The difference in idle power usage between the E5 2603v2 and E5 2697v2 is minimal, typically around 1-3 watts.
The difference in idle power usage between the E5 2603v2 and E5 2697v2 is minimal, typically around 1-3 watts.
When examining my signature, it becomes clear the machines I use are quite similar. One runs the fastest E5 2697v2 model, while the other is the slower E5 2603v2. The first handles most of my dockers and tasks, whereas the second is mostly idle. I already had the motherboard ready, so I purchased the 2603v2 for less than ten bucks on eBay. No disappointment if it eventually gets discarded. Recently, the 2697v2 has dropped to around $100 USD. It looks like the value of older Socket 2011 parts is falling. Right now, I don’t feel the need for extra speed on my second UnRAID machine, but I could upgrade later. The current hardware could last me a decade, and I might add more features or dockers to it. For a machine that sits mostly idle, the power usage difference between idle and light loads is what matters most.
You already have the chips, so why not try it yourself? There are many power monitoring tools for those devices on Linux, and they display the energy usage. From my experience, the difference isn’t big—usually under 10 watts when idle, but it could be higher with light use or certain docking stations.
Most of my home entertainment system is built around UnRAID #1, and taking UnRAID #2 out for maintenance feels like a hassle. I often have to shift a sofa and push it off an Ikea table onto the floor to access it. It's quite frustrating.
The processor includes a power display on the chip. By checking its reading, you can determine the exact power usage of each CPU. It's quite precise. You may choose to halt VMs or dockers for more accurate idle measurements.