devices and models >> 2 HW Platforms ZCU102 ZCU104 Ultra96 PYNQ Face detection Pose estimation Video analytics Lane detection Object detection Segmentation DNN Specific Instruction Set Convolution, Max Pool etc. ˃ Any Network, Any Image Size ˃ High Frequency & High Compute Efficiency ˃ Supported on U200 – 3 Instances U250 – 4 Instances Amazon F1 ˃

running at faster than real-time - Compute split between GRU units and FC layers - 24kHz sampling frequency requires 40us sampling net runtime - First PyTorch model used a 3,400us sampling net runtime

experts will be distributed across multiple devices. For each token, its MoE-related communication frequency is proportional to the number of devices covered by its target experts. Due to the fine-grained

traffic / purchase history analytics. Different companies may define ‘users’ differently based on frequency. Source: Statcounter (2/25), Google (5/25), Meta 10Q (4/25), Apple (1/25), TikTok (7/21), LinkedIn