These reference models are already compiled end-to-end. Download their artifacts, or log in below to compile your own.
// demo access
Run your own model through the compiler.
Sign up or log in to upload an ONNX model. It runs on our AWS server and streams progress back live — choose an AWS F2 FPGA load package (DCP + AFI) to deploy on your own FPGA, or a sky130 GDSII layout for full-custom silicon.
Real accounts via AWS Cognito (ap-southeast-2). Check your email for a verification code after signup.
// upload model
Upload your ONNX model.
Signed in as —. Supported: .onnx files (CNN / YOLO-family detectors), up to 200MB. Your file is sent to our AWS server on run — progress updates live, and you get an AWS-ready FPGA load package or a sky130 GDSII layout at the end.
⌁
Drag and drop your .onnx file here, or
// pipeline run
Compiling model.onnx
Running on our AWS server — this page polls the job and updates each stage live. Output is an AWS F2 load package (DCP + AFI), or a sky130 GDSII layout.