C28x CPU + FPU + VCU + TMU and CLA
TMS320F2837xD Microcontroller Workshop - Architecture Overview 1 - 7
CPU Pipeline
F
1
F
2
D
1
D
2
R
1
R
2
E
C28x CPU Pipeline
Protected Pipeline
Order of results are as written in
source code
Programmer need not worry about
the pipeline
8-stage pipeline
F
1
F
2
D
1
D
2
R
1
R
2
E
F
1
F
2
D
1
D
2
R
1
R
2
E
F
1
F
2
D
1
D
2
R
1
R
2
E
F
1
F
2
D
1
D
2
R
1
R
2
E
F
1
F
2
D
1
D
2
R
1
R
2
E
F
1
F
2
D
1
D
2
R
1
R
2
E
F
1
F
2
D
1
D
2
R
1
R
2
E
A
B
C
D
E
F
G
W
W
W
W
W
W
W
W
E & G Access
same address
R
1
R
2
E W
D
2
R
1
R
2
E W
F1: Instruction Address
F2: Instruction Content
D1: Decode Instruction
D2: Resolve Operand Addr
R1: Operand Address
R2: Get Operand
E: CPU doing “real” work
W: store content to memory
H
The C28x uses a special 8-stage protected pipeline to maximize the throughput. This protected
pipeline prevents a write to and a read from the same location from occurring out of order.
This pipelining also enables the C28x CPU to execute at high speeds without resorting to
expensive high-speed memories. Special branch-look-ahead hardware minimizes the latency for
conditional discontinuities. Special store conditional operations further improve performance.
With the 8-stage pipeline most operations can be performed in a single cycle.