Skip to content

generator v2021.05.25

Compare
Choose a tag to compare
@carlushuang carlushuang released this 25 May 14:12
· 308 commits to develop since this release
f9937a2

This generator is mainly for a feature named "global memory access pattern", or gmap for short.

As the name suggested, gmap is used to dump the memory access pattern of input/weight/output tensor, for organized for each individual block, and for each individual read/write request.

This feature is controlled by an environment variable IGEMM_DUMP_GMAP, example to use this feature:

python3 igemm_codegen.py config/igemm_bwd_gtc_gfx908_nhwc_fp16.config
cd out/
IGEMM_DUMP_GMAP=1 ./conv_driver.exe    convfp16 -n 2 -c 1024 -H 40 -W 52 -k 512 -y 1 -x 1 -p 0 -q 0 -u 2 -v 2 -l 1 -j 1 -g 1 -t 1 -F 2  --in_layout NHWC --fil_layout NHWC --out_layout NHWC

Currently only support NHWC fwd/bwd, fp32/fp16. More layout precision support is to be added.