Shipping Competitive RTL with Open Source Tools
A phase-by-phase RTL workflow that keeps modeling, generation, verification, synthesis, and timing open without closing the door to commercial backends.
HardMatrix
Featured Posts
Latest Posts
Building a PyTorch Backend for a Custom Accelerator
Custom AI hardware creates value when repeated model operations can be moved into workload-specific datapaths for better latency, power, and integration. This post shows a small PyTorch-facing prototype that connects ordinary tensor code to that hardware path.
Shipping Competitive RTL with Open Source Tools
A phase-by-phase RTL workflow that keeps modeling, generation, verification, synthesis, and timing open without closing the door to commercial backends.
Lightweight RTL Verification Flow with cocotb
A lightweight verification workflow for building RTL with golden models, standard streaming interfaces, and composable blocks.
Parallelizing Linear Feedback Systems with Matrices
Many “serial feedback” circuits look inherently sequential, but if their state update is linear over a finite field, we can raise the transition matrix to process many input symbols per clock.