git404hub

what is geak fr?

aakashvarma/geak — explained in plain English

Analysis updated 2026-05-18

0PythonAudience · researcherComplexity · 5/5Setup · hard

tl;dr

GEAK uses coordinated Claude Code AI agents to automatically profile and speed up AI model serving and individual GPU kernels on AMD Instinct MI GPUs.

vibe map

mindmap
  root((GEAK))
    What it does
      GPU kernel optimization
      Model serving speedup
      AI agent orchestration
    Tech stack
      Claude Code
      ROCm
      Triton
      HIP
    Use cases
      Whole model throughput
      Single kernel speedup
      Batch GPU optimization
    Audience
      ML engineers
      AMD GPU researchers

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

what do people make with this?

VIBE 1

Automatically raise the serving throughput of a large language model running on sglang or vLLM.

VIBE 2

Speed up one specific AMD GPU kernel written in Triton, HIP, or CK without manual profiling.

VIBE 3

Get a written report explaining which optimizations were tried and why each one was accepted or rejected.

VIBE 4

Batch-optimize many GPU kernels at once across multiple GPUs with automatic GPU locking.

what's the stack?

PythonClaude CodeROCmTritonHIP

how it stacks up fr

aakashvarma/geak0xallam/my-recipe0xhassaan/nn-from-scratch
Stars00
LanguagePythonPythonPython
Last pushed2022-11-22
MaintenanceDormant
Setup difficultyhardmoderatemoderate
Complexity5/52/54/5
Audienceresearchergeneraldeveloper

Figures from each repo's GitHub metadata at analysis time.

how do i run it?

Difficulty · hard time til it works · 1h+

Requires an AMD Instinct MI GPU, ROCm 6+, and a pre-built kernel toolchain such as PyTorch or Triton set up beforehand.

in plain english

GEAK is a tool that uses AI agents to automatically speed up code that runs on AMD Instinct MI GPUs, the specialized chips used for running large AI models. It is built on top of Claude Code, Anthropic's coding assistant, and uses a system of scripted workflows to coordinate multiple AI agents working together on a performance problem, rather than relying on one agent to figure everything out unsupervised. The project ships two related tools. The main one, called e2e_workflow, looks at an entire AI model being served to real users through a serving system such as sglang or vLLM, and works to raise its overall throughput, meaning how many requests it can handle per second. It does this by first measuring where the model actually spends its GPU time, then trying the cheapest fixes first, such as adjusting configuration settings, before moving on to rewriting or generating faster low-level GPU code for the parts that matter most. The second tool, kernel_workflow, works on a single piece of GPU code at a time, called a kernel, and is what the first tool calls automatically whenever it decides a specific kernel needs to be rewritten for speed. Every change the system makes is checked before being accepted. It compares the before and after performance directly on the same hardware, confirms the output is still correct, and only keeps a change if it delivers a real, measurable improvement. Every run produces a written report describing what was tried, what worked, and what the final result was, so a person can review the reasoning afterward rather than trusting a black box. Getting started requires an AMD Instinct MI GPU, the ROCm software stack that lets you use it, and Python. You install GEAK with a single command, set up API access to an AI model, then launch Claude Code and simply describe in plain language what you want optimized, such as naming a model and its serving settings, or pointing at one specific piece of GPU code. The project reports that its kernel-level tool roughly doubled the speed gains achieved by an earlier version of the same idea, based on tests across a dozen sample kernels. This is aimed at machine learning engineers working on AMD hardware who want to speed up model serving without manually profiling and rewriting GPU code by hand.

prompts (copy fr)

prompt 1
Use path_to_GEAK/e2e_workflow to optimize inference throughput for my model served with sglang on 4 GPUs.
prompt 2
Use path_to_GEAK/kernel_workflow to optimize this specific GPU kernel and explain what changed.
prompt 3
Explain how GEAK verifies that a GPU kernel optimization actually improved performance before accepting it.
prompt 4
Help me set up ROCm and the prerequisites needed to run GEAK on an AMD Instinct MI GPU.

Frequently asked questions

what is geak fr?

GEAK uses coordinated Claude Code AI agents to automatically profile and speed up AI model serving and individual GPU kernels on AMD Instinct MI GPUs.

What language is geak written in?

Mainly Python. The stack also includes Python, Claude Code, ROCm.

How hard is geak to set up?

Setup difficulty is rated hard, with roughly 1h+ to a first successful run.

Who is geak for?

Mainly researcher.

peek the repo → explain another one

This repo across BitVibe Labs

double-check against the repo, no cap.