Wow, great stuff! Thanks for sharing. Exactly what I needed.
I think I found some minor errors in instruction pre-decoding, PD-xxxx10x0 and PD-0xx0xx0x have wrong values I believe.
Search found 4 matches
- Thu Mar 26, 2020 6:22 pm
- Forum: General Discussions
- Topic: Balasz, breakNES and visual6502 attribution
- Replies: 15
- Views: 1897
- Mon Mar 18, 2019 12:12 pm
- Forum: Emulation and Simulation
- Topic: Emulator recommendation
- Replies: 3
- Views: 1948
Emulator recommendation
I want to run code extracts in an emulator to evaluate code size and execution speed. In addition to being able to execute the code in debug like fashion I also want to know number of clock cycles spent, instruction trace, memory access maps, register state, memory inspector, ...
Is there a tool ...
Is there a tool ...
- Mon Mar 18, 2019 11:58 am
- Forum: Programming
- Topic: Calculating log and trig - algorithms or tables
- Replies: 19
- Views: 8546
Re: Calculating log and trig - algorithms or tables
In a BASIC ROM, you need to have subroutines implementing multiply and divide anyway. So from a code size perspective, it costs very little to make use of them. Using a Taylor series (which has a regular structure, so you can loop its generator code) might well end up as smaller code (albeit slower ...
- Fri Mar 01, 2019 4:33 pm
- Forum: Programming
- Topic: Calculating log and trig - algorithms or tables
- Replies: 19
- Views: 8546
Re: Calculating log and trig - algorithms or tables
Stumbled into this topic from a web search on Cordic.
Clock frequency 198.4 KHz
Precision 30 bits, including sign
Time for CORDIC operation 5 ms
In my implementation for 65C816 (precision: 128 bits) a sin/cos with arbitrary argument can be computed in 70/80 ms at 4MHz clock
Cordic: 992 ...
Clock frequency 198.4 KHz
Precision 30 bits, including sign
Time for CORDIC operation 5 ms
In my implementation for 65C816 (precision: 128 bits) a sin/cos with arbitrary argument can be computed in 70/80 ms at 4MHz clock
Cordic: 992 ...