There came a point in time where an idea dawned upon me which I wanted to try out: how would my Power Mac G5 run while cooled with frigid, outside Arctic air? I thought about it for a while, not wanting to damage the precious build I’ve carefully crafted for so long.
With this project behind me, I’m happy to report I was able to an extended test, and this Power Mac G5 has since remained operational, despite exposure to extreme temperature differences.
The first test
Although I conducted a test run the night before, I felt too guilty about potentially damaging the hardware to try it out for more than a few minutes. The next morning I mustered up some courage and decided to do it again, anyway. The first test was basically seeing how hot the CPUs would get while running a few heavy apps all at the same time: Photoshop CS4, TenFiveTube 720p, Minecraft, and an instance of Virtual PC all at the same time. Unsurprisingly, the CPUs were pegged to 100%.
Although CPU A would reach as high as 31 degrees Celsius during this run, CPU B remained ~10c behind at all times. At idle, CPU A would stay around 15.2c while CPU B remained around 7.8c. When placing the Power Mac G5 up on the window, I wouldn’t let it sit powered off for long. It’d get plugged up and fired up rather quickly, to reduce exposure time to the cold air in case of condensation. Miraculously, this never happens.
The second test
Setting it up on the window, the cold gust sweeps into the room, I hurry to close the window as much as the Power Mac G5 would allow – and I swear there are light crackling noises coming from inside the machine. I hope it’s not the GPU. So far since the experiment, time has been on my side, nothing was damaged in the process.
Temps in F for the AM experiment |
Temps in C for the AM experiment |
When streaming 720p in TenFiveTube, this Dual 2.0 2003 Power Mac G5, paired with a flashed X850XT 256 MB, Sorbet Leopard, 8 GB RAM, and an SSD RAID0 Array hovers at or below 30c for both CPUs. The warmest point is only 36.4c for the memory controller heatsink, and this test was conducted with the 120mm custom fan off – to simulate real world Power Mac G5 temperatures without assisted airflow.
Next up was to see how the game Amnesia: The Dark Descent truly performed without restriction (or as little as can be), on this Power Mac G5.
In Conclusion
- The Power Mac G5 appears to perform slightly better and more steadily with higher settings when introduced to subfreezing outside air, although this isn’t noticeable when using Finder/managing files. Amnesia appears to perform better, Minecraft not as much.
. - Unless I was just lucky not to break anything, this seems to be quite durable for being able to handle such cold temperatures.
. - All in all, this was a fun, quick and easy experiment to push the absolute limits and thermals of what a Power Mac G5 could do! Not particularly scientific as it was more spur of the moment, and I didn’t want to really damage the hardware. Perhaps sometime I’ll revisit this subject.
