Not a critical point of your post, but in my experience CPU Mark scores change over time as new CPUs are added. I don’t know the details (do new data points revise the score? does testing methodology change?) But, those changes can be significant over years. Every time you add a new CPU, you should re-check/update the scores for all the others too for proper relative comparisons.
You might explicitly list the mobo chipset and socket rather than inferring it from the CPU and mobo name, or maybe there’s a separate mobo details sheet.
Y assume you have info about PSU, PCIe slots and networking. OS and other software? IP and IPMI addresses?
Also, the performance you can get on a CPU today will not necessarily be the same performance in 12-24 months’ time with microcode updates. Things can theoretically get faster, but they more typically get slower as the CPU manufacturers end up mitigating vulnerabilities along the way.
What’s happening is we inflate objects (could be a “Build” or a “Cluster”) with details from related subcomponent sheets (say, “Memory”, “CPU”, or “Motherboard”) if those exist, then we run these compatibility check functions to determine the drop down lists for each sub-component type, i.e.: only list compatible options given all other already selected sub-components.
Speaking of which…
What compatibility checks does everyone here do, that I should definitely not overlook?