Some people would consider this comment "FUD and baloney" because of the number of confounding factors you're jumping over. You're measuring wall socket voltag

Some people would consider this comment "FUD and baloney" because of the number of confounding factors you're jumping over. You're measuring wall socket voltage on two different servers to make a commentary on CPU power usage, but that's including storage, power delivery efficiency, cooling solutions, etc You didn't describe what "we load it up" means for the Intel server, or how that compares to the AMD server, or what load is for that even (you're saying you keep the AMD server pegged at 100% CPU, or you have 100% of your expected load?). Presumably you didn't get a dual socket CPU for giggles, so the load isn't trivially comparable between both systems by either measure. If you're buying servers because of fallacies about entire brands, please leave the procurement to someone more experienced. That kind of stuff is more for building a gaming PC, server hardware is way too varied for "brand X is Y". You need to examine things at the SKU level.