Energy-Efficient
Multimedia Systems Group
Professor Vivienne Sze

DNN Benchmarking Form

Please download and complete this spreadsheet. Once completed, save and upload while completing the form below.

We welcome feedback on the benchmarking metrics and approaches.
Please contact us at dnn-benchmarking@mit.edu

Uploaded Completed XLS File*
Chip Name*
(text)
Publication Info*
(bibtex)
Contact Info*
(email)
Process Technology*
(nm) (volts)
Chip Area/multiplier*
(mm2)
On-Chip Memory/multiplier*
(kB)
Measured or Simulated*
- Measured
- Simulated
If Simulated, Synthesis or PnR*
- Synthesis
- PnR
- n/a
Please select which categories you support
Layer Types*
- CONV layer
- FC layer
AlexNet
- Dense
- Sparse
VGG16
- Dense
- Sparse
GoogleNet
- Dense
- Sparse
ResNet-50
- Dense
- Sparse