Intel® Agilex™ Variable Precision DSP Blocks User Guide

ID 683037
Date 11/17/2022
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents

10.3.1. General Tab

Table 108.  General Tab
Parameter IP Generated Parameter Value Default Value Description
Operation Mode
Choose the operation mode operation_mode

fp32_mult

fp32_add

fp32_mult_add

fp32_mult_acc

fp32_vector1

fp32_vector2

fp16_sumof2mult

fp16_sumof2mult_add_fp32

fp16_sumof2mult_acc

fp16_vector1

fp16_vector2

fp16_vector3

fp32_mult_add Select the desired floating-point operation mode.
Enable fp32_chainin use_chainin

No

Yes

No

Select to enable chainin feature.

When you enable the chainin feature, the result from the multiplier is added or subtracted by the input from chainin port.

Enable fp32_chainout enable_chainout

No

Yes

No Select to enable the chainout port.
FP32 Operation
Perform subtraction in fp32_adder fp32_adder_subtract

No

Yes

No

Select Yes to set FP32 adder to perform subtraction.

Select No to set FP32 adder to perform addition.

FP16 Representation/Operation
Select the mode for fp16 fp16_mode

FLUSHED

EXTENDED

BFLOAT16

FLUSHED Select the precision format for FP16 operation modes.
Select the width size for fp16 (Only for bfloat16 mode) fp16_input_width

16

19

16 Specify the width of FP16 data input bus.
Perform subtraction in fp16_adder fp16_adder_subtract

No

Yes

No

Select Yes to set FP16 adder to perform subtraction.

Select No to set FP16 adder to perform addition.

Exception Flag
Enable exception flag enable_exception_flag

No

Yes

No Select to enable exception flag feature.