A Look At Alder Lake Against Older CPUs Going Back To Sandy Bridge, Kaveri

[Setif]

It would be nice to compare efficiency (performance/watts), as they always mention it when announcing new CPUs. I am still wondering if their claims are true or just a marketing!

[free2create]

Yes ditto would it be possible to add a part 2 and look at performance/power consumption.

[Michael]

Yes ditto would it be possible to add a part 2 and look at performance/power consumption.

Click the openbenchmarking.org link... all the graphs should have power numbers and higher is better graphs also have perf/watt.

[polarathene]

In some tests, e.g. Blender 3.0, 12900K is more than 10 times (!) faster than 2500K. That's insane. "4 cores vs 16 cores (24 threads)" doesn't sound like it could lead to such a massive difference.

Just compare the 2500K (2011Q1) to the 12900K (2021Q4), it's not that unbelievable when benchmarking CPU perf is it?

A 3.7 GHz max turbo frequency vs 5.2 GHz max turbo, or more importantly the 5x larger cache of 6MB vs 30MB? Uses up to 2.4x the amount of wattage, more than 3x the memory bandwidth. Plus probably a bunch of other architectural improvements over that decade delta between them.

I know that the CPU cache can make a big difference for stuff like this though. It's considerably faster than having to access main memory, so having more room there brings significant speedups for workloads that can leverage it.

[onlyLinuxLuvUBack]

Maybe you could add some more tests like: blender load a 10 million triangle .obj file, imagemagick mosaic a bunch of images together.

[jaxa]

My 5700G just edges out the i5-12400, yay.