https://pranav083.github.io/tags/chipyard/Recent content in Chipyard on Pranav KumarHugo -- gohugo.ioenCopyright © 2016–2030, Pranav KumarSun, 15 Dec 2024 00:00:00 -0500https://pranav083.github.io/post/project/s1-ece5504-cache-optimization/Sun, 15 Dec 2024 00:00:00 -0500https://pranav083.github.io/post/project/s1-ece5504-cache-optimization/ <p><strong>ECE 5504 — Computer Architecture · Virginia Tech · Fall 2024</strong></p> <p><strong>Team project</strong> with Alok Mehendale and Benjamin Rice. Benchmarks were divided between us: Alok took <code>rsort</code> and <code>vvadd</code>, Benjamin took <code>towers</code> and <code>median</code>, and <strong>I took FFT and SHA-256</strong>. Results were combined for the final report.</p> <hr> <h2 id="problem">Problem</h2> <p>General-purpose processor configurations are a compromise. Given an open, parameterised SoC generator, how much performance can be recovered by tailoring the microarchitecture to a <em>specific</em> workload — and where does that tailoring stop paying off?</p>