Skip to content
View jakedves's full-sized avatar
11 followers · 20 following
  • London/Edinburgh

Achievements

Achievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: Pull Sharkx2Achievement: YOLO

Achievements

Achievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: Pull Sharkx2Achievement: YOLO

Highlights

  • Pro

Block or report jakedves

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
jakedves/README.md

Jake Davies 👨🏻‍💻

I'm a PhD student who's interested in computer architecture, high performance computing, and deep learning. Knowledgable about parallel systems, including CPU and GPU architectures, and my work focuses on the compilation stack for RISC-V accelerators.

Recent Projects

GPU Relaxation with Metal Compute

An implementation of the relaxation technique for solving differential equations using the Metal Compute API for Apple GPUs.

ReadMe Card

Deep Reinforcement Learning for NES Super Mario Bros.

Implemented RainbowDQN with CUDA acceleration (via PyTorch) to train an agent to play Super Mario Bros, achieving the highest grade ever on the unit with 92%.

ReadMe Card

Full Game Performance
Full Game Performance 		Multi-level Training 1
Multi-level Training 1 		Multi-level Training 2
Multi-level Training 2 		Multi-level Training 3
Multi-level Training 3

Advanced Raytracer

Raytracer built from scratch in C++, featuring parallelism, bounding volumes, global illumination (photon mapping), and more for a university coursework. Achieved an almost-perfect grade of 99%.

ReadMe Card

Description

Final render with photon mapping

Languages and Tools

I'm pretty open to learning cool and modern languages, like Rust or Go, for now I'm very comfortable with:

  • C, C++
  • Python
  • Swift
  • Java

Pinned Loading

  1. advanced-raytracer advanced-raytracer Public

    Description of how my raytracer was implemented in C++

    C++

  2. deep-reinforcement-learning deep-reinforcement-learning Public

    The post training video of our agent performance

    Python 1

  3. visionary visionary Public

    Computer Vision based approach to using macOS (demo in README).

    Swift 5

  4. relaxation-metal relaxation-metal Public

    Implementation of the relaxation technique in Swift and Metal compute shaders

    Metal