in-Memory Analog Matrix-Vector-Multiplier for DNN Acceleration.
Ioannis A. PapistasStefan CosemansBram RooseleerJonas DoevenspeckMyung Hee NaArindam MallikPeter DebackerDiederik VerkestPublished in: CICC (2021)
Keyphrases
- sparse matrix
- eigenvalues and eigenvectors
- memory usage
- matrix representation
- floating point
- memory requirements
- singular value decomposition
- memory space
- transformation matrix
- dot product
- linear algebra
- rows and columns
- linearly independent
- successive approximation
- fuzzy membership
- random access
- hardware implementation
- fixed point
- data sets