HDL Coder™ supports math and trigonometric functions, and several Simulink® blocks with native floating-point technology. You can customize the latency of these operators.
The table shows a list of floating-point operators that HDL Coder supports, the maximum and minimum latency, and the units in the last place (ulp) error.
| Function | Minimum Output Latency | Maximum Output Latency | Units in the Last Place (ULP) error |
|---|---|---|---|
| Addition/Subtraction | 7 | 12 | 0 |
| Multiplication | 8 | 8 | 0 |
| Division | 32 | 32 | 0 |
| Absolute value | 0 | 0 | 0 |
| Unary minus | 0 | 0 | 0 |
| Minimum/maximum | 3 | 3 | 0 |
| Relational operators | 3 | 3 | 0 |
| Reciprocal | 19 | 19 | 1 |
| Reciprocal square root | 17 | 17 | 2 |
| Square root | 28 | 28 | 2 |
| Exponential | 23 | 23 | 3 |
| Power of two | 2 | 2 | 0 |
| Natural logarithm | 20 | 20 | 3 |
| Numeric to Single data type conversion | 6 | 6 | 0 |
| Single to numeric data type conversion | 6 | 6 | 0 |
| Trigonometric sine | 27 | 27 | 1 |
| Trigonometric cosine | 27 | 27 | 2 |
| Arctangent | 36 | 36 | 2 |
| Arctangent 2 (ArcTangent (input1 / input2)) | 42 | 42 | 3 |
HDL Coder supports several Simulink blocks including math and trigonometric blocks. In the Math Operations library, these blocks are supported:
Math Function where Function can be:
reciprocal
log
exp
Trigonometric Function where Function can be:
sin
cos
sincos
cos + jsin
atan
atan2
The table shows the list of supported blocks in other block libraries.
| Block Library | Supported blocks |
|---|---|
Discrete | Zero Order Hold and the set of delay blocks including Integer Delay and Tapped Delay. |
HDL Operations | Multiply-Add and HDL Counter |
HDL Subsystems | All blocks are supported. This library includes the State Control block and subsystems that use enable and reset ports with the State Control block. |
Logic and Bit Operations | Compare To Constant and Compare To Zero |
Lookup Tables | Prelookup, direct Lookup Table (n-D), and n-D Lookup Table blocks. |
Model Verification | All blocks are supported. This library includes blocks such as Assertion and Check Dynamic Range. |
Model-Wide Utilities | All blocks are supported. This library includes the DocBlock and Model Info blocks. |
Ports & Subsystems | Enable, reset, input, and output ports, and model references, and subsystem blocks are supported. |
Signal Attributes | All blocks are supported. This library includes blocks such as Bus to Vector and Data Type Propagation. |
Signal Routing | All blocks are supported. This library includes blocks such as Mux and Bus Selector. |
Sources | Inport, Constant, and Ground blocks. |
Sinks | All blocks are supported. This library includes blocks such as Display, To File, and Spectrum Analyzer. |
Biquad Filter
Dead Zone and Dead Zone Dynamic
Delay block in RAM-based mode or with frame-based input.
Switch block with input to the control port as a floating-point type.
Dot Product in complex mode with Architecture as Tree or Linear.
Sum of Elements with complex input types.
Trigonometric Function with Function as sin or cos in CORDIC Approximation
method.
MinMax block with mode other than value.
MATLAB System blocks.
Polar to Cartesian
Discrete FIR Filter with CoeffMultipliers HDL
block property set to csd or factored-csd.
For the Data Type Conversion block:
Stored Integer (SI) mode
for Input and output to have equal setting is
not supported.
Saturate on integer overflow check box must be left cleared.
Note:
For blocks that support |
HDL Coder does not support:
Complex and vector data types.
Blocks operating at a continuous sample time.
Double data types.
The streaming optimization.