diff options
-rwxr-xr-x | VA/dimes-hetsim-hbm-roofline.sh | 8 | ||||
-rwxr-xr-x | VA/dimes-hetsim-hbm.sh | 10 | ||||
-rwxr-xr-x | VA/dimes-hetsim-nmc.sh | 38 |
3 files changed, 31 insertions, 25 deletions
diff --git a/VA/dimes-hetsim-hbm-roofline.sh b/VA/dimes-hetsim-hbm-roofline.sh index f987f5a..348f0e6 100755 --- a/VA/dimes-hetsim-hbm-roofline.sh +++ b/VA/dimes-hetsim-hbm-roofline.sh @@ -4,7 +4,7 @@ cd baselines/cpu make -B NUMA=1 mkdir -p log/$(hostname) -fn=log/$(hostname)/$(date +%Y%m%d)-roofline +fn=log/$(hostname)/dimes-hetsim-roofline # upstream uses 167772160 * 2 * int32 == 2.5 GiB input and 1.25 GiB output for DPU version @@ -12,7 +12,7 @@ fn=log/$(hostname)/$(date +%Y%m%d)-roofline for i in $(seq 0 7); do echo "single-node execution ($i/8)" >&2 - parallel -j1 --eta --joblog ${fn}.${i}.joblog --header : \ + parallel -j1 --eta --joblog ${fn}.joblog --resume --header : \ ./va -i {input_size} -a {ram} -b {ram} -c {cpu} -t {nr_threads} -w 0 -e 5 \ ::: nr_threads $(seq 1 16) \ ::: cpu $i \ @@ -20,6 +20,4 @@ for i in $(seq 0 7); do ::: input_size 167772160 done -) > ${fn}.txt - -xz -f -v -9 -M 800M ${fn}.txt +) >> ${fn}.txt diff --git a/VA/dimes-hetsim-hbm.sh b/VA/dimes-hetsim-hbm.sh index 66456e5..a7c9ef4 100755 --- a/VA/dimes-hetsim-hbm.sh +++ b/VA/dimes-hetsim-hbm.sh @@ -4,7 +4,7 @@ cd baselines/cpu make -B NUMA=1 mkdir -p log/$(hostname) -fn=log/$(hostname)/$(date +%Y%m%d) +fn=log/$(hostname)/dimes-hetsim-hbm # upstream uses 167772160 * 2 * int32 == 2.5 GiB input and 1.25 GiB output for DPU version @@ -12,7 +12,7 @@ fn=log/$(hostname)/$(date +%Y%m%d) echo "single-node execution (1/2)" >&2 -parallel -j1 --eta --joblog ${fn}.1.joblog --header : \ +parallel -j1 --eta --joblog ${fn}.1.joblog --resume --header : \ ./va -i {input_size} -a {ram} -b {ram} -c {cpu} -t {nr_threads} -w 0 -e 5 \ ::: nr_threads 1 2 4 8 12 16 \ ::: cpu $(seq 0 7) \ @@ -21,13 +21,11 @@ parallel -j1 --eta --joblog ${fn}.1.joblog --header : \ echo "multi-node execution (2/2)" >&2 -parallel -j1 --eta --joblog ${fn}.2.joblog --header : \ +parallel -j1 --eta --joblog ${fn}.2.joblog --resume --header : \ ./va -i {input_size} -a {ram} -b {ram} -c {cpu} -t {nr_threads} -w 0 -e 40 \ ::: nr_threads 32 48 64 96 128 \ ::: cpu -1 \ ::: ram $(seq 0 15) \ ::: input_size 167772160 -) > ${fn}.txt - -xz -f -v -9 -M 800M ${fn}.txt +) >> ${fn}.txt diff --git a/VA/dimes-hetsim-nmc.sh b/VA/dimes-hetsim-nmc.sh index f622b47..fa9e975 100755 --- a/VA/dimes-hetsim-nmc.sh +++ b/VA/dimes-hetsim-nmc.sh @@ -1,17 +1,16 @@ #!