APPENDIX
CONCLUSION
The results of this benchmark study show that the iAPX 88/10 significantly outperformed both the Z80A and Z80B for the benchmark programs used. Table 8 shows that the iAPX 88 is faster than both the Z80A and the Z80B, and that the iAPX 88 uses fewer lines of code, less memory and cheaper memory than the Z80.
The iAPX 88 did particularly well in the programs which were word oriented. It was also efficient to pro- gram due to the powerful instruction set and flexible ad- dressing modes. Both processors do have useful string instructions and a loop instruction with an automatic counter. The Z80 has faster interrupt response, but was slower and less efficient than the iAPX 88 for nearly all other benchmarks.
In view of these results, it appears that the iAPX 88 is a better choice for applications where high throughput, low development cost and low memory cost are impor- tant considerations.
Table 8. Performance Breakdown
| Performance Ratio of |
Performance Category | iAPX 88 to zeo |
Execution Speed (Z80A) | iAPX 88/10 is 3.79X faster |
Execution Speed (Z80B) | iAPX 88/10 is 2.52X faster |
Execution Speed (Z80A)* | iAPX 88/10 is 4.77X faster |
Execution Speed (Z80B)** iAPX 88/10 is 3.20X faster
Execution Speed (Z80B)*·· iAPX 88/10 is 3.83X faster
Ease of Programming | iAPX 88/10 is 2.51X more |
| efficient |
Coding Efficiency | iAPX 88/10 is 1.97X more |
| efficient |
NOTES:
*iAPX 88 vs Z80A with comparable memory (Z80A with 1 wait state). "iAPX 88 vs Z80B with comparable memory (Z80B with 1 wait state).
·*"'iAPX 88 vs Z80B with comparable memory (Z80B with 2 wait states).
9 | AFN·01664A |
