Hardware and deep learning
Across
- 3. The place to store activations in a tiny device
- 6. Process of identify suitable CNN architecture
- 8. The non-dominated points during DSE
- 12. Using INT8 instead of FP32 representation
- 14. Determines the range of a floating point number
Down
- 1. An array structure used in most CNN accelerators
- 2. A common objective metric used in NAS
- 4. A CNN which proposed depthwise and pointwise convolutions
- 5. Helps map a CNN graph efficiently to given hardware
- 7. Removing unimportant weights from a CNN model
- 8. The layer which helps reduce the input dimensions for next layer
- 9. A commonly used bio-inspired algorithm for DSE
- 10. A lossless compression scheme used in Eyeriss
- 11. Accelerators help to reduce this metric
- 13. A device to implement custom digital logic