Single-threaded tests, such as the single-core Cinebench and POV-Ray benchmarks, highlight the Core i3-8100's only real disadvantage against previous-gen Kaby Lake models with higher clock rates. It takes a threaded metric to demonstrate the benefit of Coffee Lake's higher core counts.
The stock Ryzen 3 models aren't very attractive compared to Core i3-8100 in our multi-core Cinebench and POV-Ray tests. But overclocking shines a spotlight on the value of an unlocked ratio multiplier. Tuned Ryzen 3s nearly match the Core i3-8100, exposing a surprisingly close race in this type of workload. As expected, the bulkier Ryzen 5 models are even more effective, similar to the Core i5-8400.
Encoding & Compression
Core i3-8100 falls below the -7100 (and much of the rest of the field) during our LAME benchmark. AMD's 1300X and 1500X models hold small clock speed advantages over the -8100, but they manage to carve out a sizeable lead.
The multi-core 7-Zip compression workload naturally favors Ryzen 5 and Core i5, but an overclocked Core i3-8350K also offers incredible performance. The Core i3-8100 and tuned Ryzen 3 1300X fare similarly. The decompression workload goes AMD's way, with tuned Ryzen 5 CPUs leading the test pool.
There's a larger delta between Intel and AMD processors during our HandBrake x265 test compared to the x264 benchmark. This is likely due to the former's heavy use of AVX instructions. The Core i3-8100 beats both Ryzen 3 models in the x264 metric, though its lead is largely academic after we overclock AMD's CPUs. During the x265 workload, a heavier mixture of AVX instructions gives the -8100 a much larger lead. It even outpaces the Ryzen 5 models.
We also provide results from y-cruncher, a single- and multi-threaded program that computes Pi using AVX instructions. We tested with version 0.7.3.9474, which includes Ryzen optimizations. The y-cruncher tests show just how well Intel's processors perform in both single- and multi-threaded AVX workloads.
MORE: Best Cheap CPUs
MORE: Intel & AMD Processor Hierarchy
MORE: All CPUs Content
Just a glaring sign that the articles value comparison is off, in real life.
The performance loss might change the results here significantly. I know it's a lot of work, but I think we need new benchmarks for pretty much all Intel's new models. The current numbers are no longer accurate. Maybe retest some of AMD's to see if there is much, if any effect on their performance as well.