Aeroflex Circuit Technology SCD5271 REV 1 12/22/98 Plainview NY (516) 694-6700
2
DESCRIPTION
The Aeroflex ACT5271 is a highly integrated
superscalar microprocessor that implements a
superset of the MIPS IV Instruction Set
Arch itect ur e(ISA ). It h as a h ig h perf orm an ce 64 -bit
integer unit, a high throughput, fully pipelined 64-bit
floating point unit, an operating system friendly
memory management unit with a 48-entry fully
associa tive TLB, a 32 K Byte 2 -way set associative
instr uction cac he, a 32 KByte 2 -way set assoc iative
data cache , and a high-perfor mance 64-bi t system
interface with support for an optional external
secondary ca che. The ACT5271 can issue both an
inte ger and a f loating point inst ruction in the sa me
cycle.
The ACT5271 is ideally suited for high-end
embedded control applications such as
internetworking, high performance image
manipulation, high speed printing, and 3-D
visualization.The ACT5271 is also applicable to the
low end workstation market where its balanced
integer and floating-point performance and direct
support for a large secondary cache (up to 2MB)
provide outstanding price/performance
HARDWARE OVERVIEW
The ACT5271 offers a high-level of integration
targeted at high-performance embedded
applications. The key elements of the ACT5271
are brief ly des c ribed below.
Superscalar Dispatch
The ACT5271 has an efficient asymmetric
superscalar dispatch unit which allows it to issue an
inte ger ins t ructi on an d a fl oating-poi nt c om putat ion
instruction simultaneously. With respect to
superscalar issue, integer instructions include alu,
branch, load/store, and floating-point load/ store,
while floating-point computation instructions
include floating-point add, subtract, combined
multiply-add, converts, etc. In combination with its
high throughput fully pipelined floating-point
execution unit, the superscalar capability of the
ACT5 27 1 pro vide s un pa rallele d p rice /perfo rman ce
in computationally intensive embedded
applications.
CPU Regist ers
Like all MIPS ISA processors, the ACT5271 CPU
has a simple, clean user visible state consisting o f
32 ge neral purpos e regis te rs, tw o specia l purpo se
registers for integer multiplication and division, a
program c ounter, and no c ondition code bits .
Pipeline
For integer operations, loads, stores, and other
non-floating-point operations, the ACT5271 uses
the simple 5-stage pipeline also found in the
ACT52xx family, R4600, R4700, and R5000. In
addition to this standard pipeline, the ACT5271
uses an extended seven stage pipeline for
floating-point operations. Like the ACT5270 and
R5000, the ACT5271 does virtual to physical
tra ns lati on in par allel wit h c ac he ac c es s .
Integer Unit
Like the other members of the ACT52xx family
and R5000, the ACT5271 implements the MIPS IV
Instruction Set Architecture, and is therefore fully
upward compatible with applications that run on
processors implementing the earlier generation
MIPS I-III instruction sets. Additionally, the
ACT5271 includes two implementation specific
ins truction s not fou nd in the baselin e MIPS I V ISA
but that are usef ul in the embe dded marke t pla ce.
De s c ribed in de t ail in the QE D R M 5271 da t as heet,
these instructions are integer multiply-accumulate
and 3-operand integer multiply.
The ACT5271 integer unit includes thirty-two
general purpose 64-bit registers, a load/store
architect ure with single cycle ALU operations (add,
sub, logical, shift) and an autonomous multiply/
divide unit. Additional register resources include:
the HI/LO result registers for the two-operand
int eger m ultiply / div ide op erat ions, a nd t he pro gram
counter(PC).
Register File
The ACT5271 has thirty-two general purpose
registers with register location 0 hard wired to zero.
These registers are used for scalar integer
operations and address calculation. The register
file has two read ports and one write port and is fully
bypassed to minimize operation latency in the
pipeline.
ALU
The ACT5271 ALU consists of the integer adder/
subtractor, the logic unit, and the shifter. The adder
performs address calculations in addition to
arithmetic operations, the logic unit performs all
logical and zero shift data moves, and the shifter
performs shifts and store alignment operations.
Each of these units i s opti mi zed to perf orm all ti o ns
in a s ingle processo r c y cl e.
For additional Detail Information regarding the
operation of the Quantum Effect Design (QED)
RISCMark ACT5271, 64-Bit Superscalar
Microprocessor see the latest QED datasheet
(R ev is ion 1. 0 J uly 1998).