RISC-V and Floating Point
Summary
A report on the RISC-V architecture's floating point capabilities and updates.
Similar Articles
Math is hard (OpenBSD on VAX)
OpenBSD developer details the painful quirks of VAX floating-point exceptions and how they complicate kernel porting.
Novel Aspects of IEEE SA P3109 Arithmetic Formats for Machine Learning
The IEEE P3109 draft standard defines a parameterized family of binary floating-point formats and operations tailored for machine learning, supporting configurable width, precision, signedness, and infinities, along with extensive rounding modes including stochastic rounding and a novel scale-invariant approximation measure called kappa-approximation.
Conditions in the Intel 8087 floating-point chip's microcode
A detailed examination of the conditional tests used in the Intel 8087 floating-point coprocessor's microcode, part of a reverse-engineering effort to understand its algorithms.
The stack circuitry of the Intel 8087 floating point chip, reverse-engineered
The article details the reverse-engineering of the Intel 8087 floating-point coprocessor's stack circuitry, explaining how the chip's stack-based register architecture and microcode ROM enabled fast floating-point operations.
@pupposandro: https://x.com/pupposandro/status/2054241934164492328
The article announces support for DFlash and PFlash speculative decoding in llama.cpp for AMD Strix Halo iGPUs, demonstrating significant speedups in inference performance using ROCm.