RISC-V and Floating Point
Summary
A report on the RISC-V architecture's floating point capabilities and updates.
Similar Articles
Moving integer division to floating-point is trivial
A technical blog post explains how to perform integer division and remainder using floating-point division and fused multiply-add, with constraints on operand bit widths, and discusses practical considerations for SIMD and rounding modes.
Intermediate Floating-Point Precision
This article explores how intermediate floating-point precision in C++ code depends on compiler settings, CPU flags, and architecture, particularly on x87 FPU, and how this affects performance and calculation results.
Faster floating point math with Rust’s new API
Rust 1.98 introduces a new API to enable faster floating-point math by allowing more aggressive compiler optimizations, while still giving developers control over rounding errors.
Pushing the limits of RISC-V emulation
This blog post explores how to make RISC-V execution faster on non-RISC-V machines using an ahead-of-time recompiler that connects basic blocks with tail calls and leverages Clang's preserve_none calling convention, as part of the OpenVM project at Axiom.
Formally Verified Synthesizable Floating-Point Data Types in ARCH HDL
This paper presents the design and end-to-end formal verification of IEEE-754 binary32 and bfloat16 arithmetic for ARCH HDL, a hardware description language intended for AI model generation. The operators are proven correctly rounded using a hybrid approach combining exhaustive SMT equivalence checking and Lean 4 proofs, with synthesizable SystemVerilog output.