seller made a mistake and they didn't actually have the fan in stock and they gave me a refund so ill probably have to improvise something.

Would it be possible to swap the fan with a passive heatsink like this one , or does it need a fan?

20220501_180110.jpg

I would add a fan to it, if only because adding a fan is really kinda trivial and it will definitely help with cooling and especially if you intend to use it for extended periods of time. Also it will definitely help prolong the life (and thus its functionality) if you keep it cool.

Personally I always tend to favor leaving the slot directly below the main graphics card empty for airflow.

Howdy folks , bit of an update , the fan came in and looks about right except for the fact that there's a rubber seal thing on the back that I'm guessing is for the oil.
Would it be okay to leave it off since i'd want the block to sit flush with the gpu die , or would oil leak everywhere if I did?

I'm just worried if I leave it on the fan will sit awkwardly or wont do its job properly, cheers.

I went along and mounted the new fan and its spinning.
GPU booted up and displayed the video bios , but my pc has been restarting randomly , so either the power socket is dodgy or the psu is starting to go.
But that's another job for another day.