-
|
I'm currently hacking together a 8080 CPU (without a microoperations ROM) and when I do a speed test from the menu on the design it gives a number like 10735,0 kHz which sounds very very high but would have been very nice if it was true... :-) When I measure the actual simulated frequency of the clock (set at Maximum) it turns out that I only get about 57 kHz. That's quite a difference. |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
|
The speed test clocks the model for one second and calculates the frequency from the number of clock cycles that could be calculated in this second. |
Beta Was this translation helpful? Give feedback.
-
|
Ah yes. I forgot how the number of changing nets affect the speed - not just the total numer of nets in the design. Having the design idle I get the clock ticking away at about 800khz. Executing NOPs which affect just a small portion of the nets reduces the speed to 275khz. When running a diagnostics program I'm down to 50-60Khz. I do measure the clock itself, not the number of instructions executed. But even the 800 kHz in idle mode is an order of magnitude off from the reported figure. I'll play around a bit more and see. |
Beta Was this translation helpful? Give feedback.
The speed test clocks the model for one second and calculates the frequency from the number of clock cycles that could be calculated in this second.
However, if the processor is running without a program, it could be that only NOPs are executed, which is very fast.