VLIW4 SIMD Engine
Configuration
This compute unit implements the TeraScale 3 architecture. It consists of a single block with the following execution units.
| Data type | Execution rate |
| FP32 | VLIW unit with 16 lanes, executing 4 operations/cycle |
|---|---|
| INT32 | VLIW unit with 16 lanes, executing 4 operations/cycle |
| FP64 | 16 lanes, executing one operation per cycle |
Block diagram
ALU
ALU
ALU
ALU
ALU
ALU
ALU
ALU
ALU
ALU
ALU
ALU
ALU
ALU
ALU
ALU
DP
DP
DP
DP
DP
DP
DP
DP
DP
DP
DP
DP
DP
DP
DP
DP