Predict
Compare
Optimize
Hardware
Visualize
Explain
KPG Visualization
Interactive Kernel Pipeline Graph — see every GPU kernel in the inference pipeline.
Model
Loading...
Runtime
vLLM
SGLang
TensorRT-LLM
By GPU Type
By AWS SKU
GPU Type
Loading...
Number of GPUs
1
2
4
8
AWS Instance Type
Loading...
Precision
FP16
BF16
FP8
TP Degree
Generate Visualization
Processing...