for hw expert designers, a Frankestein GFX CARD for msx

Page 2/2
1 |

By rolandve

Master (254)

rolandve's picture

23-01-2021, 21:53

Grauw wrote:

..........
For a BASIC speed test, an additional factor is the efficiency of the BASIC interpreter implementation, this varies wildly and makes it difficult to perform raw CPU speed tests with it. But in practice (in assembly), I think probably an MSX and C64 are similar in terms of processing speed. At least I’ve not really seen strong claims that the one is significantly faster than the other.

My point is the opposite, It depends on the definition of raw. If you want a out of the box performance indicator, you would simply look at the performance of the system as a whole. The basic test, where you "for/next and print a value" is a test that out of the box works. Does it have a meaning? It tell's which system gives most cpu power to the user from basic. If an architecture eats 90% of the cpu cycles in waits for peripherals, if the basic interpreter is bad, well that's part of the machine.

Good performance indicators? Look at the demo scene and compare demo's. Demo coders aim to work around every limitation in a system in order to get the best performance.

By PingPong

Prophet (3789)

PingPong's picture

23-01-2021, 22:00

Maybe we are a bit OT?

Page 2/2
1 |