The Magician of Riga never goes undetected, even when his name is code 😀
Could it be the amount of ram?

Nexxen wrote on 2023-07-10, 21:51:

The Magician of Riga never goes undetected, even when his name is code

Even the great Tal likely wouldn't have sacrificed that much CPU power. 😀

While it runs fine in a virtual machine, CSTal (like many DOS games) appears immune to the effects of TSRs such as POWER and DOSIDLE. That causes significant heating of the physical processor and attendant high levels of fan anxiety.

Not sure what "bug" you're referring to here. DOSBox is an emulator, the CPU it implements will have much lower performance than your real one.

jmarsh wrote on 2023-07-11, 12:19:

Not sure what "bug" you're referring to here. DOSBox is an emulator, the CPU it implements will have much lower performance than your real one.

That begs a question: Is the processor always being emulated by DOSBox, irrespective of the setting?
I assumed including cycles=max would've enabled some kind of CPU passthrough mechanism.

drSolomon wrote on 2023-07-11, 14:38:

jmarsh wrote on 2023-07-11, 12:19:

Not sure what "bug" you're referring to here. DOSBox is an emulator, the CPU it implements will have much lower performance than your real one.

That begs a question: Is the processor always being emulated by DOSBox, irrespective of the setting?
I assumed including cycles=max would've enabled some kind of CPU passthrough mechanism.

Yes, it's not a VM but an emulator. Dosbox always presents an emulated, non-existent CPU.

konc wrote on 2023-07-11, 15:07:

Yes, it's not a VM but an emulator. Dosbox always presents an emulated, non-existent CPU.

Ahh...Ok, so it's always emulated. That makes sense.
The Wiki entry under Performance states that an Intel Core 2 Duo 3.33 GHz can be used to emulate a Pentium II 300MHz in DOSBox. One comparison between an Intel Core 2 Duo E8600 3.33GHz and the Intel Core i3-2350M 2.30GHz shows the two neck-and-neck in terms of performance, according to userbenchmark. This lead me to think that my i3 should be capable of no less than 100% on that benchmark—probably be a bit more.
It must be core=normal that's slowing it down. I can live with an i3 that's barely up Pentium 60 specs 😀
As jmarsh suggested, this is not a bug; I was simply unclear as to what was meant by max cycles. I'm still getting the precise hang of the terminology.
And thanks for clarifying that point about emulation. Much appreciated.

You have to realize that DOSBox is not multi-threaded. So more cores will not help. What does help is a higher GHz rating, or just generally a newer CPU.

Here is what I get on my notebook with default dosbox settings:

_Rob wrote on 2023-07-12, 11:24:

You have to realize that DOSBox is not multi-threaded. So more cores will not help. What does help is a higher GHz rating, or just generally a newer CPU.

Here is what I get on my notebook with default dosbox settings:

With dynamic core I can get 99% of a Pentium Pro 200. Setting core to normal, which is what I prefer, comes with an acceptable penalty now that I understand why it happens.