No figures for a Xeon machine in particular, but some other figures for a general idea (from The Carbon Footprint of Everything by Mike Berners-Lee, 2020 second

No figures for a Xeon machine in particular, but some other figures for a general idea (from The Carbon Footprint of Everything by Mike Berners-Lee, 2020 second edition, page 140): Manufacturing a computer: * 326 kg CO2e: 13-inch MacBook Pro, 128GB storage * 475 kg CO2e: 15-inch Dell Precision 5530, 256GB storage * 620 kg CO2e: 16-inch MacBook Pro, 1TB storage Using it: * 4g CO2e per hour on 13-inch MacBook Pro * 6g CO2e per hour on 16-inch MacBook Pro * 20g CO2e per hour on average-efficient laptop * 98g CO2e per hour on desktop computer with screen * 130g CO2e per hour on gaming PC with screen The figures for using it are based on the US grid's typical mix of coal, renewables, nuclear, etc. In places with a greener grid, such as the UK, the figures are different (there's a different edition of the book for the US and UK version, and possibly other regions as well.) So we don't have figures for the emissions for a Xeon machine, either in manufacturing or usage. And these figures are estimates, of course, not precise measurements. But they help us to get a good picture of the scale of the impact. Let's assume that you use your computer 12 hours a day, 365 days a year. On a standard gaming PC with a screen, that'd be 569400g CO2e per year, or 569kg. The big question then is how much more inefficient is a Xeon machine than a standard gaming PC. If the Xeon machine emits 1.33x more than a standard gaming PC, that'd put its yearly usage emissions at 757kg CO2e, so if you used your machine for 5 years, than the difference in emissions would be 940kg CO2e between the Xeon and standard gaming PC, which would probably be enough to justify buying a new machine. But if it's only 1.1x more inefficient, than that difference would be 285 kg, not nearly enough to justify a new machine (and probably still the case even if you use the machine for 10 years.) Ultimately, I'm not sure that we have the information necessary to definitively answer this, but all other things being equal, I'm skeptical that the CPU alone will make an enormous difference in power draw in day to day usage. A newer CPU will be more efficient when at 100% usage, but does it draw significantly less power when at idle? I don't know. But if anyone else has found any information that could help, I'd love to check it out. That said, there are plenty of second-hand machines that are more efficient than a Xeon machine. Once there are no second-hand efficient machines left, it'd be worthwhile to dig in deep on the emissions of old Xeon machines versus newer more efficient machines, but right now you can just buy used and refurbished efficient machines, so I'd probably just recommend that.