Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Pedro Passamani

macrumors member
Original poster
Jan 20, 2023
45
79
I was reading about the G5 processor earlier today and noticed that the fastest clockspeed it was able to achieve without liquid cooling was 2.3 GHz. That got me wondering, is there a big increase in power consumption and heat dissipation between 2.3 GHz and 2.5/2.7 GHz to warrant the use of a LCS? I've seen people converting their Quad G5s to use aircoolers, so why didn't Apple ship aircooled 2.5 and 2.7 GHz (maybe I'm stretching a little bit with 2.7) systems? I think that the Xserve, for example, with all that cooling power available, should've easily gotten past 2.3 GHz.
 
I have 2 Quad core systems. A 2.3Ghz quad core (I grabbed 2 dual core 2.3 and put them on a quad motherboard) and a 2.5ghz quad core. Both systems are air cooled. The 2.5ghz system is noticeably louder but other than that I cant really tell the difference and both run stable. I cant really even tell if one is faster than the other. The temperatures on the quad 2.5 are around the 60C range at idle and can go up to 70C+ while the 2.3 runs much cooler. I assume thats why they went with a liquid cooled solution (and to charge more :) )
 
Just a quick shot. The topic has a deep history and a lot of info.

The reason some people pull out the liquid cooling system out of the Quad 2,5GHz and the Dual 2,7Hz is that a number of them spilled out and eventually harmed the system. There were two manufacturers of the liquid cooling system Apple used. There was the discussion, if one of them was more likely to fail, but with increasing time the likeliness/difference became less clear. Also a lot became leaky, when they were bought used via Ebay/Craigslist and shipped rather then picked up personally (well you could still harm the connections of the hoses when you're driving badly yourself in case of personal pickup).

Reasons for overheating and or loud G5s can be dried out thermal paste, wrongly applied thermal paste (from stock, people found the loudness ok, because it is a workstation or the increase was not that apparent in the beginning). Wrongly calibrated fans via firmware after a CPU swap or thermal paste reapplying or because the firmware recognizes that not the original CPU was put in, though the same GHz and Generation.
and old worn out fans.

Re temperatures, there have to be the original apple documents around the net. If you have a G5 test with a voltmeter, this can be quiet interesting. If the G5 has heat problems that one isn't aware because of only a slight increase of temperature or fan noise, the power consumption can be higher, though.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.