Page 1

ND 073 BCD Arithmetic for ND-500

A special Hardware Module for Use in Commercial Data Processing on the ND-500/CE Series of Computers.

In order to increase the speed of COBOL programs on the ND-500, a special BCD-processor is provided as an extension of the ND-500 CPU.

The processor performs arithmetic on Binary Coded Decimal (BCD) numbers, and converts data fields between BCD-format and ASCII-format, and between BCD – and binary (integer) format.

This new processor can handle automatic scaling on all operands and multiplication.

The following instructions are implemented:

BCD Instructions

• PADD: Add packed decimal numbers with automatic scaling.
• PADDR: Same as PADD, rounding the result.
• PSUB: Subtract Packed decimal numbers with automatic scaling.
• PSUBR: Same as PSUB, rounding the result.
• PCOMP: Compare packed decimal numbers.
• PMPY: Multiply packed decimal numbers with automatic scaling.
• PMPYR: Same as PMPY, rounding the result.
• PSHIFT: Shift packed decimal number with automatic scaling.
• PSHIFTR: Same as PSHIFT with rounding.
• PPACK: Convert from ASCII coded decimal number to packed decimal number with automatic scaling.
• PPACKR: Same as PPACK, rounding the result.
• PUPACK: Convert from packed decimal number to ASCII coded decimal number with automatic scaling.
• PUPACKR: Same as PUPACK, rounding the result.
• Wn WPCONV: Convert from ordinary binary to packed decimal numbers.
• Wn PWCONV: Convert from packed decimal numbers to ordinary binary.

The instructions operate on packed decimal numbers with from 1 to 31 decimal digits. The decimal point ranges from -32 to 31 counted from the least significant digit.

What is a Packed Decimal Number (BCD-number)

Each decimal digit is represented by four bits. Therefore two digits can be placed adjacent in a byte. In the right-most byte of the the string of bytes representing your number, the last four bits are used to keep the sign. Each decimal digit is encoded as follows:

Digit	Bit Pattern	Hexa-decimal
0	0000	0
1	0001	1
2	0010	2
3	0011	3
4	0100	4
5	0101	5
6	0110	6
7	0111	7
8	1000	8
9	1001	9
+	1010	A
-	1011	B
+	1100	C
-	1101	D
+	1110	E
unsigned	1111	F

Every number will fill an integral number of bytes right justified. If the leftmost 4 bits are not needed, they will always contain a zero.

---

Page 2

What is an ASCII Coded Decimal Number

In the ASCII format each decimal digit occupies one byte (8 bits). The four high-order bits of a byte are called the zone. The four low-order bits are encoded in the same way for the packed decimal digits.

A decimal number in this format may have four different sign representations:

1. Embedded trailing sign.
   The right-most byte occupies both the least significant digit and the sign.

2. Separate trailing sign.
   The right-most byte occupies the ASCII-coded sign only.

3. Embedded leading sign.
   The left-most byte contains both the most significant digit and the sign.

4. Separate leading sign.
   The left-most byte occupies the ASCII-coded sign only.

When a sign is embedded the following codes are used:

	0	1-9
positive	173B	101B-111B
negative	175B	112B-122B

    _     _                 
  _| |_ _| |_               
 |_   _|_   _|              
   |_|   |_|                

   _   _    _                    
  | | | |  (_)                   
  | |_| | ___  ___ ___  ___  ___ 
  |  _  |/ _ \/ __/ __|/ _ \/ __|
  | | | |  __/\__ \__ \  __/\__ \
  \_| |_/\___||___/___/\___||___/

      ____  ____                 
     / ___||  _ \                
    | |    | |_) |               
    | |___ |  __/                
     \____||_|                   

NOTE: NORSK DATA reserves the right to change specifications without notice.

---