Repository navigation
Annotations
5 notices
|
Summarize comparison
qd1 iops 15,962 -> 15,664 baseline ratio 0.98, latency 39 us -> 40 us; qd8 iops 41,547 -> 41,352 baseline ratio 1.00, latency 170 us -> 171 us; qd32 iops 46,106 -> 47,548 baseline ratio 1.03, latency 675 us -> 655 us; qd64 iops 18,638 -> 53,150 baseline ratio 2.85, latency 3,420 us -> 1,188 us; qd128 iops 13,077 -> 66,171 baseline ratio 5.06, latency 9,773 us -> 1,923 us; qd256 iops 9,285 -> 82,983 baseline ratio 8.94, latency 27,547 us -> 3,077 us
|
|
Benchmark native Photon WorkPool implementation
offset=138448896 bytes, size=12546048 bytes, selected-file allocated coverage=96.74%
|
|
Benchmark native Photon WorkPool implementation
AVX-512 accelerated linearized B+tree search not observed
|
|
Benchmark baseline implementation
offset=138448896 bytes, size=12546048 bytes, selected-file allocated coverage=96.74%
|
|
Benchmark baseline implementation
AVX-512 accelerated linearized B+tree search not observed
|
background
wait
wait-all
cancel
parallel
Loading