---
title: "Jay Gala (@Jaygala223) | Hashnode"
description: "AI Software Engineer at Intel. Read the latest articles by Jay Gala on Hashnode."
url: "https://hashnode.com/@Jaygala223"
---

@Jaygala223

# Jay Gala

@Jaygala223 · Mumbai, India · Joined December 2022

AI Software Engineer at Intel

[Website](https://www.leetcode.com/jaygala223)

### [I Wrote a GPU Matmul Kernel From Scratch in Triton. Here's Everything I Learned](https://galacodes.hashnode.dev/i-wrote-a-gpu-matmul-kernel-from-scratch-in-triton-here-s-everything-i-learned?utm_source=hashnode&utm_medium=feed)

I recently started learning Triton, OpenAI's Python-based language for writing GPU kernels. My project: build a matrix multiplication kernel from scratch, step by step, until it's competitive with PyT

15 min read · [galacodes.hashnode.dev](https://galacodes.hashnode.dev/i-wrote-a-gpu-matmul-kernel-from-scratch-in-triton-here-s-everything-i-learned?utm_source=hashnode&utm_medium=feed)

[Jun 23](https://hashnode.com/posts/i-wrote-a-gpu-matmul-kernel-from-scratch-in-triton-here-s-everything-i-learned/6a3a18393c0ead73f6b0dd07)

### [GPUs: The Hardware That Power AI](https://galacodes.hashnode.dev/gpus-the-hardware-that-power-ai?utm_source=hashnode&utm_medium=feed)

GPUs revolutionized AI with parallel computing, transforming video game technology for models like ChatGPT and Claude

18 min read · [galacodes.hashnode.dev](https://galacodes.hashnode.dev/gpus-the-hardware-that-power-ai?utm_source=hashnode&utm_medium=feed)

[Nov 30, 2025](https://hashnode.com/posts/gpus-the-hardware-that-power-ai/692c032a7f0960cfc2a4a799)

### [Speculative Decoding: From Theory to Implementation](https://galacodes.hashnode.dev/speculative-decoding?utm_source=hashnode&utm_medium=feed)

The complete guide to understanding the concept of speculative decoding in LLM inference and implementing it from scratch

15 min read · [galacodes.hashnode.dev](https://galacodes.hashnode.dev/speculative-decoding?utm_source=hashnode&utm_medium=feed)

[1](https://hashnode.com/posts/speculative-decoding/690b1ada1921d85b235d9b58) [Nov 5, 2025](https://hashnode.com/posts/speculative-decoding/690b1ada1921d85b235d9b58)

### [Bits Don't Lie: Datatypes in Modern LLMs](https://galacodes.hashnode.dev/bits-dont-lie-datatypes-in-modern-llms?utm_source=hashnode&utm_medium=feed)

Let’s talk about the different datatypes that are being used in modern LLMs like GPT, LLaMa and the like. The most common ones that you might have heard: FP32, BFloat16, Float16, INT8, etc. These are all standard data types and available in PyTorch a...

6 min read · [galacodes.hashnode.dev](https://galacodes.hashnode.dev/bits-dont-lie-datatypes-in-modern-llms?utm_source=hashnode&utm_medium=feed)

[May 30, 2025](https://hashnode.com/posts/bits-dont-lie-datatypes-in-modern-llms/6839a06496fd310e54f38d84)

### [Introduction to LLM inferencing](https://galacodes.hashnode.dev/introduction-to-llm-inferencing?utm_source=hashnode&utm_medium=feed)

A complete introduction to LLM Inference with KV Cache and sample code.

7 min read · [galacodes.hashnode.dev](https://galacodes.hashnode.dev/introduction-to-llm-inferencing?utm_source=hashnode&utm_medium=feed)

[Oct 26, 2024](https://hashnode.com/posts/introduction-to-llm-inferencing/671c905ccd3f2679dc35ab00)
