Revert "MLBEDSW-6312: Find block config improvement"
This reverts commit d2b5510697e7789f5a416f9d80d3cb640eecc092.
Signed-off-by: Tim Hall <tim.hall@arm.com>
Change-Id: Ia3043bc9c27fe2f72f3ab2f6f7341b3a9adb4231
diff --git a/ethosu/vela/architecture_allocator.py b/ethosu/vela/architecture_allocator.py
index d27f126..e7e7471 100644
--- a/ethosu/vela/architecture_allocator.py
+++ b/ethosu/vela/architecture_allocator.py
@@ -53,7 +53,6 @@
self.acc_type = SHRAMElements.Acc32
self.is_partkernel = False
self.bank_size = 0
- self.ifm_depth_buf_scaling = 0
def get_shram_memory_access_range(self):
# Returns the SHRAM memory access range used by this shared buffer,
@@ -86,18 +85,12 @@
acc_bits: int,
acc_granule: int,
lut_banks: int,
- ifm_depth_buf_scaling: int,
- cores: int,
) -> Union[SHRAMLayout, None]:
assert (acc_bits > 0) and (acc_granule > 0)
assert (ifm_bits >= 8) and ((ifm_bits % 8) == 0) and (ifm_granule > 0)
- # Scale depth with cores
- ifm_depth = round_up_divide(ifm_block.depth, ifm_depth_buf_scaling)
- ofm_depth = round_up_divide(ofm_block.depth, cores)
-
# Aways need IFM space
- ifm_bytes = ifm_block.elements_wh() * round_up((ifm_depth * ifm_bits) / 8, 8)
+ ifm_bytes = ifm_block.elements_wh() * round_up((ifm_block.depth * ifm_bits) / 8, 8)
ifm_banks = round_up_divide(ifm_bytes, shram.bank_size_bytes) * 2
ifm_banks = round_up(ifm_banks, ifm_granule)
@@ -109,7 +102,7 @@
# If not elementwise then we need accumulator space
if ew_usage == ElementwiseUsage.No:
- acc_bytes = (ofm_block.elements_wh() * round_up(ofm_depth, 8) * acc_bits) // 8
+ acc_bytes = (ofm_block.elements_wh() * round_up(ofm_block.depth, 8) * acc_bits) // 8
acc_banks = round_up_divide(acc_bytes, shram.bank_size_bytes) * 2
acc_banks = round_up(acc_banks, acc_granule)
acc_start = acc_start - acc_banks
@@ -259,14 +252,6 @@
config = ArchitectureBlockConfig()
config.is_partkernel = is_convolution and _choose_kernel_method(ifm_shape, ifm_bits, kernel)
- # IFM is not broadcasted for pooling and depthwise ops and for elementwise
- # when there's no elementwise-broadcasting in depth
- elemwise_buf_scalable = npu_op_type == NpuBlockType.ElementWise and (
- not ifm2_shape or ifm_shape.depth == ifm2_shape.depth
- )
- ifm_depth_buf_scaling = arch.ncores if is_pooling or is_depthwise or elemwise_buf_scalable else 1
- config.ifm_depth_buf_scaling = ifm_depth_buf_scaling
-
# Accumulator & granule settings
config.acc_type = _acc_type(npu_op_type, ifm_bits, scaled)
@@ -290,9 +275,8 @@
# Weights fetch (for operators that have them)
weight_fetch_wh = (kernel.area_width() * kernel.area_height()) if is_convolution else 0
- ofm_ublock_depth = arch.ofm_ublock.depth * arch.ncores
search_space = Shape4D.min(ofm_shape, Shape4D(arch.ofm_block_max.to_hwc()))
- search_space = Shape4D.round_up(search_space, Shape4D(arch.ofm_ublock.to_hwc()).with_depth(ofm_ublock_depth))
+ search_space = Shape4D.round_up(search_space, Shape4D(arch.ofm_ublock.to_hwc()))
# Block WHC search, loops across the search space looking for best efficiency
best_cost = math.inf
@@ -328,8 +312,6 @@
acc_bits,
acc_granule,
lut_banks,
- ifm_depth_buf_scaling,
- arch.ncores,
)
if layout:
@@ -427,14 +409,6 @@
config = ArchitectureBlockConfig()
config.is_partkernel = is_partkernel
- # IFM is not broadcasted for pooling and depthwise ops and for elementwise
- # when there's no elementwise-broadcasting in depth
- elemwise_buf_scalable = npu_op_type == NpuBlockType.ElementWise and (
- not ifm2_shape or ifm_shape.depth == ifm2_shape.depth
- )
- ifm_depth_buf_scaling = arch.ncores if is_pooling or is_depthwise or elemwise_buf_scalable else 1
- config.ifm_depth_buf_scaling = ifm_depth_buf_scaling
-
# Accumulator & granule settings
config.acc_type = _acc_type(npu_op_type, ifm_bits, scaled)
@@ -457,17 +431,7 @@
block_config_opt = fit_block_for_ofm(arch, ofm_shape, kernel, block_config)
layout = _try_block_config(
- arch.shram,
- ew_usage,
- block_config_opt,
- ifm_block,
- ifm_bits,
- ifm_granule,
- acc_bits,
- acc_granule,
- lut_banks,
- ifm_depth_buf_scaling,
- arch.ncores,
+ arch.shram, ew_usage, block_config_opt, ifm_block, ifm_bits, ifm_granule, acc_bits, acc_granule, lut_banks
)
if layout is None:
return None