OK.
You made me take one of R5200 O2s off the shelf and actually look. So this part is LIKELY a Tantalum Cap. However there is a problem, on my system (see Attachment) it has its own heatsink! I cannot make much sense of the engraving on the heatsink. The heatsink has small sides that "hug" the component on the board, and the heatsink is only soldered on one side (top) to the board, I carefully looked below (cannot really see in the photo) and it's NOT touching the board in any way on the bottom. But this heatsink is totally covering the component. So I cannot read the component itself. However, to me, the component is BLACK, not yellow/orange. It could still be a Tantalum (some tants are black, like poly tantalums are black).
So somewhere, this is supposed to a heatsink in your system. If you cannot find it...then that may add to your issues/suspicions. Now, since this is close to the RAM (and there is a similar layout of a cap in this position, C177, (though no heatsink cover) on the Octane motherboard.
Octane Mainboard Close-up Link
It MIGHT share the same value (C177 on the Octane, zoom in to read it) is 330uf @ 63V. Again, I'm guessing, it's a capacitor. But the exact value may only be found by carefully desoldering an example O2 board. It has a heatsink, your's is gone now (unless you found it rattling around).
Now as to the reason, it could be a one-off or it could be a failure point for RAM DC VCC or a Power supply issue or the cap is related to the IO ports near it. There is precedent for Tantalums outright failing (I'm told it occurs like 0.1% on brand new caps, and grows larger with aging). There has been a documented situation (In an Indy or a Challenge S, I don't remember) where a user had a tantalum cap blow (for I think the serial port), they soldered on a new one and were good to go.
Because of this one failure issue, there are two things suppliers/designers do: 1. There is a testing procedure (if you have the equipment, I don't) that basically runs the Cap in a quick pulse close to rated voltage/Amperage and sees if it fails, if not it passes. 2. designers massively derate the caps. Like 50% or more. So you'll often see tantalums that are for like 63v but are used on a 12v or less line. It's the margin, even if they might explode at rated values, they really shouldn't at half or less.
We don't really know the reason yours failed, you'd have to successfully desolder the blown mess, and check the board for a short at either side of the cap (both pads against local VCC rails) in Diode mode to see if there is a static short that caused an overdraw of current and popped the cap.
Without the heatsink, if even if the cap spontaneously blew, you need to come up with a third party heatsink or the right one because a heatsink indicates...they know it will overheat without one at some given point.
Sorry this happened to you, as these systems get older I'm convinced this will be much more frequent. I've stopped booting my older systems until I have the equipment to verify as many of the caps as I can. Normally Tantalum caps are more rugged, have some recovery ability, and are very stable during temperature changes. SGI used them to be more high-end at the time. Thought they do fail (it's just very rare). But they could be forced to fail if you draw too much power through them (for some reason).
The real blocker here is, most of the SGI system are inaccessible to probe while they are assembled. The short detection may or may not work without the PSU connected. But given this part's placement. I'd suspect a short in the main VCC to RAM voltage converter, that maybe blew the cap or if the cap is really related to one of the IO ports in the area, that could be the reason and it's not RAM based at all. But without further probing...that's pure speculation on my part.
Really, we don't know at this level. Most of these are still so cheap...nobody but a hobbyist would take the time to probe or care. Until prices rise a lot to pay for a real technician's time. It's nearly always been cheaper to just get another used part.
Also, not that this is really related, the O2 has a motherboard failure rumor, if the board is removed from the case BEFORE the PSU is removed, it can be ruined. There has been a rumor (forever, for me) that the PSU holds a charge, even after you unplug it. The designers didn't design the O2 to have the mainboard removed, while the PSU has a charge. So you can damage the mainboard if you remove it first. When I read this years ago, I always (no matter how long the O2 has been sitting) as a habit, remove the PSU (or at least slide it partially out), then I remove the mainboard. So I never accidentally remove the mainboard first. I do this for ALL systems, due to this rumor. This doesn't mean that's the answer to your failure, I'm just adding it on as education for any readers that never heard this. Remove or internally disconnect (when available) the PSU first on all system models!
Also, however told you that was a fuse, you may not want to put too much stock in their advice. If that were a fuse...it wouldn't erupt, violently, from the board with a burnt pad and all that..it would blow open, correctly, and intact. A fuse is made to blow, and stays in tact while it does so (in a controlled manner), not splits and flies apart. This is an uncontrolled, self-desoldered, mess. It's not a fuse, not including the fact the stencil label for this part indicates it's a capacitor.
:: 100MHz R4400SC, 96MB memory, 16GB SD card
:: 180MHz R5000SC, 128MB Memory, 17GB Hard drive
NEED -> 
![[Image: oJeSZw4.png]](https://i.imgur.com/oJeSZw4.png)