diff options
author | Tim Hall <tim.hall@arm.com> | 2021-06-17 17:02:31 +0100 |
---|---|---|
committer | Tim Hall <tim.hall@arm.com> | 2021-06-17 17:02:31 +0100 |
commit | 789e6f3acd1a377dfba80aa18d513579fd33fc93 (patch) | |
tree | a3f44fafb91f26032c2d273aef6c602da05d4715 /ethosu/vela/npu_performance.py | |
parent | a5e8c1c3470409566723919e878b17297a52c54b (diff) | |
download | ethos-u-vela-789e6f3acd1a377dfba80aa18d513579fd33fc93.tar.gz |
vela: Improve block configuration and weight buffering algorithm
- Update block config selection to take into account partial
IFM fetches at edge of non-whole OFM block data.
- Change to scheduler depth slicing for networks in MLBEDSW-4637
for improved buffering. This helps general performance by buffering
larger depth slices.
- Bug fix for opt_max_schedule always being fitted to SRAM which
prevented the optimisation step running in some cases.
Signed-off-by: Tim Hall <tim.hall@arm.com>
Change-Id: I97642c5adec3bb684b1daabf2b81574c27d4eef2
Diffstat (limited to 'ethosu/vela/npu_performance.py')
-rw-r--r-- | ethosu/vela/npu_performance.py | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/ethosu/vela/npu_performance.py b/ethosu/vela/npu_performance.py index 5c61c7db..21b420bf 100644 --- a/ethosu/vela/npu_performance.py +++ b/ethosu/vela/npu_performance.py @@ -410,6 +410,7 @@ def _estimate_conv_cycles(arch, op_type: Op, faf_type: Op, query: PerformanceQue def measure_mem2mem_cycles(arch, from_mem_area, to_mem_area, to_transfer): from_cycles = to_transfer // arch.memory_bandwidths_per_cycle[from_mem_area] + from_cycles += arch.memory_latency[from_mem_area][BandwidthDirection.Read] to_cycles = to_transfer // arch.memory_bandwidths_per_cycle[to_mem_area] return max(from_cycles, to_cycles) |