Set affinity in benchmark runs

ci/run-benchmarks.sh

@@ -12,7 +12,9 @@ if [[ $# -ne 0 ]]; then
 fi
 
 # we use taskset to pin the process to a specific set of cores and their HTs to reduce benchmark result variance
-# this set is chosen to all be located well on the Threadripper 2920X CPU in our current CI benchmark system "gpuc2"
-bash /root/capture-backtrace.sh taskset -c 4-11,16-23 test/benchmarks \
+# this set is chosen to all be located on the second die of the Threadripper 2920X CPU in our current CI benchmark system "gpuc2"
+# since this is a very rare system with only 3 cores per CC, we use a hardcoded thread pinning strategy which places the 4 BE worker threads on HTs
+export CELERITY_THREAD_PINNING=7,8,9,10,11,22,23
+bash /root/capture-backtrace.sh taskset -c 6-11,18-23 test/benchmarks \
     --reporter celerity-benchmark-md::out=gpuc2_bench.md \
     --reporter celerity-benchmark-csv::out=gpuc2_bench.csv