PACE1750A
SINGLE CHIP, 15MHz to 40MHz,
CMOS 16-BIT PROCESSOR
FEATURES
Implements the MIL-STD-1750A Instruction Set
Architecture
Single Chip PACE Technology CMOS 16-Bit
Processor with 32 and 48-Bit Floating Point
Arithmetic
DAIS Instruction Mix Execution Performance
Including Floating Point Arithmetic
15, 20, 30, and 40 MHz Operation over the
Military Temperature Range
Extensive Error and Fault Management and
Interrupt Capability
24 User Accessible Registers
Single 5V ± 10% Power Supply
TM
Power Dissipation over Military Temperature
Range
1.3 MIPS at 20 MHz
1.9 MIPS at 30 MHz
2.6 MIPS at 40 MHz
< 0.30 watts at 20 MHz
< 0.35 watts at 30 MHz
< 0.40 watts at 40 MHz
TTL Signal Level Compatible Inputs and
Outputs
Multiprocessor and Co-processor Capability
Built-In Function (BIF) for User Defined
Instructions
Two programmable Timers
Available in:
– 64-Pin DIP or Gull Wing (50 Mil Pin Centers)
– 68-Pin Pin Grid Array (PGA)
– 68-Lead Quad Pack (Leaded Chip Carrier)
Integer DAIS Mix Performance
3.9 MIPS at 40 MHz
Conventional Integer Processing Mix
Performance
5.0 MIPS at 40 MHz
Instruction Execution at 40 MHz over the
Military Temperature Range
0.10 µsec Integer Add/Sub
0.57 µsec Integer Multiply
0.70 µsec Floating Point Add/Sub
1.07 µsec Floating Point Multiply
GENERAL DESCRIPTION
levels of prioritized internal and external interrupts, and a
faults and exceptions handler controlling internally and
externally generated faults.
The PACE1750A is a general purpose, single chip, 16-bit
CMOS microprocessor designed for high performance
floating point and integer arithmetic, with extensive real
timeenvironmentsupport. Itoffersavarietyofdatatypes,
includingbits, bytes, 16-bitand32-bitintegers, and32-bit
and48-bitfloatingpointnumbers. Itprovides13addressing
modes, including direct, indirect, indexed, based, based
indexed and immediate long and short, and it can access
2 MWords of segmented memory space (64 KWords
segments).
Themicroprocessorachievesveryhighthroughputof2.6
MIPS for a standard real time integer/floating point
instructionmixata40MHzclock. ItexecutesintegerAdd
in 0.1 µs, integer Multiply in 0.575 µs, Floating Point Add
in 0.7 µs, and Floating Point Multiply in 1.075 µs, for
register operands at a 40 MHz clock speed.
ThePACE1750Ausesasinglemultiplexed16-bitparallel
bus. Status signals are provided to determine whether
the processor is in the memory or I/O bus cycle, reading
and writing, and whether the bus cycle is for data or
instructions.
The PACE1750A offers a well-rounded instruction set
with 130 instruction types, including a comprehensive
integer,floatingpoint,integer-to-floatingpointandfloating
point-to-integer set, a variety of stack manipulation
instructions, high level language support instructions
such as Compare Between Bounds and Loop Control
Instructions. It also offers some unique instructions such
as vectored l/O, supports executive and user modes, and
providesanescapemechanismwhichallowsuser-defined
instructions using a coprocessor.
Thebasicbuscycleis4clockslong. ThePACE1750Awill
extend the cycle by insertion of wait states in the address
and data phases (in response to RDYA and RDYD
signals, repectively) and will hold the machine in HI-Z if
this CPU has not acquired the bus. A typical non-bus
cycle is three clocks long. However, variable length
cyclesareusedforsuchrepetitiveoperationsasmultiply,
divide, scale and normalize, reducing significantly the
numberofCPUCLOCKSperoperationstepandresulting
in very fast integer and floating point execution times.
The chip includes 16 general purpose registers, 8 other
user-accessible registers, and an array of real time
application support resources, such as 2 programmable
timers, a complete interrupt controller supporting 16
Do c um e nt # MICRO-3 Re v. C
Re vise d Oc to b e r 2005