Intel 210200-002 manual IAPX 88/10

APPENDIX

In Table 5 the iAPX 88 and the MC6809 are compared for "Ease of Programming" by counting the number of lines of code required for each benchmark. The iAPX 88 used a smaller number of lines of code than the MC6809 for eight of the ten programs. As in coding ef- ficiency, the greatest differences occurred in the two in- terrupt response benchmarks, with the MC6809 again having an advantage in the Single-Vectored Interrupt,

and the iAPX 88/10 using fewer instructions in the Multi-Vectored Interrupt. For the other programs, the iAPX 88's use of string instructions, and its ability to handle 8-bit or 16-bit data allowed the algorithms to be implemented in fewer lines of code. The Adjusted Average Normalized Lines of Code was 2.67 showing that the iAPX 88 used less lines of code than the MC6809 by a factor of more than 2.6 to 1.

	Table 5. Ease of Programming
	Lines of Code		Normalized Lines
Benchmark Program	iAPX 88110	MC6809	iAPX 88/10	MC6809
Computer Graphics	15	87		5.80
16-Bit Multiply	4	28		7.00
Vector Add	8	8		1.00
Block Move	7	14		2.00
Block Translate	10	13		1.30
Character Search	8	9		1.13
Word Shift	2	9		4.50
Reentrant Call	26	23		0.88
Single-Vectored Interrupt	15			0.Q7
Multi-Vectored Interrupt		8		8.00
Average Normalized Number of Lines of Code·				3.17
Adjusted Average Normalized Number of Lines of Code·				2.95

·See notc. Table 2, for description of average calculations.

IAPX 88/10	IAPX 88/10
1.00	1.00
			6809
			A~E	6809
		3	alB	AVERAGE AD::ED
8809			6809	'2':i4A~IlE
				2.81
ADJUSTED	8809		ADJUSTED
AVERAGE	ADJUSTED		A~E
.54	AVERAGE
			2.10
	.48
6809	8809			F
AVERAGE
.27	AVERAGE		~
				UO
			1.00
HIGHEST SPEED	SPEED WITH EQUAL MEMORY		BYTES OF CODE	LINES OF CODE
	ACCESS TIME
Graph I. Normalized Average Throughput:		Graph II. Normalized Average Memory Use and Lines
5 MHz iAPX 88110 vs 2 MHz 6809			of Code: iAPX 88/10 vs 6809

26 AFN 01532A