Research Scientist, Systems

San Francisco /
Research /
Full-time
Company Description 

Adept is an ML research and product lab building general intelligence by enabling humans and computers to work together creatively. We’ve raised a $65M Series A from Greylock and Addition and several angel investors, and were recently highlighted by Fortune

Check out our demo and blog to see what the future holds.

Team Culture 

We're looking for team members who are energized by our ambitious mission and excited to join a fast-paced startup environment, working closely together in our new office in San Francisco.

About the role
At Adept, we use large amounts of computational resources for neural net training and inference and strive to constantly improve our ability to use our compute resources. This role will work with other researchers to co-design systems improvements (such as 8-bit or lower training) in conjunction with network architecture.

Looking into the future, we also want to make sure we pick the right computational platforms that match our current and projected needs. In this role you would help us explore the future hardware landscape and make recommendations based on the kinds of networks we believe hold the most promise.

Prior experience with Neural Networks is not necessary, but a desire to learn about them is.

What you'll achieve

    • Write custom efficient CUDA code
    • Optimize communication algorithms
    • Change the neural network architecture
    • Understand the numerics of reduced precision and train neural networks at reduced precision
    • Optimize or modify the internals of PyTorch
    • Improve GEMM performance over CuBLAS
    • Implement GPU kernels for 8-bit inference and training

Reasons you might be a fit

    • You enjoy figuring out why something isn’t fast and then making it faster
    • You understand how hardware (especially GPUs) works at the lowest levels and are able to take advantage of this knowledge by coding in C++/CUDA and assembly/ptx
    • You are adept at using profilers to debug performance problems across the stack (networking, memory, compute, disk, etc...) and able to devise solutions once issues are identified

Skills you'll need to bring

    • Experience with high-performance networking: MPI, NCCL, RDMA, Infiniband, RoCE, etc.
    • Experience with GPUs in large scale clusters
    • Experience with virtualization and container architecture in cloud environments
    • Strong drive and autonomy to own problems and build solutions from end-to-end
This role reports to Erich Elsen 

Others you will learn from:  
Anmol Gulati
Ashish Vaswani
Augustus Odena
David Luan
Fred Bertsch 
Kelsey Schroeder 
Maxwell Nye 
Niki Parmar

Adept is an equal opportunity employer. We're excited about candidates who will raise the bar of our team, regardless of specific experiences -- we encourage applicants from a range of backgrounds to apply.