Back
CNN Acceleration on Analog Compute
HPC
AI & ML
Staff Aug
The ACCELR team has expertise in building hardware accelerated high performance computing systems on truly exotic computes. This is demonstrated by our contribution to the AI modeling exercises and CNN inference acceleration on low power analog compute. Talk to our team for more information on our AI acceleration experience.
Modeling of CNN network behavior on Analog Compute
High-level modeling of quantization noise and it's impact on CNN inference
CNN Inference Acceleration
Contributed to building frontend and backend SDK components such as functional models, kernels, run-time components.
