Block config optimisation for 256/512 configurations
- 256 and 512 configuration variants execute 1D convolutions
in an optimised manner compared to their 2x2 microblock
dimensions. This commit takes this into account to improve
Conv1D throughput on these configurations.
Signed-off-by: Tim Hall <tim.hall@arm.com>
Change-Id: I6ecdf6e4a219e356327b22f8393f50ee8817af23
diff --git a/ethosu/vela/test/test_architecture_allocator.py b/ethosu/vela/test/test_architecture_allocator.py
index 94768fc..a199e9d 100644
--- a/ethosu/vela/test/test_architecture_allocator.py
+++ b/ethosu/vela/test/test_architecture_allocator.py
@@ -106,6 +106,7 @@
Block.from_shape(config.ofm_block.as_list()),
arch,
block_type,
+ ofm_shape,
ifm_shape,
ifm2_shape,
is_partkernel=config.is_partkernel,