Upon comparing Intel's Skylake (slightly old) but very optimised version (Skylake-SP which is used for LGA 3647 xeon processors) I found that in benchmarks from Cinebench to Geekbench the AMD Zen processor only suffered at most a 11% performance loss due to architectural design and I was even assuming that the AMD processor was running at max turbo boost (or very close to it) - If it hadn't been significantly less than max then there would have been almost no performance difference between the architectures.

For this i used the Intel Xeon Bronze 3106 and the AMD Ryzen Threadripper 1900x for comparison.

I used the Xeon Bronze because it has the same number of cores and very similar amounts of cache (if not a bit more which may have even given it more of an advantage) to the 1900X and no turbo boost (so comparison was easier). I used the AMD TR 1900X because the turbo boost was minimal (only 4.0GHz compared to base 3.8GHz clock speed).

The Bronze has a clock of 1.7GHz with nearly all other factors very similar or the same as the 1900X and so i used the clock as a comparison indicator.

Assuming the 1900X clock speed was at max turbo then 4.00 / 1.7 = 2.35 which means that the 1900X clock is 235% more than the Bronze one.

Lets compare the benchmark results (single core benchmarks to eliminate the hyperthreading differences (Bronze has no hyperthreading)):

- The increase from Bronze (B) to 1900X (X) for Cinebench R11.5 was 230%

- The increase from B to X for Cinebench R15 was 224%

- The increase from B to X for Geekbench 3 was 230%

235% (clock difference) - 224% (Lowest benchmark difference) = 11%

11 / 224 = 0.04910714285714 * 100 = 4.91% ≈ 5% increase in performance all down to the factors that weren't the same between the processors (mostly architecture e.g. Pipelining and perhaps other small factors).

These benchmarks were taken from www.cpu-monkey.com but all other sources showed a very similar trend.

So what do you think? Has AMD finally bridged the design and architectural gap between them and Intel while still keeping their processors cheap?

What do I think? I think you've committed a cardinal sin of mathematics and performed mathematical operations between two different percentages, and you've drawn conclusions from that. You can't do that.