Visible to Intel only — GUID: iov1691736827429
Ixiasoft
4.3.1. General-Purpose Register File
4.3.2. Arithmetic Logic Unit
4.3.3. Multipy and Divide Units
4.3.4. Floating-Point Unit
4.3.5. Custom Instruction
4.3.6. Reset and Debug Signals
4.3.7. Control and Status Registers
4.3.8. Trap Controller
4.3.9. Memory and I/O Organization
4.3.10. RISC-V based Debug Module
4.3.11. Error Correction Code (ECC)
4.3.12. Branch Prediction
4.3.13. Lockstep Module
Visible to Intel only — GUID: iov1691736827429
Ixiasoft
4.3.4. Floating-Point Unit
The floating-point unit (FPU) implements the single precision floating point instructions. The FPU operates on data stored in thirty-two 32-bits floating-point registers, implemented using M20K memories.
Below are the characteristics of the FPU:
- Based on RISC-V “F” Standard Extension for Single-Precision Floating-Point
- Supports floating-point fused multiply-add instructions.
- IEEE 754-2008 compliant except for:
- Simplified rounding
- Subnormal supported on a subset of operations
- Consumes resource in a typical system as below1:
- 960 ALMs
- Five M20Ks memories
- Five DSP blocks
Note: The Nios® V/g processor adopts the GNU floating point software emulation for double precision floating point operation.
1 System using Arria® 10 FPGA devices.