which earn me advertising fees or commissions.
As an Amazon Associate I earn from qualifying purchases.
2013 Mac Pro: Power Usage (Watts)
Related: 2013 Mac Pro, computer display, gear, GPU, Mac Pro, Macs, SSD, video
The 2013 Mac Pro is much more power efficient than the 2010/2012 tower models.
A Watts Up electricity meter @AMAZON was used to assess power consumption under several scenarios.
Operational power usage
3.5 GHz 6-core with D500 GPUs and 64GB memory and 256G flash drive attached to 30" 2560 X 1600 display, wired keyboard and mouse.
Figures fluctuate somewhat, approximate values as observed.
- Idle: ~ 46 watts.
- All cores driven by MemoryTester compute: ~150 watts.
- Memory Test compute plus Unigine Valley demo running simultaneously (to exercise the GPU): 210 - 240 watts.
But it appears that only one GPU is being used, so that if both were used power usage might rise to ~320 watts.
- Batch processing raw files to TIF: 90 - 145 watts.
These figures are terrific for the Mac Pro itself, but a specious view when examined critically in the context of a working system.
Your author’s first thought was that his electric bill would drop significantly since workdays involved 14 hours or so of Mac Pro usage. But it is a specious conclusion because the real power usage has been moved outside of the Mac Pro: several hard drives, dual PCIe SSDs, three SATA SSDs, eSATA compatibility box, and a USB3 expansion box. All those separate boxes are less efficient than with the same contents housed internally. In sum, the power usage might be worse with the 2013 Mac Pro, but at best it won’t be much of a win. Still, for those users with minimal external gear, the power savings will be real.
Thunderbolt 4 hub and ports!
Any Mac with Thunderbolt 3.
- Sleep mode: 2.5 watts, assuming screen sleeps at the same time.
- An OS X bug in display sleep (Mini DisplayPort to NEC PA301W) was seen to incur a 45-50W consumption even as the machine appeared to be in sleep mode. Also, the display frequently would not go off, though the mouse would "stick" indicating no screen updates and this was confirmed by wiggling the mouse and observing screen updates that had not been shown in the interim.