diff options
| author | Johan Alfvén <johan.alfven@arm.com> | 2022-06-10 15:40:58 +0200 |
|---|---|---|
| committer | Johan Alfvén <johan.alfven@arm.com> | 2022-06-20 08:47:44 +0200 |
| commit | 5c30971e1025eae428f6a74ca5f828919b2c34d4 (patch) | |
| tree | a697bd8478eeebcc63c4261615f63a4ce921fb74 /BUGS.md | |
| parent | 5cc4c76988eb72c89fe92e96ae670b1828fedf88 (diff) | |
| download | ethos-u-vela-5c30971e1025eae428f6a74ca5f828919b2c34d4.tar.gz | |
MLBEDSW-6347: Improved fast storage allocator
- The fast storage allocator only looked at tensor size, giving priority
to larger tensors. The problem with this method is that it does not
consider the actual read/write access of the tensor. So, a smaller
tensor size can cause higher memory transactions than a bigger one.
- The solution is to calculate the read/write access of the tensor and
add that score to the decision when deciding where to place the tensors.
Signed-off-by: Johan Alfven <johan.alfven@arm.com>
Change-Id: I59eb9bd3a44a0238b576cfd8f09ff27012b99070
Diffstat (limited to 'BUGS.md')
0 files changed, 0 insertions, 0 deletions