After deciding that I would redo the MacPro case that I had previously posted about, I decided to go for the G5 as the case for an X99/Xeon system that I am piecing together over the next few months. I used the Mountain Mods rear panel and mobo tray combo to get the G5 ready to accept an mATX. I wanted to avoid the wiring hell that I had going on with the Pro case, so I decided to reuse the old PowerMac PSU case, gut it, and strip down a Corsair SFX 600. The guts of the Corsair, placed on standoffs epoxied to the floor of the old case PSU, left plenty of room to hide excess cable while also offering a better overall access to devices in need of power.
the nekid PSU case...
checking for placement. I may have to go with individually wrapped wire as I will need a bit more length on some.
I had some cutting to do re the lid of the PSU. I think I can get away with the top shelf remaining stock as it has wider openings...
OK, top shelf back on. Time to check how it mates with the G5 innards...
I had to pull one of the 80mm fans to get the dingus in...
And here we see the classic fitment, IEC meets the butt of the G5...
PowerMac G5 case mods: reusing the old psu case...
Started by m.oreilly , Jan 19 2020 01:50 AM
4 replies to this topic
Posted 19 January 2020 - 01:50 AM
Posted 10 February 2020 - 02:17 AM
I still have some cable tying to do, but this here x99 Micro2/10 core Xeon build fits in pretty neat:
Posted 10 February 2020 - 03:57 AM
Mostly used parts: 580 reference radeon, EVGA micro2 board, sf600 psu from another build, the E5 2660 v3 Xeon is from a server pull, NVMe Sammy 970 Pro from last PCIe-3 winders build. New Noctua fans and 32 gigs DDR4, along with a $19 cpu cooler (which is really working! 25C for pretty much all (20) cores)...
Posted 12 February 2020 - 09:37 PM
Looks good Mike.
Posted 14 February 2020 - 04:10 AM
It really is quite an agile system. I thought the low initial clock of the Xeon would hinder general performance, but it really does zip along. I am using an older version of 10, so am wondering if the performance is because of the lack of injection of altered Intel microcode re Spectre (I suspended updates for a bit. If the performance drops, I would take a guess that corrective microcode had been put in place re a Windows update...).
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users