diff options
author | Tim Hall <tim.hall@arm.com> | 2021-07-19 21:22:46 +0100 |
---|---|---|
committer | Tim Hall <tim.hall@arm.com> | 2021-07-19 21:22:46 +0100 |
commit | daed1529848465aab221a30008f455f9fa03c8d4 (patch) | |
tree | d98ec1b3138616a45651abee63c843a87c646c18 /ethosu/vela/tflite_writer.py | |
parent | a8c003909f6cfdbe7d71d40196920dd9e01c2467 (diff) | |
download | ethos-u-vela-daed1529848465aab221a30008f455f9fa03c8d4.tar.gz |
MLBEDSW-4812: Deep speech performance block config update
Deep speech was exhibiting poor performance in its first three
layers due to poor SHRAM utilisation.
- Given a choice between multiple identical-cost block configs,
the allocator was choosing the first one it encountered. This
commit biases the choice towards blocks with a larger IFM
fetch area to improve SHRAM utilisation.
Signed-off-by: Tim Hall <tim.hall@arm.com>
Change-Id: I2ff18a13444b8812cb451a606ff692bf290e7d20
Diffstat (limited to 'ethosu/vela/tflite_writer.py')
0 files changed, 0 insertions, 0 deletions