Skip to content

Commit cef44f1

Browse files
committed
Make it clear that we talk about fio scores at one point
1 parent 0ea9cc3 commit cef44f1

File tree

1 file changed

+6
-6
lines changed

1 file changed

+6
-6
lines changed

doc/final-report/final-report.md

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -992,12 +992,12 @@ benchmark may give too rosy a picture if the physical machine was not saturated
992992
at the time the benchmark was run, or the benchmark was not run long enough, or
993993
the benchmark does not use the same amount of I/O parallelism as the deployment.
994994

995-
Finally, the dev laptop’s scores of 262 k IOPS on 1 core and 487 k on 2 cores
996-
are well below the laptop’s rated 1.4 M IOPS. The explanation is that the 1.4 M
997-
is for the SSD device as a whole and it takes multiple cores to saturate the
998-
SSD, while the benchmark numbers above are for a single core. On the same
999-
machine, the scores of the `fio` and `blockio-uring` benchmarks for 8 cores are
1000-
1,449 k and 1,337 k IOPS respectively.
995+
Finally, the dev laptop’s `fio` scores of 262 k IOPS on 1 core and 487 k on
996+
2 cores are well below the laptop’s rated 1.4 M IOPS. The explanation is that
997+
the 1.4 M is for the SSD device as a whole and it takes multiple cores to
998+
saturate the SSD, while the benchmark numbers above are for a single core. On
999+
the same machine, the scores of the `fio` and `blockio-uring` benchmarks for
1000+
8 cores are 1,449 k and 1,337 k IOPS respectively.
10011001

10021002
## Setup of the primary benchmark
10031003

0 commit comments

Comments
 (0)