Intuitive Random Selection (with Marc LeBlanc)
The post introduces a new arrayless random selection algorithm presented by Marc “Mahk” LeBlanc. Traditional approaches either build a filtered array—consuming memory and cache bandwidth—or use a replace‑retain statistical method that relies on costly modulus operations and complex fairness proofs. LeBlanc’s technique removes these drawbacks, offering a memory‑efficient, uniformly random selector without the need for extra calculations. The video walks through the evolution from a naïve selector to the refined algorithm.
How to Use uops.info
The twelfth video of the Performance‑Aware Programming series walks readers through uops.info, a web resource that catalogs micro‑operation (uop) breakdowns for modern CPUs. It explains how to navigate the site, search for specific x86 instructions, and interpret the latency, port...
The Ethics of Generative AI
In episode 4 of Waiting Through AI, host and non‑expert discuss with longtime AI specialist Dimitri the ethical dimensions of generative AI creating artifacts such as images, text, and code. They distinguish legal compliance from moral considerations, using AI‑generated art...
In-Order Interleaving
The tenth video in the Performance‑Aware Programming series, titled “In‑order Interleaving,” explains how arranging dependent instructions in a strict order can minimize pipeline stalls and boost execution speed. The tutorial walks through practical code examples hosted in the accompanying GitHub...
My Thoughts on Notch's DLSS Post
In this episode the host breaks down Notch's controversial tweet that "DLSS fundamentally makes no sense," exploring the technical and economic arguments behind the claim. He explains how DLSS uses GPU silicon for both rendering and neural‑network‑based upscaling or frame...
Should You Be A Carpenter?
In the inaugural episode of Wading Through AI, host and non‑expert conversationalist teams up with veteran AI researcher Dimitri Spanos to explore whether knowledge‑based careers are at risk from AI, humorously asking if everyone should become a carpenter. Dimitri outlines...
Dependency Chain Stalls
The ninth video in the Performance‑Aware Programming series highlights that a CPU’s ability to extract instruction‑level parallelism is bounded by dependency chains. When later instructions must wait for earlier results, the pipeline stalls, limiting throughput. The post underscores the need...