Blog on Akeel Ather Medina

Seeing Timbre: Ryuichi Sakamoto

Thu, 15 Oct 2026 00:00:00 +0500

I’ve been wanting to make a post about music for a long time. I’m not the best writer though, but I think about music often. Specifically, I’ve been interested in a few pieces by Ryuichi Sakamoto. They’re all unique in their own way, and at first I wanted to explore the phenomenon of his composition to describe a real object. For example, give a listen to Obsidian from Crystalline. It sounds like Obsidian. I don’t have the words to describe how that is. So, I’ve tried to turn some of his songs into images. For this post, I’ve used these:

Reflections on my MSc Thesis

Fri, 11 Sep 2026 00:00:00 +0500

I completed my MSc thesis on the 22nd of August, and left Edinburgh on the 26th. Everything went by in a blur, after a life changing year abroad. Today is the 11th of September, it’s been just over 2 weeks, and I wanted to write a short reflection piece. You can read my thesis if you’re interested in profiling KV cache offload, I’ll be making it available as soon as it is graded.

Reflections on ISC SCC 2026

Thu, 10 Sep 2026 00:00:00 +0500

As a participant representing TeamEPCC in the ISC Student Cluster Competition 2026, I’ve recently had the chance to work on a custom-built HPC cluster that we shipped to Hamburg for the in-person competition. I wanted to give my overall reflections on the competition and my experience.

The SCC has never included llama.cpp as a micro-benchmark, usually focusing on HPC-specific benchmarks like HPL and HPCG. This year introduced llama.cpp, which allows LLM inference with minimal setup on a wide range of hardware. It is a plain C/C++ implementation with no dependencies. It runs surprisingly well on Apple silicon! Understanding how to use llama.cpp is fairly simple, but maximizing performance for specific hardware could be a whole MSc thesis. The source code is very well written, and it is easy to find and read specific implementations of custom kernels. But the majority of performance comes from tuning the runtime parameters, so understanding the workload, and how those parameters change the character of your problem on your hardware, is what actually matters. That makes it quite different from the other benchmarks we were given, like HPL.

C++ Design Patterns

Thu, 05 Jun 2025 19:11:23 +0500

Based on the following repository - cpp-design-patterns

Background

I was recently browsing this repository on Github, and was inspired to do something similar. I hadn’t written C++ in over 2 years, since taking a course in GPU-Accelerated Computing. But, with my Masters starting in a few months, I decided to start looking in to Modern C++. I noticed this library had some interesting usage of static, and it was the first thing I wanted to investigate. I went down a rabbit hole of the static-initialization order, this video by The Cherno (I got through my course on OOP 5 years ago because of him!), and Scott Meyers’ Effective C++ (Item 4), where the Meyers Singleton originates, alongside the Factory Creational Pattern.

Understanding Simultaneous Multithreading (SMT)

Tue, 06 May 2025 00:00:00 +0500

While reading Performance Analysis and Tuning on Modern CPUs by Denis Bakhvalov, I had a surprising realization. Before this book, I had a completely incorrect understanding of what a thread is. Somewhere in my mental model of CPU microarchitecture, the “thread” was a muddy concept that vaguely correlated to a physical core, but it was more of an abstraction. I never interacted much with them because of Python’s Global Interpreter Lock (GIL), well-known for nulling any performance gain from SMT in compute-bound workflows, which was my main use case.