Architectural Implications of Neural Network Inference for High Data-Rate, Low-Latency Scientific Applications

Open in new window