In an old document in this server I compared the performance of the 6502 processor against its main competitor, the Z80, and also against an hypothetical RISC CPU. In that document I was, maybe, a bit too harsh with my criticisms about the poor chip and a long, long, time after I got some feedback from John Brooks and Kent Dickey, starting a new series of tests.
This is the original "On the 6502. A brilliant or sloppy design?" onthe6502.pdf
And this the response I got from Kent Dickey "The 6502 Ain’t That Bad" 6502.response.pdf
And after some more tests, many of them done by John Brooks I think we can summarize then on this document update.pdf
This kind of test is nothing but new. As Kent pointed out, a similar comparison was published long time ago in the Byte magazine:
As a last addition to these Byte magazine results, just take into account that John Brook's sieves calculates the primes under 16384 in just 117ms on a 2.5MHz Z80 (or around 150ms on a 1MHz 6502). About the same time it took on a Cray-1 supercomputer back in 1983, while also requiring just 1KB of memory for the sieve. Programmers like him ought to be paid a lot more!!!
