What mainboard, what BIOS, what exact type of 486 CPU?

And also what benchmark. A dx2-66 with L1 cache off performs around the level of a slow 386, according to 3dbench.

Yes, typical situation, I know this is specific but generally is what like for intel and amd 486dx 33, 66 and 100 with internal cache off, external cache on?

Cheers,

pentiumspeed wrote on 2021-10-25, 16:38:

Yes, typical situation, I know this is specific but generally is what like for intel and amd 486dx 33, 66 and 100 with internal cache off, external cache on?

Cheers,

Are you drunk? Read your post again and see if the sentence structure and wording makes any sense.

Is it even possible to have external cache on when internal is off?

Yes, when I say internal cache off this implies CPU?

External cache on meant motherboard's cache.

I thought comma meant "and"?

Well?

Cheers,

"generally is what like", we might guess is "generally what is it like"

486 was designed to rely on L1 cache, so performance penalty for disabling it is steep.

For a really rough guess then you can say a cache disabled 486 is as fast as a 386sx at the same clock, DX33 would be similar to 386SX33 but the clock multi variants are more affected. It's gonna vary by workload because 486 will still do some instructions in less cycles, but other functions on CPU or on motherboard will be extra crippled vs 386SX equivs that were designed and tuned to work without the cache. For multipliers, I would make a guess at it being multi-1 = number of base clock steps boost vs 386SX i.e. a DX2-50 would have a 25mhz base clock and the boost from the multi would onlt be equivalent to going one step up, so 33mhz 386SX equivalence, whereas a DX4-75 with a 3x multi on same base clock might be more like 386sx40 equivalent, as would a DX2-66

But it's all like saying things such as "If a banana were a lime how orangey would it taste compared to an apple if it were a grapefruit?"