...
I run three G5 servers, a G4 server, and Intel based XServe systems, in racks. The 'noise' that people complain about doesn't at all bother me, and in fact, my RAID array is louder than any XServe system. I expect that if you don't rack-mount it, it's louder than if you do, for I barely hear all of the above, plus RAID arrays, and other servers, one room away, and with any audio or video playing, I don't hear them at all.
Well, maybe he should test himself, where his personal noise barrier is. Everyone's perception is different, as we see in this thread too.
Though, you may be right, that rack mounting them changes the noise reflection.
... (It's the reason that the original macintosh had no hard disc.)
+1 Didn't know that.
...Honestly, why do I care what noises come from a system?
because you have to sit in the room (desktops) or next to its room (servers)
15K RPM video drive sounds like a whirlybird, but it does the job that I require. ...
I often heard, that 15k rpm drives are good for video, but I see that in PowerPCs Velociraptors (10k) and SSDs (not thoroughly tested though!) do not make a difference really, when encoding video. Are you using it on an Intel-Mac?
I suggest using XServes as servers, not desktop systems. The main problem with noise, that I can contemplate, is from vibration, running them on a desk. That my racks are cabinets, may also dampen it.
From others posts and my experience, it seems to be different. Maybe e had different Xserve models. With mine, on the desktop I did not have vibration, I just had a jetplain, that I could hear downstairs in the kitchen. Or maybe me (and the others) had some that were dusty, had bad thermal paste on the CPUs or defective fans.
Again, having it in a rack, might do the difference, as you say, though I can't imagine that much difference.
@ OP: just start it without doing anything with it. Wait 30min. and see if you can stand it. If not, go in the other room and see, if that might be a solution for you.
If you want a PowerPC server. A PowerMac G4 would still be a costly server energy wise, but a fraction of the Xserve (arround 50-70W).
If you decision goes between a PowerMac G5 and an Xserve G5 not as a server, but for rendering (and the noise is no problem), you might compare buying costs. If Apples specs a true, it should be equal in CPU extensive tasks.
I would be interested, were the PM G5 tops out in a really extensive everyday task, energy wise. If the Xserve is really 290W top and the PM G5 450W top (I would rather think something around the same). I would guess, that the Xserve is not as fast. I mean, even with the same clock speed, shouldn't the Xserve be slower. Because, how should you save electricity, when you use the same CPU-chip?
As a 'desktop', the only notable features of an XServe, versus a g5 tower, as the hot-swap drive bays, the larger RAM capacity, the rather common dual-Gigabit ethernet, and a proper serial port. They use a PCI-X video cards, so if you aren't using the system entirely remotely, you will eat up one slot with a video card, leaving only one PCI-X slot for expansion, which will usually be either a RAID card, or an additional NIC.
He can have 16GB RAM, Dual GE plus 4x PCIe slots with a late 2005 PowerMac G5, though, with less noise.
