for example, is 8kb*8 cache faster than 32kb*4, and 32kb*8 faster than 128kb*4?
think mkarcher, feipoa, jake or ?? had commented about it long ago. did you search ? search.php?keywords=double+bank+cache yields 15 pages of hits and have to get to work, maybe you can look thru them.
Hate posting a reply and then have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor. Stuff: https://archive.org/details/@horun
in short - no.
some motherboards may favor specific L2 cache configuration but there is no pattern that establishes one of the options as dominant.
It's not necessarily "faster" by itself, but on some motherboards you may be able to achieve tighter cache timings with double bank if the chipset/cache controller has interleaving support.
Nearly every 486 mainboard does bank interleaving in dual-bank configurations. If your board manages a X-1-1-1 burst in single bank mode at your desired FSB clock, you don't get any advantage from running dual-bank mode. Modern 486 chipsets usually have no issues with 2-1-1-1 at FSB33 even with a single bank.
At FSB40, 2-1-1-1 or 3-1-1-1 might work with 2 banks only. In that case, 2 banks is faster.