/bin/bash mkdir -p log/$(hostname) baselines/cpu/log/$(hostname) -fn=log/$(hostname)/$(date +%Y%m%d) +fn=log/$(hostname)/dimes-hetsim-nmc # upstream uses 167772160 * 2 * int32 == 2.5 GiB input and 1.25 GiB output for DPU version run_benchmark_nmc() { local "$@" + set -e sudo limit_ranks_to_numa_node ${numa_rank} - if make -B NR_DPUS=${nr_dpus} NR_TASKLETS=${nr_tasklets} BL=10 WITH_ALLOC_OVERHEAD=1 WITH_LOAD_OVERHEAD=1 WITH_FREE_OVERHEAD=1; then - bin/host_code -w 0 -e 100 -i ${input_size} -x 1 - fi - return $? + make -B NR_DPUS=${nr_dpus} NR_TASKLETS=${nr_tasklets} BL=10 WITH_ALLOC_OVERHEAD=1 WITH_LOAD_OVERHEAD=1 WITH_FREE_OVERHEAD=1 + bin/host_code -w 0 -e 40 -i ${input_size} -x 1 } export -f run_benchmark_nmc @@ -20,7 +19,7 @@ export -f run_benchmark_nmc echo "NMC single-node operation (1/2)" >&2 -parallel -j1 --eta --joblog ${fn}.1.joblog --header : \ +parallel -j1 --eta --joblog ${fn}.1.joblog --resume --header : \ run_benchmark_nmc nr_dpus={nr_dpus} nr_tasklets=16 input_size={input_size} numa_rank={numa_rank} \ ::: numa_rank 0 1 \ ::: nr_dpus 64 128 256 512 768 1024 \ @@ -28,15 +27,13 @@ parallel -j1 --eta --joblog ${fn}.1.joblog --header : \ echo "NMC multi-node operation (2/2)" >&2 -parallel -j1 --eta --joblog ${fn}.2.joblog --header : \ +parallel -j1 --eta --joblog ${fn}.2.joblog --resume --header : \ run_benchmark_nmc nr_dpus={nr_dpus} nr_tasklets=16 input_size={input_size} numa_rank={numa_rank} \ ::: numa_rank any \ ::: nr_dpus 1536 2048 2304 \ ::: input_size 167772160 -) > ${fn}.txt - -xz -f -v -9 -M 800M ${fn}.txt +) >> ${fn}.txt cd baselines/cpu make -B NUMA=1 @@ -45,7 +42,7 @@ make -B NUMA=1 echo "CPU single-node operation (1/2)" >&2 -parallel -j1 --eta --joblog ${fn}.1.joblog --header : \ +parallel -j1 --eta --joblog ${fn}.1.joblog --resume --header : \ ./va -i {input_size} -a {ram} -b {ram} -c {cpu} -t {nr_threads} -w 0 -e 40 \ ::: ram 0 1 \ ::: cpu 0 1 \ @@ -54,13 +51,26 @@ parallel -j1 --eta --joblog ${fn}.1.joblog --header : \ echo "CPU multi-node operation (1/2)" >&2 -parallel -j1 --eta --joblog ${fn}.2.joblog --header : \ +parallel -j1 --eta --joblog ${fn}.2.joblog --resume --header : \ ./va -i {input_size} -a {ram} -b {ram} -c {cpu} -t {nr_threads} -w 0 -e 40 \ ::: ram 0 1 \ ::: cpu -1 \ ::: nr_threads 24 32 \ ::: input_size 167772160 -) > ${fn}.txt +) >> ${fn}.txt + +make -B NUMA=1 NUMA_MEMCPY=1 + +( + +echo "CPU single-node operation with setup cost" >&2 + +parallel -j1 --eta --joblog ${fn}.1.joblog --resume --header : \ + ./va -i {input_size} -a {ram} -b {cpu} -c {cpu} -t {nr_threads} -w 0 -e 40 \ + ::: ram 0 1 \ + ::: cpu 0 1 \ + ::: nr_threads 1 2 4 8 12 16 \ + ::: input_size 167772160 -xz -f -v -9 -M 800M ${fn}.txt +) >> ${fn}.txt |