Hi! I am running the stream benchmark[1] in GEM5 in SE, like this
./build/X86/gem5.opt --debug-flags=ExecTicks ./configs/example/se.py --mem-type=SimpleMemory -n 1 -c /gem5tests/stream The bandwidth reported is 2000 MB/s. Now if I run it on two CPUs in parallel, like this ./build/X86/gem5.opt --debug-flags=ExecTicks ./configs/example/se.py --mem-type=SimpleMemory -n 2 -c /gem5tests/stream;/gem5tests/stream the reported bandwidth is still ~2000 MB/s, reported by both processes! I expected it to be in the order of 1000 MB. The same happens with different memory types, etc. Can someone explain why the memory bandwidth is not halved? Thank you! _______________________________________________ gem5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
