New in 26.09Detection
YOLO26 End-to-End Detection
Compiles the entire NMS-free YOLO26 detector into a single int8 program on the Chimera GPNPU, with CCL custom-op kernels for the PSA attention blocks and the...
SDK 26.09Open the demo
Every network running on the Chimera GPNPU, newest and most complete first. 400+ models, demos, and building blocks, each with a runnable notebook or the full source graph behind it.
Compiles the entire NMS-free YOLO26 detector into a single int8 program on the Chimera GPNPU, with CCL custom-op kernels for the PSA attention blocks and the...
Sarashina2.2-3B quantized to W4A16 (SmoothQuant, int4 weight-only) and run as one Chimera kernel with custom attention and projection ops, gated against ONNX...
Compiles the INT8 WaveFormer sEMG gesture-classification encoder through CGC (only the attention core stays a custom op) and validates ISS logits against ONNX...
INT8 ModernBERT-base masked language model compiled for a single-core QC-N with 1 MB OCM through nine custom ops, validated on the ISS against ONNX Runtime.
Runs W4A8 Qwen3-8B autoregressive decode on a 4-core QC-P ISS at ~18-19 tokens/s, validates a prefill decoder against ONNX Runtime and compiles the full...
Runs the PI0.5 vision-language-action pipeline as one fused CCL kernel on the Chimera GPNPU, with a live SigLIP vision tower feeding LM prefill and the Action...
Quantizes and compiles YOLOv8-Pose for the Chimera GPNPU and runs batched pose-estimation inference, drawing bounding boxes and keypoints on the host.
Quantizes, compiles, and runs YOLOv8 object detection on the Chimera GPNPU with the Instruction Set Simulator, splitting the graph to keep postprocessing on...
Runs the static W8A8 SmoothQuant pipeline that turns a HuggingFace Qwen2.5-0.5B-Instruct checkpoint into a fixed-shape INT8 ONNX for the Chimera GPNPU, with...
Create a free DevStudio account to search the full catalog, open the demo notebooks, and see cycle-accurate benchmarks, source code, and intermediate representations for every model.